Omega‐3 fatty acids for intermittent claudication

Abstract

Background

Peripheral artery disease (PAD) is a progressive disorder characterised by stenosis or occlusion of arteries, or both, due to arteriosclerosis. Intermittent claudication (IC) and diminished walking ability are often present as the main symptoms of PAD. Omega‐3 fatty acids have been used in the treatment and prevention of coronary artery disease, although current evidence suggests they may be of limited benefit. Peripheral arterial disease and coronary artery disease share a similar pathogenesis. It is uncertain whether omega‐3 fatty acids benefit people with IC. This is an update of the review first published in 2004 and updated in 2013.

Objectives

To evaluate the benefits and harms of omega‐3 fatty acid supplementation in people with intermittent claudication.

Search methods

We used standard, extensive Cochrane search methods, and searched the Cochrane Vascular Specialised Register via the Cochrane Register of Studies, CENTRAL, MEDLINE Ovid, Embase Ovid, and two trials registers on 19 April 2024.

Selection criteria

We included randomised controlled trials (RCTs) of omega‐3 fatty acids versus placebo or non‐omega‐3 fatty acids in people with intermittent claudication.

Data collection and analysis

We used standard Cochrane methods. Our primary outcomes were quality of life, pain‐free walking distance, and maximal walking distance. Secondary outcomes were ankle‐brachial index, revascularisation procedures in the lower limb, amputation rate/frequency, lipid levels, blood pressure, all‐cause and vascular mortality, non‐fatal vascular events, and adverse effects of therapy. We used GRADE to assess the certainty of the evidence for each outcome.

Main results

We included 15 RCTs (1830 participants) comparing omega‐3 fatty acid supplementation with placebo or alternative therapies. The follow‐up was four weeks to six years. The majority of the studies had unclear risk of bias, and many could not be included in our meta‐analysis, so were reported narratively.

The evidence is very uncertain about the effect of omega‐3 fatty acids on quality of life. One study measured quality of life but did not present any data. The study authors reported there was no improvement in any of the eight self‐reported quality‐of‐life parameters in the SF‐36 questionnaire between entry and 16 weeks for the intervention group. No results were presented for the control group (very low‐certainty evidence). Omega‐3 fatty acids may result in little to no effect on pain‐free walking distance (mean difference (MD) 1.01 metre (m), 95% confidence interval (CI) ‐34.23 to 36.24; 3 studies, 147 participants; very low‐certainty evidence), or maximal walking distance (MD ‐4.18 m, 95% CI ‐37.10 to 28.74; 3 studies, 164 participants; very low‐certainty evidence).

Omega‐3 compared with a control may have little to no effect on ankle‐brachial index (MD ‐0.02, 95% CI ‐0.08 to 0.04; 3 studies, 168 participants; very low‐certainty evidence). One study assessed the incidence of revascularisation procedures (lower limb angioplasty/bypass surgery) and rate of amputation (progression of critical limb ischaemia/amputation) in the lower limb. Results showed that omega‐3 may have little to no effect on either outcome (very low‐certainty evidence).

Seven studies reported adverse events. Details of reporting varied amongst studies, and we were unable to combine the results. A total of 47 adverse effects were reported in the intervention groups compared to 33 events in the control groups (7 studies, 488 participants; low‐certainty evidence). The evidence suggests that omega‐3 results in little to no difference in adverse events.

Meta‐analyses showed no differences between intervention and placebo groups for cholesterol, triglycerides, or blood pressure.

Two studies assessed mortality. All‐cause mortality and vascular mortality were reported by one study, and vascular mortality by another. We were unable to pool the studies, but both studies individually reported there were no differences between the omega‐3 and the control groups.

There was no difference between the intervention and placebo groups for the incidence of non‐fatal coronary events (odds ratio (OR) 0.59, 95% CI 0.13 to 2.60; 2 studies, 141 participants), or the incidence of non‐fatal stroke/transient ischaemic attack (OR 0.95, 95% CI 0.13 to 6.77; 2 studies, 110 participants).

Authors' conclusions

The evidence is very uncertain about the effect of omega‐3 fatty acids in people with intermittent claudication on quality of life, walking distance (pain‐free or maximal), ankle‐brachial index, and the incidence of revascularisation procedures or frequency of amputation in the lower limb. The evidence suggests that omega‐3 results in little to no difference in adverse events.

Further high‐quality research is needed to fully evaluate short‐ and long‐term effects of omega‐3 fatty acids on the most clinically relevant outcomes in people with intermittent claudication.

Plain language summary

What are the benefits and risks of omega‐3 fatty acids for a limp due to blood restriction (intermittent claudication)?

Key messages

We are very uncertain whether taking omega‐3 fatty acid supplements has any effect on the quality of life, walking distance, or blood flow to the legs of people with intermittent claudication.

We are very uncertain whether taking omega‐3 fatty acid supplements has any effect on blood pressure, or blood levels of different types of cholesterol or triglycerides (build‐up of fatty material).

Taking omega‐3 may not influence fatal or non‐fatal circulation problems or other side effects, but we are very uncertain about the results.

More, well‐designed studies with more people are required to help understand whether omega‐3 supplements carry any benefits or risks.

What is intermittent claudication?

Intermittent claudication is pain in the calf of the leg due to a lack of blood needed to supply those muscles with oxygen during exercise or movement, which causes a person to slow or stop moving. It is the most common symptom reported by people with long‐standing lower limb arterial disease, which results when the arteries that supply blood to the lower limbs are narrowed. This narrowing most commonly occurs with atherosclerosis, a result of the accumulation of fatty materials, such as cholesterol and triglycerides.

How is intermittent claudication treated?

People with lower limb arterial disease are advised to stop smoking, control their blood glucose, participate in structured exercise treatment, take special medicines, and undergo more complex treatments, such as angioplasty (unblocking of a blood vessel) or surgical repair.

What did we want to find out?

We wanted to find out if omega‐3 fatty acid supplements were better than dummy treatments (that do not contain any omega‐3) or alternative treatments to improve:

· Quality of life

· Walking distance

· Blood flow in the legs

· Cholesterol and triglyceride levels

· Blood pressure

· Number of angioplasties (opening blocked or narrowed arteries) or surgical repairs

We also wanted to find out if omega ‐3 was associated with any unwanted side effects.

What did we do?

We searched for studies that examined omega‐3 fatty acids compared with dummy treatments or alternative therapies in people with intermittent claudication. We wanted to find out if omega‐3 supplements were better than dummy treatments or alternative therapies to improve the outcomes.

We compared and summarised the results of the studies and rated our confidence in the evidence, based on factors, such as study methods and sizes.

What did we find?

We found 15 studies involving 1830 people with symptoms of intermittent claudication, lower limb arterial disease, or both. There were 971 people in the largest study and 18 in the smallest. Study locations included the United Kingdom, Australia, the Netherlands, Switzerland, the United States, and Spain. The average age of people who participated was between 62 and 69 years. Four studies only recruited men.

Omega‐3 fish oil supplements were used in 11 studies, and foods containing omega‐3 were used in the other four studies. Different dosages were used. Studies lasted between four weeks and approximately six years. All except one used a dummy treatment for comparison. Some were funded by the manufacturers of omega‐3 supplements and food.

Main results

Compared with no omega‐3 supplements:

Omega‐3 supplements may have little to no effect on quality of life, walking distance, blood flow in the leg, or the amount of angioplasty, surgical repair, or amputation, but we are very uncertain about the results.

Omega‐3 may have little to no effect on cholesterol or triglyceride blood levels, or blood pressure.

There did not appear to be any difference in fatal or non‐fatal circulation‐related problems between people taking omega‐3 and those not taking omega‐3, but we are very uncertain about the results.

Only 7 studies reported whether people experienced unwanted side effects, such as an upset stomach or headache. The evidence suggests there is little to no difference between those who took omega‐3 and those who did not take omega‐3 for side events.

What are the limitations of the evidence?

Our confidence in the evidence for each of the most important outcomes was very low to low. This was because of the way the studies were designed, the small numbers of people involved, and the differences in results between studies. Many studies also reported their results in a way that meant we could not include them in our statistical analyses.

How up‐to‐date is this evidence?

The evidence is up‐to‐date to 19 April 2024.

Summary of findings

Summary of findings 1. Supplementation with omega‐3 fatty acids compared to placebo for people with intermittent claudication.

Omega‐3 fatty acids compared to placebo for intermittent claudication
Patient or population: people with intermittent claudication Setting: outpatient clinics Intervention: omega‐3 fatty acids Comparison: control (placebo or active control)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with control	Risk with omega‐3 fatty acids
Quality of life (SF‐36) follow‐up: 16 weeks	‐	‐	‐	40 (1 RCT)	⊕⊝⊝⊝ VERY LOW a	Numerical values were not stated in the paper for control or intervention groups. P values were only reported with no improvement found in the intervention group (P > 0.05).
Pain‐free walking distance (metres, higher values are better) follow‐up: 16 weeks to 12 months	The mean PFWD in the control group ranged from 81.5 to 127 metres	The mean PFWD in the intervention group was 1.01 metres more (34.23 metres fewer to 36.2 metres more)	‐	147 (3 RCTs)	⊕⊝⊝⊝ VERY LOW b	Omega‐3 may have little to no effect on pain‐free walking distance, but the evidence is very uncertain.
Maximal walking distance (metres, higher values are better) follow‐up: 16 weeks to 12 months	The mean MWD in the control group ranged from 143.6 metres to 166 metres	The mean MWD in the intervention group was 4.18 metres fewer (37.10 metres fewer to 28.74 metres more)	‐	164 (3 RCTs)	⊕⊝⊝⊝ VERY LOWc	The evidence is very uncertain about the effect of omega‐3 on maximal walking distance.
Ankle‐brachial index (ratio: blood pressure in lower limb to blood pressure in upper limb, higher value is better) follow‐up: 16 weeks to 12 months	The mean ABI in the control group ranged from 0.58 to 0.81	The mean ABI in the intervention group was 0.02 lower (0.08 lower to 0.04 higher)	‐	168 (3 RCTs)	⊕⊝⊝⊝ VERY LOWd	Omega‐3 may result in little to no difference in ABI, but the evidence is very uncertain.
Revascularisation procedures in lower limb	1 RCT reported 1 event in the intervention group versus 3 in the control group for lower limb angioplasty/bypass surgery		‐	120 ( 1 RCT)	⊕⊝⊝⊝ VERY LOWe	We are unable to draw any conclusions about this, as any effects in this study cannot solely be attributed to omega‐3.
Amputation rate/frequency	1 RCT reported 1 event in the intervention group versus 0 in the control group for progression of critical limb ischaemia/ amputation		‐	120 ( 1 RCT)	⊕⊝⊝⊝ VERY LOWe	We are unable to draw any conclusions about this, as any effects in this study cannot solely be attributed to omega‐3.
Adverse effects	A total of 47/245 adverse effects occurred in the intervention groups. A total of 33/243 adverse effects occurred in the control groups			488 (7 RCT)	⊕⊕⊝⊝ LOWf
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). ABI: ankle brachial index; CI: confidence interval; MWD: maximal walking distance; PFWD: pain‐free walking distance; RCT: randomised controlled trial
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aWe downgraded the certainty of evidence by two levels for serious risk of bias limitations (performance and attrition bias), and twice for imprecision (one study with very low number of participants and uncertain data).
^b We downgraded the certainty of evidence by three levels due to risk of bias limitations (performance), imprecision (three studies with small numbers of participants), and inconsistency (heterogeneity, I² = 70%).
^c We downgraded the certainty of evidence by three levels due to risk of bias limitations (performance) and imprecision (three studies with small numbers of participants), and inconsistency (heterogeneity, I² = 52%).
^d We downgraded the certainty of evidence by two levels for serious risk of bias limitations (performance) and twice for imprecision (three studies with small numbers of participants, and the CI of the point estimate includes both benefit and harm).
^e We downgraded the certainty of evidence by three levels for serious risk of bias limitations (detection, attrition, and other bias) and imprecision (one study with small number of participants).
^f We downgraded the certainty of evidence by two levels for serious risk of bias limitations (selection, detection, attrition and other bias).

Background

Description of the condition

Peripheral artery disease (PAD) is a progressive disorder characterised by stenosis or occlusion of arteries, or both, due to arteriosclerosis (Campia 2019). PAD is an important cause of amputation worldwide. It significantly affects a person's quality of life, and is associated with a greatly increased risk of major cardiovascular disease (CVD) events and mortality (Fowkes 2017).

Smoking, diabetes, hypertension, and hypercholesterolaemia are major risk factors for PAD. In 2015, 236.62 million people globally, aged 25 and older, were living with this disorder, including 72.91% in low‐ and middle‐income countries (Song 2019). This health condition also becomes more common with advancing age. In 2021, the prevalence rate in the USA amongst those aged 40 and older was 4.3% (95% confidence interval (CI) 3.1% to 5.5%), rising to 14.3% (93% CI 10.8% to 18.2%) amongst individuals aged 70 and older. This corresponds to approximately four million people (Wing 2021).

People suffering from PAD are at higher risk of cardiovascular events and limb‐related complications, such as ischaemia or amputations (Jindeel 2017). This condition also affects the arteries in the lower extremity more than the upper one (Shu 2018). Symptoms of PAD include intermittent claudication and pain at rest, which is indicative of critical limb ischaemia. About 50% of those with PAD will be asymptomatic. People with pain at rest and tissue loss, referred to as critical limb ischaemia, will sometimes undergo an amputation. Intermittent claudication (IC) and diminished walking ability have been presented as the main symptoms of PAD (Hamburg 2017; Kirby 2023). Notably, IC is defined as activity‐induced dysfunction and ischaemia‐induced muscle pain relieved by rest (Shu 2018), which typically occurs when there is an inadequate supply of skeletal muscle perfusion during exercise (Hamburg 2017).

Peripheral artery disease is further related to a significant 10‐year risk of morbidity and mortality. Approximately 25% of people with PAD develop worsening claudication, 5% of the cases require amputations, 10% to 20% of affected individuals are in need of revascularisation, and 30% of sufferers die from cardiovascular events (such as heart attack and stroke) as a result of concomitant coronary artery disease (CAD), cerebrovascular atherosclerosis, or both (Benjamin 2016). A history of claudication by itself, compared to no claudication in the lower limbs, only slightly augments the risk of amputations after 10 years. Nevertheless, amongst people with PAD, those living with diabetes are 15 times more likely to undergo amputations (Wing 2021).

The ankle‐brachial index (ABI) is used to verify PAD (Gerhard‐Herman 2017). This is a non‐invasive measurement of arterial occlusion pressure, using a Doppler ultrasound. To enhance the specificity of the given index, 0.9 is typically used as the cut‐off point for the diagnosis of PAD. ABI is often recommended for people with PAD as an appropriate parameter to assess cardiovascular risk, for morbidity and mortality (Frank 2019).

Management of IC includes targeting risk factors. Interventions include smoking cessation, glycaemic control, structured exercise therapy, drug therapy (e.g. cilostazol, antiplatelet, statin, and antihypertensive drugs), and more complex interventions, such as balloon dilation (angioplasty), stents, atherectomy, and bypass surgery (Gerhard‐Herman 2017).

Omega‐3 fatty acids have been used in the treatment and prevention of coronary artery disease, although current evidence suggests they may be of limited benefit (Abdelhamid 2020; Demaison 2002; Kinsella 1990). PAD and CVD share similar pathogenesis and risk factors. It is uncertain whether omega‐3 fatty acids may benefit people with PAD and IC.

Description of the intervention

Omega‐3 polyunsaturated fatty acids (PUFAs) consist of longer and shorter chain omega‐3 fats. The longer chain omega‐3 fats (LCn3) also contain eicosapentaenoic acid (EPA, 20:5), docosahexaenoic acid (DHA, 22:6), and docosapentaenoic acid (DPA, 22:5 (Ishihara 2019)). EPA and DHA are known as the active forms of the omega‐3 family, and are found in seafood, especially fatty fish, many supplements, and concentrated pharmaceutical preparations (Innes 2020). DPA, another LCn‐3 and a metabolite of DHA, is further thought to be generated through internal metabolic pathways rather than dietary intake (Mohebi‐Nejad 2014). Notably, the human body cannot efficiently produce some n‐3 FAs (EPA and DHA), but alpha‐linolenic acid (ALA, 18:3), as a shorter chain n‐3 FA in plants, such as flaxseeds, canola oil, and walnuts, can be partially converted into LCn‐3 within our body (Blondeau 2015).

How the intervention might work

N‐3 FAs, such as EPA and DHA, can have multiple biological effects. The proposed mechanisms for the atheroprotective effects of n‐3 FAs against cardiovascular diseases include reducing platelet‐mediated thrombin generation, along with attenuating the elements of thrombus formation (Adili 2019), improving endothelial function and arterial stiffness with parallel anti‐inflammatory effects (Tousoulis 2014), altering lipid profiles, especially lowering serum triglyceride concentrations (Abdelhamid 2020), modulating T‐cell differentiation, and giving rise to various prostaglandins and specialised pro‐resolving lipid mediators that promote the resolution of tissue injury and inflammation (Mason 2020). At a molecular level, n‐3 FAs are assumed to achieve their effects through different molecular pathways, including alterations in the physical and chemical properties of cellular membranes, direct interactions with and the modulation of membrane channels and proteins, gene expression regulation via nuclear receptors and transcription factors, some changes in eicosanoid profiles, and the conversion of the n‐3 FAs into bioactive metabolites (Mozaffarian 2011).

Why it is important to do this review

Atherosclerosis is a systemic disorder with a similar risk profile to CVD, independent of the specific arterial system involved. For this reason, the biological effects of n‐3 FAs may be potentially beneficial to people with PAD. However, evidence to support this issue is uncertain, and recent guidelines do not recommend the use of FAs for the prevention of cardiovascular events in people with PAD (Frank 2019).

This is an update of a Cochrane review first published in 2004 (Sommerfield 2004), and previously updated in 2013 (Campbell 2013). Campbell and colleagues included nine studies with 425 participants, and found that omega‐3 fatty acids had little haematological advantage in individuals with IC, and may not improve clinical outcomes (quality of life, walking distance, ankle brachial pressure index, or angiographic findings). This review update aimed to identify studies investigating n‐3 FAs in people with IC to determine the available evidence of any benefits and harms.

Objectives

To evaluate the benefits and harms of omega‐3 fatty acid supplementation in people with intermittent claudication.

Methods

Criteria for considering studies for this review

Types of studies

We included all randomised controlled trials (RCTs) of omega‐3 fatty acid (n‐3 FA) supplementation versus placebo or alternative therapies. We included cross‐over RCTs, but excluded quasi‐RCTs and cluster‐RCTs.

Types of participants

We included trials involving adults (18 years of age or older) with intermittent claudication (IC) secondary to atherosclerotic disease. If a study included participants both with and without IC, and participants with IC were more than 80% of the total participants, we planned to include the study; if participants with IC were less than 80%, we planned to exclude the study.

Intermittent claudication is one of the first symptoms that presents in PAD. If a person has this symptom, he or she may also experience other, more severe symptoms, such as pain at rest and tissue loss. Accordingly, we did not exclude studies with participants who experienced other, more severe symptoms of PAD, such as pain at rest and tissue loss.

We excluded studies involving participants with acute lower‐extremity ischaemia, or those who had undergone lower‐extremity vascular surgery or angioplasty within the previous three months. This was because the effects of surgery might make it difficult to attribute any changes in their health condition to n‐3 FA supplementation.

Types of interventions

We included studies in which the intervention was any form, dose, or duration of n‐3 FA supplementation. Supplementation could include FA capsules, powders, or dietary manipulations, such as oily fishes, fish oils (made from oily fishes or a mixture of fish or cod liver oil), flaxseeds, canola oil, common purslane, mustard seeds, or walnuts as a food, capsule, or oil, or foods enriched with n‐3 FA (such as bread or eggs).

We included studies that compared n‐3 FA supplementation with placebo, fats without n‐3 FA (e.g. olive, corn, sunflower oils, and other types of fats), no intervention or typical dietary advice, or foods without n‐3 FA enrichment.

Types of outcome measures

The primary and secondary outcomes listed below did not determine the eligibility of studies for our review. For studies with multiple outcome measurements, we selected the longest follow‐up from each study.

Primary outcomes

• Quality of life: measured using the Short Form (SF)‐36 questionnaire and other validated measurements, such as the EuroQoL, the Group Quality of Life Questionnaire based on five dimensions (EQ‐5D), or the Vascular Quality of Life Questionnaire (VascuQol)

• Pain‐free walking distance (PFWD)

• Maximal walking distance (MWD)

Secondary outcomes

• Ankle‐brachial index (ABI)

• Revascularisation procedures in the lower limb

• Amputation rate/frequency

• Lipid levels: total cholesterol, high‐density lipoprotein (HDL) cholesterol, low‐density lipoprotein (LDL) cholesterol, very low‐density lipoprotein (VLDL) cholesterol, and triglycerides

• Blood pressure: systolic and diastolic

• All‐cause and vascular mortality

• Non‐fatal vascular events

• Adverse effects of therapy: any, such as skin rash, dizziness, headache, gastrointestinal upset

Search methods for identification of studies

Electronic searches

The Cochrane Vascular Information Specialist conducted systematic searches of the following databases for randomised controlled trials and controlled clinical trials, without language, publication year or publication status restrictions.

• Cochrane Vascular Specialised Register (in the Cochrane Register of Studies (CRSO); searched 19 April 2024)

• Cochrane Central Register of Controlled Trials (CENTRAL; 2024, Issue 3), in the CRSO (searched 19 April 2024)

• MEDLINE (Ovid MEDLINE® Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE®; 1946 to 19 April 2024)

• Embase Ovid (1972 to 19 April 2024)

• CINAHL EBSCO (1982 to 19 April 2024)

The Information Specialist modelled search strategies for other databases on the search strategy designed for CENTRAL. Where appropriate, they were combined with adaptations of the highly sensitive search strategy, designed by Cochrane to identify randomised controlled trials and controlled clinical trials (Lefebvre 2023). Search strategies for major databases are provided in Appendix 1.
The information specialist also searched for the following trials registries on 19 April 2024.

• World Health Organization International Clinical Trials Registry Platform (who.int/trialsearch)

• US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (ClinicalTrials.gov)

Searching other resources

For this update, we contacted the authors of the newly included studies, and reviewed the reference lists of the newly included studies to identify other potentially relevant studies. We searched literature from the websites of Roche Pharmaceutical (Roche 2017), and Seven Seas Pharmaceuticals (Seven Seas 2007), manufacturers of omega‐3 rich foods (Deans Food 2007), and websites of nutritional organisations dedicated to omega‐3 fatty acids for relevant references (Fish Foundation 2007; Omega‐3 2007).

Data collection and analysis

Selection of studies

We imported the results of the searches into Covidence (Covidence), and two review authors (MM, MR) independently reviewed the abstracts to identify trials that potentially met the inclusion criteria. If it was not feasible to include or exclude a study on the basis of its title or abstract, we retrieved the full‐text article. The same two review authors (MM, MR) then independently assessed the full‐text reports to determine if they met the inclusion criteria. Any conflicts were resolved by a third review author (TB). The main authors of the RCTs were contacted to request additional information if the reports did not provide sufficient information to determine whether the study should be included in this review. The study selection process is illustrated in the PRISMA flow diagram in Figure 1.

1.

PRISMA flow diagram

Data extraction and management

Two review authors (LJ, ES) independently reviewed each study, and extracted the following information, using a piloted data extraction form: country where the study was undertaken; number, age, and sex of participants; inclusion criteria; exclusion criteria; treatment; control; duration; outcomes, study funding, and declarations of interest. Disagreements were resolved by a third review author (MM), if needed. When necessary, we contacted the main authors of the RCTs to request additional information. If a study was not published in English, we translated it into English with Google Translate, and then two review authors independently extracted the data.

Assessment of risk of bias in included studies

Two review authors (MM, LJ) independently assessed the risk of bias in the included studies using Cochrane's RoB 1 tool. Any disagreements were resolved by discussions with a third review author (TB), if necessary. We contacted the authors of the RCTs if insufficient information was provided in their publications. The risk of bias in each study was assessed, following the guidance in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2023). We assessed the following domains of trial quality.

• random sequence generation (selection bias)

• allocation concealment (selection bias)

• blinding of participants and personnel (performance bias)

• blinding of outcome assessment (detection bias)

• incomplete outcome data (attrition bias)

• selective reporting (reporting bias)

• other biases

We gave trials a rating of low risk, unclear risk, or high risk of bias for each of these domains. We also defined an overall summary judgement for each study as:

• low summary risk of bias for all key domains (at low risk of selection bias, performance bias, and detection bias);

• unclear summary risk of bias for one or more key domains (at unclear risk of selection bias, performance bias and detection bias);

• high summary risk of bias for one or more key domains (at high risk of selection bias, performance bias, and detection bias).

Measures of treatment effect

We performed statistical analyses according to the statistical guidelines in the Cochrane Handbook (Higgins 2023), using Review Manager (RevMan 2024).

Since most of the data were continuous, we extracted mean and standard deviation (SD) values for each outcome, and analysed the mean difference (MD), with 95% confidence intervals (CI) as the summary statistic. If appropriate, we pooled results using a meta‐analysis. Some papers presented standard errors (SE) rather than SDs. Therefore, we converted these data to SDs, using the formula SD = √N x SE. Most of the studies reported post‐intervention data. The meta‐analysis was based on post‐intervention data. Studies that reported mean changes from baseline were presented narratively (McKenzie 2023). We used the odds ratio (OR) with 95% CI to pool dichotomous outcomes.

Unit of analysis issues

The unit of analysis was the individual participant in all studies included in this review. We included one study with a cross‐over design, and we used the data from the first phase only (Mackay 2012).

To avoid unit‐of‐analysis errors in future updates, we will address them as follows:

• If multi‐arm studies are included, we will try to combine relevant intervention groups into a single group, or separate comparisons into different forest plots, to avoid arbitrary omission of relevant groups and double‐counting of participants;

• For events that may re‐occur in the same participant during follow‐up (e.g. revascularisation procedures, non‐fatal vascular events, adverse effects, etc.), we will count the number of participants with at least one event to avoid a unit‐of‐analysis error;

• For outcomes assessed at multiple time points, we will select the longest follow‐up from each study.

Dealing with missing data

When we encountered missing data, we contacted the original study investigators to request them. When we did not receive the requested data, we only analysed the available data. Therefore, all analyses were based on the number of participants assessed for each outcome within each study.

Assessment of heterogeneity

We assessed the included studies for clinical heterogeneity in dose, duration of therapy, and control used. We evaluated trials heterogeneity using Chi² and I² statistics, which described the variability in effect estimates that were due to heterogeneity between studies, rather than chance. The I² is given as a percentage, with a measure of 0%, meaning little to no variability in effect estimates between the studies, and progressive amounts of variability illustrated with increased I² percentage values (Higgins 2023). If tests for heterogeneity found I² > 50%, we used a random‐effects model. Otherwise, we used a fixed‐effect model for meta‐analysis.

Assessment of reporting biases

We planned to assess reporting bias with funnel plots if we included more than ten studies in a meta‐analysis (Higgins 2023). As we did not include more than ten studies in any analysis, we did not do this.

Data synthesis

We pooled data using a fixed‐effect model meta‐analysis with subgroups, when data were available. We used a random‐effects model when tests for heterogeneity found I² > 50%. We also considered the direction and magnitude of effects and degree of overlap between CIs. For outcomes for which we were unable to pool data, we described the results narratively.

Six of the studies measured outcomes on more than one occasion (Carrero 2005; Carrero 2006; Hammer 2019; FLAX‐PAD 2011; Leng 1998; Mori 1992). Therefore, we had to make a decision on which values to consider. Hammer and colleagues measured outcomes at three and six months (Hammer 2019). As the six‐month measurement was three months after cessation of therapy, we considered the three‐month values in the analysis. Mori and colleagues took measurements at 2, 4, and 10 weeks (Mori 1992). As the 10‐week measurements were taken six weeks after cessation of therapy (by which time the effects of omega‐3 fatty acid supplementation might have subsided), we used the four‐week values in the analysis. Leng and colleagues measured outcomes at 6, 12, 18, and 24 months (Leng 1998). We used the 24‐month data because we felt that the maximal effect of the intervention therapy would be observed at that time. Carrero and colleagues took measurements at 3, 6, 9, and 12 months (Carrero 2005). We used the 12‐month values, so the maximal effects of therapy would be analysed. Carrero and colleagues measured outcomes at 6 and 12 months (Carrero 2006). Again, we used the 12‐month values, for the same reasons as mentioned above. Edel and colleagues measured the outcomes at 1, 6, and 12 months (FLAX‐PAD 2011). Therefore, we used the 12‐month data.

Subgroup analysis and investigation of heterogeneity

There were insufficient data to perform subgroup analyses (i.e. short‐ and long‐term measurements of outcomes, severity of disease, presence of diabetes, or different age groups). We considered a separate meta‐analysis for the short‐ and long‐term outcomes (i.e. one year and less versus more than one‐year measurement). However, we excluded the two studies that measured long‐term outcomes from the meta‐analysis, as Leng and colleagues administered additional active substances to their intervention groups, and the ORIGIN study provided only baseline data for outcomes of interest (Leng 1998; ORIGIN 2019).

If we include at least 10 studies in each subgroup in future updates, we will undertake subgroup analysis on the severity of disease; comorbidities, such as diabetes; different age groups; and different sexes to assess their effects on outcomes, including maximal walking distance, pain‐free walking distance, and ankle‐brachial index.

We chose these subgroups for these reasons.

People with symptomatic PAD may experience different severities of the disease, based on the Fontaine classification, including stage II (mild, moderate, and severe claudication), III (rest pain) and IV (necrosis or gangrene of the limb, or both (Hardman 2014)).

Comorbidities, such as diabetes mellitus, are known to raise the incidence of PAD, accelerate the progress of the disease, and increase the severity of the disease (Barnes 2020).

Prevalence of PAD increases with age (40 to 50 years: 2%, 51 to 60 years: 3%, 61 to 70 years: 4%, 71 to 80 years: 7%, 81 to 90 years: 13.5%, 91 to 100 years: 24% (Savji 2013)).

PAD is a condition that affects women as often, or more commonly, than men. Sex‐related differences in pathophysiology and risk factors may be accompanied by a late start, and often atypical symptoms in women with PAD, in addition to the overall disease burden (Pabon 2022; Savji 2013).

Sensitivity analysis

If we include sufficient studies in future updates, we will undertake sensitivity analyses to assess the possible effects on maximal walking distance, pain‐free walking distance, and ankle‐brachial index, by excluding: studies with very small doses of omega‐3 fatty acid, unpublished studies, and cross‐over studies.

We included one study with a cross‐over design, but did not undertake a sensitivity analysis, as we did not include it in any meta‐analysis (Mackay 2012).

Summary of findings and assessment of the certainty of the evidence

Our main findings for the most clinically relevant outcomes of this review are presented in the summary of findings table that compares the results between the omega‐3 fatty acids and placebo/active control (Table 1). We created it using GRADEproGDT (GRADEpro GDT). We used GRADE methods to assess the certainty of the evidence as high, moderate, low, or very low, using the criteria of within‐study risk of bias, inconsistency, directness of evidence, imprecision, and publication bias (Schünemann 2020). We judged the following outcomes to be the most clinically relevant:

• Quality of life

• Pain‐free walking distance

• Maximal walking distance

• Ankle‐brachial index

• Revascularisation procedures in the lower limb

• Amputation rate/frequency

• Adverse effects

Results

Description of studies

See Figure 1.

Results of the search

Due to changes in Cochrane standards and recommendations since the last version of this review was published (Campbell 2013), and to ensure all records were assessed consistently by our new author team, new searches were designed and run from database inception, rather than updated from the date of the previous search.

The database searches identified 6960 records (See Figure 1). After deduplication, we screened the titles and abstracts of 6293 records. Of these, we assessed 6265 as not relevant. We assessed the full text of 28 potentially relevant records. We identified six new studies (16 records (FLAX‐PAD 2011; Hammer 2019; Mackay 2012; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019)). With the previously included studies, the total is now 15 included studies (Carrero 2005; Carrero 2006; Conway 2005; Gans 1990; Leng 1998; Mori 1992; Schiano 2008; Stricker 2008; Woodcock 1984).

We reassessed and included one previously excluded study, because it contained data about adverse events (Mackay 2012). We included one study that was previously listed as ongoing, but was now published (OMEGA‐PAD I 2015).

We excluded four new studies (four reports (Caligiuri 2014; Euctr2013‐004342‐42; Madden 2007; Ruiz‐Canela 2014)), in addition to five previously excluded studies (Berrettini 1996; Ishikawa 2010; Moller 1998; Olsson 1984; Ramirez‐Tortosa 1999).

We identified five new ongoing studies (six records (Canola‐PAD 2010; NCT02152930; NCT01256320; NCT02096757; UMIN000026607)).

Included studies

Summary details of the included studies are given in the Characteristics of included studies table.

We included a total of 15 studies in this update, which involved 1830 participants and were conducted in eight different countries: the United Kingdom (Conway 2005; Leng 1998; Mackay 2012; Woodcock 1984), Australia (Hammer 2019; Mori 1992), Italy (Schiano 2008), the Netherlands (Gans 1990), Switzerland (Stricker 2008), Canada (FLAX‐PAD 2011; ORIGIN 2019), the United States (OMEGA‐PAD I 2015; OMEGA‐PAD II 2019), and Spain (Carrero 2005; Carrero 2006). The number of participants in the included studies ranged from 18 to 971, although most recruited fewer than 100 (see Table 2 for details).

1. An overview of the characteristics of participants and interventions in the included studies.

Studies	Participants number	Participants sex	IC confirmation methods*	Intervention type	Omega‐3 doses per day	Comparison group	Duration of intervention	Measurement times
Carrero 2005	60	M	A, B	Skimmed milk containing EPA, and DHA plus oleic acid, folic acid and vitamins A, B6	200 mg of EPA, and 130 mg of DHA	Semi‐skimmed milk plus vitamins A and D	12 months	Baseline, 3, 6, 9, and 12 months
Carrero 2006	40	M	A, B	A fortified dairy product containing fish oil, oleic acid, folic acid, and other vitamins	2.1% of EPA and 1.4% of DHA	Semi‐skimmed milk	12 months	Baseline, 6, and 12 months
Conway 2005	50	M and F	A	Omega‐3 capsules	1.7 g of EPA and 1.15 g of DHA	Mixed oils capsules	16 weeks	Baseline and 16 weeks
FLAX‐PAD 2011	110	M and F	A, B	The products contain 30 g of milled flaxseed	Unclear	The products contain 30 g of milled wheat	12 months	Baseline, 1, 6, and 12 months
Gans 1990	32	M and F	A, E	Omega‐3 capsules	1.8 g of EPA and 1.2 g of DHA	Corn oil capsules	4 months	Baseline and 4 months
Hammer 2019	70	M and F	A, B, E	Omega‐3 capsules	850 mg to 882 mg EPA and DHA	Capsules containing fatty acids	3 months	Baseline, 3, and 6 months
Leng 1998	120	M and F	A, B	Omega‐3 capsules	280 mg of GLA and 45 mg of EPA	Sunflower oil capsules	24 months	Baseline, 6, 12, and 24 months
Mackay 2012	150	M and F	B	Omega‐3 capsules	850 mg to 882 mg EPA and DHA	Mixed oil capsules	6 weeks	6 weeks
Mori 1992	32	M	A, D	Omega‐3 capsules	2.8 g of EPA and 1.8 g of DHA	Olive oil capsules	4 weeks	Baseline, 2, 4, and 10 weeks
OMEGA‐PAD I 2015	80	M and F	A, B	Omega‐3 capsules	2.6 g of EPA and 1.8 g of DHA	Soybean capsules	1 month	Baseline and 1 month
OMEGA‐PAD II 2019	24	M	A, B, C	Omega‐3 capsules	2.6 g of EPA and 1.8 g of DHA	Soybean capsules	3 months	Baseline and 3 months
ORIGIN 2019	971	M and F	B, D	Omega‐3 capsules	465 mg of EPA and 375 mg of DHA	Olive oil capsules	6.2 years	Baseline
Schiano 2008	32	M and F	A, B, C, or D	Omega‐3 capsules	2 g capsules/ day (85% EPA and DHA)	Pre‐enrolment therapy	3 months	Baseline and 3 months
Stricker 2008	40	M and F	A, B, C, or D	35 mL/ day canola oil	2.24 g of ALA	Sunflower oil	8 weeks	Baseline and 8 weeks
Woodcock 1984	19	M and F	A, B	Omega‐3 capsules	1.8 g of EPA	Corn and olive oil capsules	7 weeks	Baseline and 7 weeks

^*Intermittent claudication (IC) confirmation methods: A: Participants to have symptoms of intermittent claudication; B: Confirmation of disease by Doppler pressure indices; C: Presence of stenosis or occlusion of a leg artery on duplex scanning or angiogram D: angiographic confirmation of peripheral arterial disease; E: Using standardised treadmill testing

ALA: alpha‐linolenic acid; DHA: docosapentaenoic acid; EPA: eicosapentaenoic acid; F: female; g: gram; GLA: gamma‐linoleic acid; IC: intermittent claudication; M: male; mL: millilitres

Several study authors provided additional data. Leng provided mean and standard deviation (SD) values for ankle‐brachial index (ABI) and walking distance (Leng 1998). Mori provided mean and SD values for cholesterol and triglyceride levels, along with details of the randomisation process and participant characteristics (Mori 1992). Carrero provided walking distance and information on the study's method of double‐blinding (Carrero 2005). Conway provided details on randomisation/concealment, exclusion criteria, inclusion criteria, and numerical values for each outcome (Conway 2005). Dagenais provided details on randomisation/concealment for the ORIGIN study (ORIGIN 2019).

Types of participants

Most of the studies required that participants have symptoms of intermittent claudication (IC). Ten randomised controlled trials (RCTs) confirmed disease by Doppler pressure indices (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Hammer 2019; Gans 1990; Leng 1998; Mackay 2012; OMEGA‐PAD I 2015; ORIGIN 2019; Woodcock 1984). Three studies required confirmation of disease by Doppler pressure indices and the presence of stenosis or occlusion of a leg artery, assessed by duplex scanning or angiogram (OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008). Mori, and Dagenais and colleagues also provided the angiographic evidence of PAD (Mori 1992; ORIGIN 2019). Gans and Hammer used a standardised treadmill test to ensure that the participants had stable disease (Gans 1990; Hammer 2019).

In total, 11 studies recruited both males and females, whereas four studies included only male participants. The mean age range in all RCTs was between 62 and 69 years (see Table 2 for details). Participants in all the included studies had all the symptoms of IC.

Studies differed in exclusion criteria; these are described in the Characteristics of included studies tables.

Interventions and placebos

Eleven studies used n‐3 fatty acid (FA) supplementation in the form of fish oil capsules to increase the longer chain omega‐3 fats (LCn‐3 (Conway 2005; Gans 1990; Hammer 2019; Leng 1998; Mackay 2012; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019; Schiano 2008; Woodcock 1984)).

Two studies used supplemental foods, such as fortified milk or dairy, to boost the LCn‐3 (Carrero 2005; Carrero 2006); two studies used foods with milled flaxseeds (FLAX‐PAD 2011), or canola oil (Stricker 2008), to increase alpha‐linolenic acid (ALA).

The LCn‐3 doses ranged from 0.35 g/d of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to 4.4 g/d. Four studies used a dose of LCn‐3 < 1 g/d (Carrero 2005; Hammer 2019; Mackay 2012; ORIGIN 2019); four studies used a dose of LCn‐3 < 1 to 3 g/d (Conway 2005; Hammer 2019; Schiano 2008; Woodcock 1984); four used a dose of LCn‐3 ≥ 3 g/d (Gans 1990; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; and two did not clearly state the dose (Carrero 2006; FLAX‐PAD 2011). Leng and colleagues' study differed from the other studies as the intervention group received capsules containing a combination of omega‐3 and omega‐6 fatty acids (1.12 g gamma‐linolenic acid (GLA)/ evening primrose oil daily, increasing to 1.68 g after two weeks, and 180 mg EPA, increasing to 270 mg (Leng 1998)). Doses of ALA were 2.24 g/d in Stricker and colleagues (Stricker 2008) and unclear in Leyva and colleagues (FLAX‐PAD 2011). See Table 2.

In most studies, the control group received the placebo as a capsule containing soybean, olive oil, corn oil, sunflower oil, and other types of fats (mixed oil or palm). Other studies used semi‐skimmed milk (Carrero 2005; Carrero 2006), two tablespoons a day of sunflower oil (Stricker 2008), or milled wheat as placebo (FLAX‐PAD 2011). Only Schiano and colleagues did not use a placebo preparation; the control group remained on pre‐enrolment therapy instead (Schiano 2008) It was not clear exactly what this was. See Table 2.

Three studies had active ingredients in one or both arms (Carrero 2005; Carrero 2006; Leng 1998).

Carrero 2005 used an intervention preparation containing omega‐3 fatty acids, oleic acid, and vitamins that were not included in the placebo preparation. However, the omega‐3 fatty acid dose was very small, and it would be impossible to attribute any observed effects to omega‐3 fatty acids alone.

Carrero 2006 had two treatment groups. Both received a dairy product containing EPA, DHA, oleic acid, folic acid, and vitamins A, D, E and B6,while the second treatment group also received 20 mg/day of simvastatin. There were also two control groups, which both received 500 mL/day of semi‐skimmed milk as a placebo preparation, with the addition of 20 mg/day of simvastatin for the second control group. We used the data of two groups for comparison – the group that received a dairy product containing EPA, DHA, oleic acid, folic acid, and vitamins A, D, E and B6 as the treatment, compared to the group that received 500 mL/day of semi‐skimmed milk as the placebo.

Leng 1998 evaluated the effect of both omega‐3 and omega‐6 fatty acid supplementation. This made it difficult to attribute any particular effects to omega‐3 fatty acids alone, since omega‐6 fatty acids have also been described as having total cholesterol‐lowering effects, and possible effects on modifying triglycerides, HDL, and LDL blood levels. They used linoleic acid as a placebo, even though this is a precursor compound of GLA. The authors acknowledged that the linoleic acid given to the control group may have been at least partly metabolised into GLA.

Length of studies

The duration of treatment varied between studies. The longest studies lasted 6.2 years (ORIGIN 2019), two years (Leng 1998), and one year (Carrero 2005; Carrero 2006; FLAX‐PAD 2011). The shortest duration of treatment was four weeks (OMEGA‐PAD I 2015; Mori 1992). Two studies continued to measure outcomes after the interventions finished (Hammer 2019; Mori 1992). In one, the participants were followed up to six months after a three‐month intervention (Hammer 2019). The second followed up for an additional six weeks after a four‐week intervention (Mori 1992).

See Table 2.

Outcomes

Studies assessed the effect of LCn3 and alpha‐linolenic acid (ALA) on different outcomes of interest.

• Quality of life (Conway 2005)

• Pain‐free walking distance (Carrero 2005; Carrero 2006; Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Woodcock 1984)

• Maximal walking distance (Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019)

• Ankle‐brachial index (Carrero 2005; Carrero 2006; Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Woodcock 1984)

• Revascularisation procedures in the lower limb (Leng 1998)

• Amputation rate/frequency (Leng 1998)

• Cholesterol (total, HDL, LDL, VLDL (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008; Woodcock 1984))

• Triglycerides (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008; Woodcock 1984)

• Blood pressure (systolic and diastolic (FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019))

• All‐cause and vascular mortality (FLAX‐PAD 2011; Leng 1998)

• Non‐fatal vascular events (FLAX‐PAD 2011; Hammer 2019; Leng 1998; OMEGA‐PAD II 2019)

• Adverse effects of therapy (Conway 2005; Gans 1990; Leng 1998; Mackay 2012; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008)

Ongoing Studies

We found five ongoing trials, which are described in the Characteristics of ongoing studies tables. At the time of writing this review, all of these trials were marked as complete but no data were published: completed in 2013 (Canola‐PAD 2010; NCT01256320); completed in 2016 (NCT02152930); completed in 2017 (NCT02096757); and completed in 2019 (UMIN000026607).

We tried to contact authors of all 'overdue' ongoing trials, and only one study team stated that publications were forthcoming (Canola‐PAD 2010); the others did not respond.

Excluded studies

For this update, we excluded four new studies (Caligiuri 2014; Euctr2013‐004342‐42; Madden 2007; Ruiz‐Canela 2014). Five studies were excluded in the previous version (Berrettini 1996; Ishikawa 2010; Moller 1998; Olsson 1984; Ramirez‐Tortosa 1999). See Characteristics of excluded studies for details.

Reasons for excluding studies are as follows:

• Not an RCT (Madden 2007; Moller 1998; Olsson 1984; Ramirez‐Tortosa 1999);

• IC participant data unavailable/unclear (Berrettini 1996; Caligiuri 2014; Ishikawa 2010; Ruiz‐Canela 2014);

• Terminated due to low recruitment rate (Euctr2013‐004342‐42).

Risk of bias in included studies

Details of the risk of bias are presented in Figure 2 and Figure 3, and in the Characteristics of included studies table.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study

We defined the overall summary of risk of bias judgement for each study as:

• Low risk of bias – low risk of bias for all key domains (selection bias, performance bias, and detection bias);

• Unclear risk of bias – unclear risk of bias for one or more key domains (selection bias, performance bias, and detection bias);

• High risk of bias – high risk of bias for one or more key domains (selection bias, performance bias, and detection bias)

Based on the above definition, two studies were an overall low risk of bias (FLAX‐PAD 2011; Hammer 2019); the other studies were an overall unclear risk of bias for the key domains.

Allocation

We judged seven studies at low risk of selection bias, as we judged them at low risk for both random sequence generation and allocation concealment (Conway 2005; FLAX‐PAD 2011; Hammer 2019; Leng 1998; Mackay 2012; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019).

We considered 13 studies at a low risk of bias for random sequence generation. Five studies used block randomisation (Hammer 2019; Leng 1998; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Stricker 2008); four studies used a table of random numbers (Carrero 2005; Carrero 2006; Mori 1992; Woodcock 1984); three studies used a computerised randomisation program (Conway 2005; FLAX‐PAD 2011; Mackay 2012); and one study used an automated telephone randomisation system. Three studies provided no description of the randomisation process, and we deemed them to have an unclear risk of bias (Gans 1990; ORIGIN 2019; Schiano 2008).

We judged nine studies at a low risk of bias for allocation concealment (Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Mackay 2012; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019). The remaining six studies did not describe how group allocation was concealed, and we judged them to have an unclear risk of bias (Carrero 2005; Carrero 2006; Mori 1992; Schiano 2008; Stricker 2008; Woodcock 1984).

Blinding

We judged 12 studies at a low risk of bias for performance bias, blinding of participants and personnel (Carrero 2005; FLAX‐PAD 2011; Hammer 2019; Leng 1998; Mackay 2012; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019; Schiano 2008; Stricker 2008; Woodcock 1984). Three studies reported a blind trial method, but did not provide details of the blinding process; we deemed them to have an unclear risk of performance bias (Carrero 2006; Conway 2005; Gans 1990).

We considered five studies at a low risk of bias for blinding of outcome assessment (Carrero 2005; FLAX‐PAD 2011; Hammer 2019; Schiano 2008; Stricker 2008). The ten other trials did not report whether outcome assessors were blinded, and we judged them to have an unclear risk of detection bias (Carrero 2006; Conway 2005; Gans 1990; Leng 1998; Mackay 2012; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019; Woodcock 1984).

Incomplete outcome data

Attrition bias occurs when participants who drop out of a study systematically differ from those who remain (Mohammady 2017).

We judged nine studies at low risk of attrition bias due to incomplete outcome data (Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019; Schiano 2008). We considered three studies to have an unclear risk of attrition bias, as they lacked an explicit statement to confirm there were no exclusions or withdrawals (Carrero 2006; Stricker 2008; Woodcock 1984). The number of exclusions post‐randomisation was not stated in one study (Carrero 2006). Forty participants were randomised and 40 were analysed in another, so it was implied that there were no exclusions or withdrawals. However, there was no explicit statement to confirm this (Stricker 2008). The third did not state the number of people randomised, or describe any withdrawals or exclusions (Woodcock 1984).

We judged two studies at a high risk of attrition bias due to incomplete outcome data, as exclusions were not balanced between study groups (Carrero 2005; Mackay 2012). A third study had missing outcome data that introduced an imbalance between the two groups (6 of the intervention group and 11 of the control group chose not to complete the study). In addition, 15 of the intervention group and 13 of the control group were withdrawn by the trial organisers due to either fatal or non‐fatal adverse events, so we judged this as a high risk of bias (Leng 1998)

Attrition bias can affect the quality of the study data.

Selective reporting

We judged 11 studies at a low risk of bias from selective reporting, as they all reported the prespecified outcomes as planned. We judged the Woodcock study to have a high risk of bias as data were missing for most outcomes (Woodcock 1984). We considered three studies to have an unclear risk of reporting bias as there were discrepancies in the outcomes between trial registry information and publications (FLAX‐PAD 2011; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019).

Other potential sources of bias

Three studies used additional active ingredients in the study or control arms (Carrero 2005; Carrero 2006; Leng 1998). Carrero 2005 and Carrero 2006 differed from the other studies because the intervention preparation, as well as containing omega‐3 fatty acids, contained oleic acid and vitamins not included in the placebo preparation. However, the omega‐3 fatty acid dose was very small, so it would be impossible to attribute any observed effects to omega‐3 fatty acids alone. Carrero 2006 reported a large difference between claudication distance across groups between baseline and the start of the study. The trial sample size was very small, with poor matching for age. Therefore, we judged Carrero 2005 and Carrero 2006 to have an unclear risk of other biases. Leng 1998 evaluated the effect of both omega‐3 and omega‐6 fatty acid supplementation. This made it difficult to attribute any particular effects to omega‐3 fatty acids alone, since omega‐6 fatty acids have also been described as having total cholesterol‐lowering effects, and positive effects on modifying triglyceride, HDL, and LDL blood levels (Hooper 2018). They used linoleic acid as a placebo, even though this is a precursor compound of GLA. The authors acknowledged that the linoleic acid given to the control group may have partially metabolised into GLA. Therefore, we judged this study to have a high risk of other bias.

Mori 1992 provided no data on the severity of peripheral arterial disease. We judged this study to have an unclear risk of other biases.

In OMEGA‐PAD II 2019, recruitment for the trial was slower than expected, and was ended by the principal investigator before reaching target enrolment, so we judged this to be of high risk of other bias.

We identified no other potential sources of bias for the remaining studies.

Effects of interventions

See: Table 1

Of the 15 RCTs that met our inclusion criteria, we included seven in meta‐analyses (Conway 2005; FLAX‐PAD 2011; Gans 1990; Mackay 2012; Mori 1992; Stricker 2008; Woodcock 1984). The remaining studies clearly collected relevant data, but did not report it in a format that we could use in meta‐analyses (Carrero 2005; Carrero 2006; Hammer 2019; Leng 1998; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; ORIGIN 2019; Schiano 2008). Mackay 2012 did not report any of our primary outcomes, but did report adverse effects of treatment. We narratively reported the results of the individual studies when we were unable to use them in meta‐analysis.

The studies were clinically heterogeneous for dose, duration of therapy, and placebo used. When statistical heterogeneity was detected with the Chi² test, we used and reported the random‐effects model, and presented the mean difference (MD).

We excluded three studies from the meta‐analysis as they administered additional active substances to their intervention groups, and we could not attribute any observed effects to the n‐3 FAs alone (Carrero 2005; Carrero 2006; Leng 1998). Two studies could not be included in the analysis process as they did not express mean values for any outcomes (preferring median values (25th; 75th percentile)), so we could not combine the results of these two studies with the others (Hammer 2019; Schiano 2008). We excluded two studies from the analysis since they expressed the mean change of the pre‐and post‐intervention values, and did not report the post‐intervention data for most outcomes (OMEGA‐PAD I 2015; OMEGA‐PAD II 2019). ORIGIN 2019 provided only baseline data for our outcomes of interest.

Three studies showed substantial clinical heterogeneity, and we excluded them from the meta‐analysis (Carrero 2005; Carrero 2006; Leng 1998). Leng 1998 took place over a longer time period and involved a greater number of participants than the other studies. In addition, it evaluated the effect of both omega‐3 and omega‐6 fatty acid supplementation. This made it difficult to attribute any particular effects to omega‐3 fatty acids alone. They used linoleic acid as a placebo, even though this is a precursor compound of GLA, and acknowledged that it may have partially metabolised into GLA. This study also had the highest number of withdrawals and dropouts. The intervention in Carrero 2005 and Carrero 2006 contained a very small dose of omega‐3 fatty acid, oleic acid, and vitamins that were not included in the placebo preparation. Participants in both groups were also given antiplatelet therapy and a haemorheological agent.

We considered a secondary analysis with only these three studies, but we did not undertake one, as it would be still impossible to attribute any observed effects to omega‐3 fatty acids.

Primary outcomes

We were unable to include all studies in analyses because of different outcomes and unusable data. See Table 1.

Quality of life

One study (N = 50) measured quality of life (Conway 2005).

The study authors reported the difference in quality of life using the self‐reported Short Form ‐ 36 item (SF‐36) questionnaire for individual participants between entry and completion at 16 weeks, through t‐test comparison. The evidence is very uncertain about the effect of the intervention on any of the eight quality of life parameters in the intervention group (P > 0.05). Numerical values were not given in the paper, neither were results for the control group. We assessed this outcome as having very low‐certainty evidence.

Pain‐free walking distance (PFWD)

Eight studies (N = 511) reported PFWD (Carrero 2005; Carrero 2006; Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Woodcock 1984).

We were able to combine three of them in our analysis (Conway 2005; FLAX‐PAD 2011; Gans 1990). We detected substantial heterogeneity in this outcome, so we pooled results using a random‐effects model. There may be little to no difference between groups for PFWD, but the evidence is very uncertain (MD 1.01 meters (m), 95% confidence interval (CI) ‐34.23 to 36.24; 3 studies, 147 participants; I² = 70%; P = 0.96; Analysis 1.1; very low‐certainty evidence).

1.1. Analysis.

Comparison 1: Omega‐3 fatty acids (FA) versus placebo, Outcome 1: Walking distance

We narratively reported the results of the individual studies since we were unable to combine them in the analysis.

Only Carrero 2005 and Carrero 2006 reported an effect on the intervention group. In Carrero 2005 (N = 60), PFWD increased by up to 3.5 times in the intervention group (by 279.33 m) compared to only 43.5 m in the control group (P < 0.05). In Carrero 2006 (N = 40), PFWD may increase by more than three times in the two treatment groups: enriched dairy product and enriched dairy product + simvastatin. In the enriched dairy product group, PFWD increased by 251 m compared to an increase of only 28 m in the control group (P < 0.05). In the enriched dairy product + simvastatin group, PFWD increased by 317 m compared to 54 m in the control group, which also received simvastatin (P < 0.05). Both studies used active ingredients in the intervention group, and the omega‐3 fatty acid dose was very small. Therefore, we are unable to attribute any observed effects to omega‐3 fatty acids alone.

Hammer 2019 (N = 70) used median and interquartile range (IQR) to measure PFWD, three months after treatment completion in two groups. The results demonstrated a median change of 4 m (IQR = ‐9, 30%) in the intervention group (P = 0.15), and 8 m (IQR= ‐28, 32%) in the placebo group (P = 0.6). Compared with the baseline, there is no clear difference between groups (P = 0.62).

Leng 1998 (N = 120) reported that the mean PFWD increased after 24 months by 20 m in the intervention group and by 15 m in the control group. Although the P value was not presented, the study authors stated there may be no significant difference between the two groups that might be due to omega‐6.

Woodcock 1984 (N = 19) found that after seven weeks of treatment, three participants in the intervention group noted an increase of 46 m in claudication distance, and one participant no longer had rest pain. In the control group, one had an increase in claudication distance of 73 m and one participant no longer had rest pain. Unfortunately, no data were provided to clarify or analyse these findings.

Maximum walking distance (MWD)

Four studies assessed MWD (Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019).

We were able to combine three studies in our analysis (Conway 2005; FLAX‐PAD 2011; Gans 1990). We detected substantial heterogeneity, so we pooled the results using a random‐effects model. Results showed that there may be little to no difference in post‐trial MWD between groups, but the evidence is very uncertain (MD ‐4.18 m, 95% CI ‐37.10 to 28.74; 3 studies, 164 participants; I² = 52%; P = 0.80; Analysis 1.1; very low‐certainty evidence).

Hammer 2019 (N = 70) reported that MWD results remained unchanged in both groups between pre‐ and post‐intervention. The median change of MWD was 0 m (IQR ‐21 to 49, P = 0.60) in the control group, and 0.5 m (IQR ‐29 to 38, P = 0.63) in the intervention group after the omega‐3 administration. No difference was detected between the two groups (P = 0.73).

Secondary outcomes

Ankle‐brachial index (ABI)

Eleven studies reported the ankle‐brachial index (Carrero 2005; Carrero 2006; Conway 2005; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Woodcock 1984).

We were able to combine three studies in our analysis (Conway 2005; FLAX‐PAD 2011; Gans 1990). Heterogeneity was low for this outcome, so we pooled the results using the fixed‐effect method. The evidence suggests that omega‐3 has little to no effect on post‐trial ABI values, but the evidence is very uncertain (MD ‐0.02, 95% CI ‐0.08 to 0.04; 3 studies, 168 participants; I² = 0%; P = 0.58; Analysis 1.2; very low‐certainty evidence).

1.2. Analysis.

Comparison 1: Omega‐3 fatty acids (FA) versus placebo, Outcome 2: Ankle brachial index

Here are the narrative results for the other eight studies.

Carrero 2005 (N = 60) reported a change between pre‐ and post‐intervention ABI. ABI increased from 0.46 to 0.52 in the intervention group. This was reported as being statistically significant, although the P value was not presented. There was no corresponding change in the control group.

Carrero 2006 (N = 40) reported changes between pre‐ and post‐intervention values of ABI in both groups. No P value was presented. Both studies used active ingredients in the intervention group. The omega‐3 fatty acid dose was very small, and we cannot attribute any observed effects to the omega‐3 fatty acids alone.

Hammer 2019 (N = 70) reported absolute changes in the ABI, using median and IQR, three months after treatment completion in both study groups. The results did not show a clear difference between groups after treatment (median change in the right leg was 0.02 (IQR ‐0.04 to 0.08) in the omega‐3 group (P = 0.33), and ‐0.02 (IQR –0.05 to 0.02) in the placebo group (P = 0.54)). The results did not show a clear difference between the two groups (P = 0.32).

Leng 1998 (N = 120) reported that the mean ABI had risen by 6% in the intervention and 9% in the control group after 24 months. No P values were presented, but the study authors stated there was no significant difference between the two groups that might be due to omega‐6.

The OMEGA‐PAD studies reported ABI as an MD between baseline and post‐intervention in the two groups. Both studies reported no difference between omega‐3 and placebo groups (MD 0.01, 95% CI ‐0.04 to 0.06; P = 0.67; 1 study, 80 participants (OMEGA‐PAD I 2015)); (MD 0.01, 95% CI ‐0.11, to 0.13; P = 0.87; 1 study, 24 participants (OMEGA‐PAD II 2019)).

Schiano 2008 (N = 32) reported no change in ABI between pre‐ and post‐intervention. In the control group, the median ABI was 0.67 (IQR 0.55 to 0.77) at baseline, and 0.68 (IQR 0.55 to 0.78) at the end of the study. In the treatment group, ABI remained unchanged (median 0.66, IQR 0.53 to 0.69) after omega‐3 fatty acid administration. No P values were presented.

Woodcock 1984 (N = 19) reported that the ABI did not change in either group after seven weeks. However, they did not present any data.

Revascularisation procedures in the lower limb

Leng 1998 (N = 120) reported omega‐3 results may have little to no effect on the incidence of lower limb angioplasty/bypass surgery, but the evidence is very uncertain (1 (1.7%) in the intervention group versus 3 (5%) in the control group (P > 0.05)). There was no significant difference between the two groups that might be due to omega‐6. We assessed this to be very low‐certainty evidence.

Amputation rate/frequency

Leng 1998 (N = 120) reported omega‐3 results may have little to no effect on the progression of critical limb ischaemia/amputation, but the evidence is very uncertain (1 (1.7%) in the intervention group versus 0 in the control group (no P value reported)). There was no significant difference between the two groups that might be due to omega‐6. We assessed this to be very low‐certainty evidence.

Lipid levels

Total cholesterol

Twelve studies reported total cholesterol (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008; Woodcock 1984).

There may be little to no difference in mean post‐trial total cholesterol levels between the intervention and control groups (MD 0.10 mmol/L, 95% CI ‐0.48 to 0.67; 5 studies, 206 participants; I² = 61%; P = 0.74; Analysis 1.3; low‐certainty evidence). Statistical heterogeneity (I² = 61%) was detected; therefore, we used the random‐effects model (FLAX‐PAD 2011; Gans 1990; Mori 1992; Stricker 2008; Woodcock 1984).

1.3. Analysis.

Comparison 1: Omega‐3 fatty acids (FA) versus placebo, Outcome 3: Cholesterol

Here are the narrative results for the other seven studies.

Carrero 2005 (N = 60) reported a decrease in total cholesterol in the intervention group at 12 months (from 5.43 mmol/L to 5.15 mmol/L, P < 0.05). No change was reported in the control group. No P value was reported, but study authors reported no difference in total cholesterol between the two groups.

Carrero 2006 (N = 40) reported an increase in total cholesterol in the control group that did not receive the enriched dairy product or simvastatin (from 5.07 mmol/L to 5.59 mmol/L at 12 months, P < 0.05), and no change in the intervention group that received only semi‐skimmed milk. No P values were presented, but the study authors reported no differences in total cholesterol between the two groups, which may be caused by the use of active ingredients in the intervention group, since the omega‐3 fatty acid dose was very small, and we cannot attribute any observed effects to the omega‐3 fatty acids.

Hammer 2019 (N = 70) reported that omega‐3 was more effective than control in lowering total cholesterol (P = 0.03) after three months (median change ‐0.33 mmol/L (IQR ‐0.65 to 0.18) in the intervention group (P = 0.7), and 0.10 mmol/L (IQR ‐0.21 to 0.36) in the placebo group (P = 0.20)).

Leng 1998 (N = 120) reported a change in mean total cholesterol in the intervention group (from 6.03 to 1.129 mmol/L), and no change in the control group (5.862 to 5.922 mmol/L) after 24 months. Pre‐ to post‐ intervention differences between the two groups were not statistically significant. No P values were presented, but the study authors stated there was no difference in total cholesterol detected between the two groups that might be due to omega‐6.

OMEGA‐PAD I 2015 (N = 80) reported no change in total cholesterol in either group after one month (mean change ‐0.05 mmol/L in the intervention group; 0.13 mmol/L in the placebo group). No difference was detected between the two groups (P = 0.51).

OMEGA‐PAD II 2019 (N = 24) reported no change in total cholesterol levels in either group after three months (mean change ‐0.22 in the intervention group; ‐0.14 mmol/L in the placebo group). No difference was observed between the two groups (P = 0.71).

Schiano 2008 (N = 32) reported no difference in the median of total cholesterol between pre‐ and post‐intervention in either group (P = 0.27). In the treatment group, the median total cholesterol was 5.43 mmol/L at baseline and 4.37 mmol/L after three months; in the control group, it was 4.7 mmol/L at baseline and 4.5 mmol/L after three months.

High‐density lipoprotein cholesterol (HDL)

Twelve studies evaluated HDL cholesterol (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008; Woodcock 1984).

We included five studies in the meta‐analysis (FLAX‐PAD 2011; Gans 1990; Mori 1992; Stricker 2008; Woodcock 1984). There may be little to no difference between groups in mean HDL cholesterol levels at the end of the trial (MD ‐0.07 mmol/L, 95% CI ‐0.16 to 0.02; 5 studies, 206 participants; I² = 0%; P = 0.13; Analysis 1.3; low‐certainty evidence). Fixed‐effect and random‐effects analyses did not differ.

Here are the narrative results for the other seven studies.

Carrero 2005 (N = 60) described a fall in HDL cholesterol in both groups (from 1.27 mmol/L to 1.17 mmol/L, P < 0.001 for intervention group; from 1.30 mmol/L to 1.22 mmol/L, P < 0.05 for control group). However, no difference was detected between the two groups after 12 months (P = 0.53), which may be due to the use of active ingredients in the intervention group. It is not possible to attribute observed effects only to omega‐3 fatty acids.

Carrero 2006 (N = 40) described no change in HDL cholesterol in either group after 12 months (from 1.45 mmol/L to 1.35 mmol/L for the intervention group; from 1.14 mmol/L to 1.09 mmol/L for the control group). No P values were presented, but the authors reported no difference in HDL between the two groups, which may be due to the use of active ingredients in the intervention group. It is not possible to attribute observed effects only to omega‐3 fatty acids.

Hammer 2019 (N = 70) described no change in HDL cholesterol in either group after three months (median change 0.04 mmol/L (IQR ‐0.13 to 0.14) in the intervention group (P = 0.53); 0.03 mmol/L (IQR ‐0.05 to 0.05) in the placebo group (P = 0.74)). There was no difference between the two groups (P = 0.55).

Leng 1998 (N = 120) reported that after 24 months of treatment, those in the fish oil group who had completed the study had an 18% increase in HDL levels, compared to the control group, in which there was a 17% increase. No P values were presented, but the authors stated there was no significant difference between the two groups that might be due to omega‐6.

OMEGA‐PAD I 2015 (N = 80) detected an increase of 0.05 mmol/L in the HDL cholesterol levels in the intervention group after one month (P = 0.03), but no variations in the controls (P = 0.78). However, no difference was detected between the groups at the end of the study (P = 0.41).

OMEGA‐PAD II 2019 (N = 24) reported that after three months, there was no change in HDL levels in either group. No difference was observed between the two groups (P = 0.25).

Schiano 2008 (N = 32) reported no change in the median HDL in the intervention group (1.16 to 1.24 mmol/L), or in the control group (1.37 to 1.00 mmol/L) after three months. No difference was detected between the groups at the end of the study (P = 0.74).

Low‐density lipoprotein cholesterol (LDL)

Eleven studies evaluated LDL cholesterol (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008).

We included four studies in the meta‐analysis (FLAX‐PAD 2011; Gans 1990, Mori 1992; Stricker 2008). There may be little to no difference between groups in mean LDL cholesterol levels (MD 0.28 mmol/L, 95% CI ‐0.28 to 0.85; 4 studies, 186 participants; I² = 73%; P = 0.33; Analysis 1.3; low‐certainty evidence). Significant statistical heterogeneity was detected, so we used and reported results using the random‐effects model.

Here are the narrative results for the other seven studies.

Carrero 2005 (N = 60) reported no change in LDL cholesterol in either group after 12 months (mean change ‐0.09 mmol/L in the intervention group; 0.20 mmol/L in the control group). No P values were presented, but the authors reported there was no difference between the two groups at 12 months, which may be due to the use of active ingredients in the intervention group. It was not possible to attribute any observed effects to omega‐3 fatty acids.

Carrero 2006 (N = 40) reported changes in LDL cholesterol levels. The control group that did not receive the enriched dairy product or simvastatin saw increased LDL cholesterol (from 2.71 to 2.98 mmol/L after 12 months, P < 0.05). LDL levels dropped in the other groups.

Hammer 2019 (N = 70) reported no change in LDL in either group after three months (median change ‐0.17 mmol/L (IQR ‐0.46 to 0.29) in the intervention group (P = 0.28); 0.12 mmol/L (IQR ‐0.20 to 0.27) in the placebo group (P = 0.22)). There was no difference between the two groups (P = 0.08).

Leng 1998 (N = 120) reported an increase in mean LDL in both the intervention group (2.76 to 3.1 mmol/L), and the control group (2.674 to 3 mmol/L) after 24 months. No P values were presented, but the study authors stated there was no significant difference between the two groups that might be due to omega‐6.

OMEGA‐PAD I 2015 (N = 80) detected no change in LDL in either group after one month (mean change 0.07 mmol/L (P = 0.37) in the intervention; 0.10 mmol/L in the control (P = 0.22)). No difference was detected between the two groups at the end of the study (P = 0.14).

OMEGA‐PAD II 2019 (N = 24) reported no change in LDL levels in either group after three months (mean change ‐0.21 in the intervention; ‐0.08 mmol/L in the control). No difference was observed between the two groups (P = 0.57).

Schiano 2008 (N = 32) reported no change in the median LDL cholesterol in the intervention group (3.41 to 2.80 mmol/L), or in the control group (3.00 to 2.89 mmol/L) after three months. No difference was detected between the groups at the end of the study (P = 0.25).

Very low‐density lipoprotein cholesterol (VLDL)

Only two studies evaluated VLDL cholesterol (Gans 1990; Leng 1998). There may be little to no difference between mean VLDL cholesterol levels in either group (MD 0.08 mmol/L, 95% CI ‐0.47 to 0.63; 1 study, 33 participants; P = 0.78; Analysis 1.3; low‐certainty evidence).

Leng 1998 (N = 120) reported that at six months, those who ultimately completed the study had higher VLDL cholesterol levels in the intervention group (1.2 mmol/L (SD 0.97), compared to the control group (1.06 mmol/L (SD 0.96); P < 0.05)).

Triglycerides

Eleven studies evaluated triglycerides (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008; Stricker 2008; Woodcock 1984).

We included four studies in the meta‐analysis (FLAX‐PAD 2011; Gans 1990; Mori 1992; Woodcock 1984). There may be little to no difference in triglyceride levels between groups (MD ‐0.14 mmol/L, 95% CI ‐0.59 to 0.30; 4 studies, 165 participants; I² = 66%; P = 0.52; Analysis 1.4; low‐certainty evidence). Statistical heterogeneity was detected. Therefore, we calculated and presented the random‐effects model results.

1.4. Analysis.

Comparison 1: Omega‐3 fatty acids (FA) versus placebo, Outcome 4: Triglycerides

Narrative results for the other seven studies

Carrero 2005 (N = 60) described no change in triglyceride levels in either group after 12 months (mean change ‐0.08 mmol/L for the intervention group; 0.06 mmol/L for the control group). No P values were presented, but study authors reported there was no difference between the two groups at 12 months, which may be due to the use of active ingredients in the intervention group. It was not possible to attribute any observed effects to omega‐3 fatty acids.

Carrero 2006 (N = 40) detected no change in triglyceride levels in either group after 12 months. No P values were presented.

Hammer 2019 (N = 70) reported that the median triglyceride levels declined in the treatment group (median change ‐0.19 mmol/L (IQR ‐0.72 to 0.13, P = 0.03)), but remained unchanged in the control group after three months (P = 0.81). No change was reported between groups (P = 0.08).

OMEGA‐PAD I 2015 (N = 80) reported a reduction of ‐0.38 mmol/L in the triglyceride levels in the intervention group (P < 0.0001), and no change in the control group (P = 0.2) after one month. A difference was observed in the mean change of triglyceride levels between groups (P = 0.02).

OMEGA‐PAD II 2019 (N = 24) reported no change in triglyceride levels in the intervention and control groups after three months (mean change ‐0.15 mmol/L for the intervention group; 0.06 mmol/L for the control group). No difference was observed in the mean change of triglyceride levels between groups (P = 0.26).

Schiano 2008 (N = 32) reported median values, and showed a decrease in triglyceride levels from 1.93 mmol/L to 1.21 mmol/L (P = 0.03) after three months in the treatment group. There was no change from baseline in the control group (P = 0.91). There was a difference detected between groups in relative changes from baseline after three months (P = 0.03).

Stricker 2008 (N = 40) expressed median values for this outcome, and reported no changes in median triglyceride levels in either the intervention group (1.32 to 1.29 mmol/L), or the control group (1.44 to 1.24 mmol/L) after eight weeks. No P values were presented.

Blood pressure

Six studies evaluated systolic (SBP) and diastolic (DBP) blood pressure (FLAX‐PAD 2011; Gans 1990; Hammer 2019; Leng 1998; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019).

We were able to combine the results from two studies (FLAX‐PAD 2011; Gans 1990).

The evidence suggests there may be little to no difference in SBP between groups (MD ‐4.36 mmHg, 95% CI ‐11.37 to 2.65; 2 studies, 118 participants; I² = 33%; P = 0.22; Analysis 1.5; low‐certainty evidence). Statistical heterogeneity was low; therefore, we used the fixed‐effects model.

1.5. Analysis.

Comparison 1: Omega‐3 fatty acids (FA) versus placebo, Outcome 5: Blood pressure

There may be little to no difference in DBP between groups (MD ‐3.54 mmHg, 95% CI ‐7.14 to 0.06; 2 studies, 118 participants; I² = 0%; P = 0.05; Analysis 1.5; low‐certainty evidence). No statistical heterogeneity was detected; therefore, we used the fixed‐effects model.

Narrative results for the other four studies

Hammer 2019 (N = 70) did not find any changes in median levels of SBP (P = 0.32) or DBP (P = 0.68) after three months in either group.

Leng 1998 (N = 120) detected a statistically significant reduction in SBP of 5% in the intervention group versus an increase of 0.25% in the control group (P < 0.05). There were no changes in DBP. No P values were presented, but the authors stated there were no significant differences between the two groups that might be due to omega‐6.

OMEGA‐PAD I 2015 (N = 80) detected an increase of 4 mmHg in the SBP in the intervention group after one month (P = 0.09), versus a reduction of 0.4 mmHg in the control group (P = 0.86). The authors reported a reduction of 0.7 mmHg in the DBP in the intervention group (P = 0.64), and 2 mmHg in the control group (P = 0.24) after one month. No difference was observed in the mean change of SBP (P = 0.16), or DBP (P = 0.63) between groups.

OMEGA‐PAD II 2019 (N = 24) found no change in mean differences between groups for either SBP (P = 0.79), or DBP (P = 0.86) after three months.

All‐cause and vascular mortality

Two studies reported mortality (FLAX‐PAD 2011; Leng 1998). All‐cause mortality and vascular mortality were reported by Leng 1998, and vascular mortality by FLAX‐PAD 2011. We were unable to combine data.

FLAX‐PAD 2011 (N = 110) found no difference between groups in the incidence of cardiovascular mortality; they reported one death in the intervention group (reported as a reason for loss of follow‐up). No P value was presented.

Leng 1998 (N = 120) found no difference between groups in the incidence of all‐cause mortality or cardiovascular mortality. They reported three all‐cause deaths (5%) in each group, and two cardiovascular‐related deaths (3%) in each group. No P values were presented for either outcome, but the authors reported there were no differences between groups.

Non‐fatal vascular events

The incidence of non‐fatal coronary events was reported in three studies (FLAX‐PAD 2011; Hammer 2019; Leng 1998). Leng 1998 was not included in the meta‐analysis. The meta‐analysis showed there may be little to no difference between intervention and control groups regarding this event (OR 0.59, 95% CI 0.13 to 2.60; 2 studies, 141 participants; I² = 0%; P = 0.48; low‐certainty evidence). See Analysis 1.6.

1.6. Analysis.

Comparison 1: Omega‐3 fatty acids (FA) versus placebo, Outcome 6: Vascular events

Leng 1998 (N = 120) reported no difference between treatment and control groups in the incidence of non‐fatal coronary events, with six cases (5%) in the intervention group, and nine cases in the control group. No P values were presented.

Two studies reported the incidence of non‐fatal stroke/transient ischaemic attack (FLAX‐PAD 2011; OMEGA‐PAD II 2019). The meta‐analysis showed there may be little to no difference between the intervention and control groups (OR 0.95, 95% CI 0.13 to 6.77; 2 studies, 110 participants; I² = 6%; P = 0.96; Analysis 1.6; low‐certainty evidence).

Adverse effects of therapy

Seven studies reported adverse events of therapy, from gastrointestinal upset to headache (Conway 2005; Gans 1990; Leng 1998; Mackay 2012; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008). The types of adverse events reported and details of reporting varied from study to study. The evidence suggests that omega‐3 results in little to no difference in adverse events.

Gans 1990, OMEGA‐PAD I 2015 and Schiano 2008 (N = 144) stated that neither group experienced significant side effects from fatty acid supplementation.

Conway 2005 (N = 60) reported that four people withdrew from the study because they were unable to tolerate therapy, due to either nausea or difficulty in swallowing the capsules (3 people from the intervention group and 1 in the placebo group).

Mackay 2012 (N = 150) reported a higher incidence of adverse effects, including belching, heartburn or reflux, nausea and vomiting, loose stool, epistaxis, and inability to digest the tablets, with 14 (19%) in the intervention group versus eight (10%) in the control group, but analysis indicated no clear difference (P = 0.13).

In Leng 1998 (N = 120), gastrointestinal upset was higher in the intervention group (30 (50%)) than in the control group (23 (38%); P < 0.05). However, it should be noted that the intervention group also received an omega‐6 fatty acid, so we cannot say for sure if any side effects were due to omega‐3 FA.

OMEGA‐PAD II 2019 (N = 80) reported no gastrointestinal upset during their study. One participant in the placebo group experienced a headache.

The remaining studies did not refer to side effects of therapy (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Hammer 2019; Mori 1992; ORIGIN 2019; Stricker 2008; Woodcock 1984).

Discussion

Summary of main results

This review included 15 studies (1830 participants), which examined the effects of omega‐3 fatty acid supplementation on intermittent claudication (IC). The data presented in this update agrees with the findings of the previous version of this review (Campbell 2013). Based on these studies, omega‐3 fatty acid supplementation may have little to no effect on quality of life, pain‐free walking distance (PFWD), maximal walking distance, ankle‐branchial index, the incidence of revascularisation procedures, and the rate or frequency of amputation in the lower limbs, but the evidence is very uncertain.

In addition, Omega‐3 may have little to no effect on lipid levels (total cholesterol, high‐density lipoprotein (HDL) cholesterol, low‐density lipoprotein (LDL) cholesterol, very low‐density lipoprotein (VLDL) cholesterol, and triglycerides), systolic or diastolic blood pressure, or non‐fatal vascular events. Evidence on mortality (all‐cause or vascular) is very uncertain. The evidence suggests that omega‐3 results in little to no difference in adverse events.

Overall completeness and applicability of evidence

This Cochrane review involved 1830 participants, both men and women, from eight countries. Certain populations (e.g. those with a diagnosis of diabetes) were excluded from the majority of studies; thus, findings from this review may not necessarily be applicable to these groups of people.

One newly included study was ongoing in our 2013 review (OMEGA‐PAD I 2015). We detailed five remaining ongoing studies, which were unpublished at the time of writing. We tried to contact the authors of all the ongoing trials. Some stated that publications were forthcoming; others did not reply. A possible explanation for the delay in publishing might be no clear effect from the intervention.

Studies involving any form, dose, and duration of omega‐3 fatty acid supplementation were eligible. In the majority of trials, omega‐3 fatty acid supplementation came in the form of capsules. Only four studies used dietary manipulation, such as skimmed milk (Carrero 2005), fortified dairy products (Carrero 2006), products containing milled flaxseeds (FLAX‐PAD 2011), or canola oil (Stricker 2008).

Most studies used dietary manipulation or capsules containing longer‐chain omega‐3 fats (LCn3) as an intervention; only two studies used shorter‐chain omega‐3 fats (alpha‐linolenic acid (ALA)) as an intervention (FLAX‐PAD 2011; Stricker 2008). Therefore, the reader should consider that the results of the review reflect the effects of LCn3 more than the effects of ALA.

The majority of studies were limited in their sample size and duration of follow‐up. See Table 2.

The trial periods ranged from four weeks to 12 months for the primary outcomes. These studies had no long‐term follow‐up. Eight studies did report longer follow‐up periods, but were excluded from analysis due to the administration of additional active substances to their intervention groups, and substantial clinical heterogeneity (Carrero 2005; Carrero 2006; Leng 1998); the reporting of only median values (25th; 75th percentile) for any outcomes (Hammer 2019; Schiano 2008); lack of reporting of post‐intervention data for most of the outcomes (OMEGA‐PAD I 2015; OMEGA‐PAD II 2019); and reporting only baseline data for outcomes of interest (ORIGIN 2019).

Insufficient doses of omega‐3 fatty acids may have been used in at least two of the included studies. Leng 1998 administered a small daily dose of eicosapentaenoic acid (EPA; 270 mg) for two years. Similarly, Carrero 2005 and Carrero 2006 used only 130 mg docosahexaenoic acid (DHA) and 200 mg EPA per day. Five studies used a combined dose of 1.8 g to 4.6 g of EPA and DHA (Conway 2005; Gans 1990; Mori 1992; Stricker 2008; Woodcock 1984). A recently published systematic review reported that higher LCn‐3 doses reduced serum triglyceride levels and cardiovascular and cardioarterial events (Abdelhamid 2020).

Only seven studies reported adverse effects of therapy, so it is difficult to determine whether the dose or formulation of omega‐3 fatty acid is related to adverse effects (Conway 2005; Gans 1990; Leng 1998; Mackay 2012; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Schiano 2008). The fact that three studies reported that no participants experienced side effects, suggests omega‐3 fatty acids are generally well tolerated at the doses used (Gans 1990; OMEGA‐PAD I 2015; Schiano 2008). Leng 1998 reported increased gastrointestinal side effects, despite a small daily dose of EPA, suggesting that side effects are unlikely to be dose‐related. The intervention group in this study also included omega‐6 fatty acids, making it impossible to directly attribute the side effects to omega‐3 fatty acids. The severity and burden of the side effects are unclear. Conway 2005 and Gans 1990 used very similar daily doses of omega‐3 fatty acids. Three participants from the intervention group and one participant from the control group withdrew in Conway 2005 due to nausea or the inability to take the medication; no adverse effects were reported in Gans 1990. One potential explanation for the difference between the two studies is that in Gans 1990, the dose was divided into six different capsules, whereas in Conway 2005, the full daily dose was contained in one capsule. Therefore, it is possible that the administration regimen of omega‐3 fatty acids may contribute to the occurrence of adverse effects. In comparison to other studies, Mackay 2012 reported higher gastrointestinal side effects in the omega‐3 group (19%) compared to the placebo group (10%). These side effects led to six losses to follow‐up in the intervention group. None of the studies that used dairy drinks to administer the omega‐3 fatty acids addressed side effects of therapy. Therefore, we are unable to compare this method with giving omega‐3 fatty acids in capsules.

To monitor compliance amongst participants, Leng 1998, Mori 1992, OMEGA‐PAD II 2019, and Schiano 2008 counted the number of returned capsules. Similarly, Stricker 2008 assessed compliance with the return of empty bottles. Leng 1998 directly asked participants about compliance. Ten of the studies measured either plasma or platelet phospholipid fatty acid composition amongst participants, and found increases in omega‐3 fatty acid concentrations in the intervention groups (Carrero 2005; Carrero 2006; FLAX‐PAD 2011; Gans 1990; Hammer 2019; Mackay 2012; Mori 1992; OMEGA‐PAD I 2015; OMEGA‐PAD II 2019; Woodcock 1984). As well as confirming compliance, these findings demonstrated that the omega‐3 supplements were absorbed from the gastrointestinal tract. Leng 1998 found compliance to be high amongst those who completed the study, with 95% taking at least 1.12 g gamma‐linolenic acid (GLA) and 180 mg EPA daily, and 74% taking the full dose of 1.68 g GLA and 270 mg EPA. Similarly, Mori 1992 found compliance to be 99% for the fish oil group and 98% for the control group. Schiano 2008 found compliance with omega‐3 fatty acid treatment to be satisfactory (> 94%). Mackay 2012 reported the levels of compliance in both groups were 76%. Carrero 2005 and Carrero 2006 described compliance as 'good' (assessed by regular telephone calls to participants and collection of empty dairy containers). Conway 2005, Gans 1990, ORIGIN 2019, and Woodcock 1984 did not report compliance figures.

The primary outcome, quality of life, was only investigated by one of the included studies. Clinical outcomes were generally less commonly reported than biochemical or haematological outcomes, with blood pressure and side effects evaluated by almost half of the studies. Two studies reported all‐cause and vascular mortality as an outcome (FLAX‐PAD 2011; Leng 1998), while three studies reported non‐fatal coronary events (FLAX‐PAD 2011; Hammer 2019; Leng 1998), and two reported non‐fatal strokes (FLAX‐PAD 2011; OMEGA‐PAD II 2019). The results showed no significant differences between the intervention and control groups in the fatal and non‐fatal vascular events (low‐certainty evidence).

Quality of the evidence

We used the GRADE approach to assess the body of evidence for the primary and secondary outcomes of the review. This assesses the certainty of the evidence for each outcome, taking risk of bias, inconsistency, directness of evidence, imprecision, and publication bias into consideration. The risk of bias in each included study is summarised in Figure 3. We found more evidence of moderate to serious bias. However, there was some uncertainty around the key elements; in particular, threats to the internal validity arising from possible poor blinding and allocation concealment. Some studies were also at risk of selection bias. Of the 15 RCTs, only two studies were at low overall risk of bias (low risk of selection bias, performance bias, and detection bias). More details can be found in the Characteristics of included studies tables.

There was significant clinical and statistical heterogeneity between the included studies. In addition, three trials involved the administration of additional substances to the participants, making it impossible to attribute any effects of treatment specifically to omega‐3 FA, so they were excluded from the meta‐analysis (Carrero 2005; Carrero 2006; Leng 1998). This meant that the actual number of participants and the studies included in the meta‐analysis for each outcome was further reduced, in addition to studies that did not report all outcomes in a usable way. Publication bias was another unclear issue in the present review that was not further investigated, as we did not include more than ten studies in any analysis.

Very low‐certainty evidence, due to risk of bias concerns and imprecision, supported the quality of life outcome; very low‐certainty, due to risk of bias, imprecision, and inconsistency, also supported maximal walking distance and PFWD. Very low‐certainty evidence, due to the risk of bias and imprecision, supported ankle‐brachial index. Only one study reported revascularisation procedures in the lower extremity or amputation rate, so there was little evidence for either outcome to judge certainty (Leng 1998). Seven studies reported different adverse effects for both groups. We downgraded them to low‐certainty evidence. More details can be found in the Table 1. Due to the low and very low‐certainty evidence presented, caution should be applied when interpreting the results of this review.

Potential biases in the review process

None of the review authors were involved in the included or excluded studies, and none of them had any conflicts of interest. To reduce the possibility of bias, we used a broad and comprehensive search of several databases to find eligible studies. However, there might be some studies published in journals that are not indexed in the databases, so the possibility remains that not all eligible studies were detected. In addition, study selection, data collection, and risk of bias and GRADE assessment of the included studies was independently performed by two trained review authors. To minimise any potential bias in the review process, we conducted this review by meeting the standards recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2023).

Of note, the studies were not homogeneous for sample size, dose, and duration of treatment, or timing of physiological measurements. Therefore, we had to evaluate the appropriateness of pooling the individual results. Despite excluding the three most clinically heterogeneous studies from the meta‐analysis process, the Chi² test results showed significant statistical heterogeneity (I² > 50%) in the PFWD, maximal walking distance, total cholesterol, LDL, and triglyceride analyses. For this reason, caution should be taken when interpreting these results.

The analysis in this review was carried out on the post‐intervention values alone, using the mean difference (MD) between the treatment and control groups. Two new studies compared the MD between the pre‐ and post‐intervention values in groups that were not similar to the ones included in the meta‐analysis; therefore, we reported them narratively, rather than including them in the meta‐analysis (OMEGA‐PAD I 2015; OMEGA‐PAD II 2019). This meant that the baseline values were not acknowledged by the analysis.

Since the studies were all randomised, the baseline values should have been relatively equal. However, for certain outcomes, this is not the case. For example, the baseline triglyceride levels were higher in the omega‐3 group than in the control group in Gans 1990. Even though the difference between pre‐ and post‐trial values was greatest for the intervention group, the control group had a lower mean triglyceride level at the end of the study. This led to no clear mean difference between the two.

Notably, the post‐intervention data were unavailable for a factorial‐design study with 971 participants, despite contacting the trial authors (ORIGIN 2019). As this study had the largest sample size of all the included studies, being able to pool the data may have changed the results and certainty of our evidence.

The mean age range in all the included studies was between 62 and 69 years. Therefore, the results cannot be generalised to other age groups, and this should be considered. Due to the increased probability of digestive problems in the elderly, especially when taking omega‐3 supplements, the probability of non‐compliance with the treatment regimen is higher in this age group. We planned to include studies with participants older than 18 years, and examine the effect of age, but we did not identify any studies of other age groups, although age similarity reduces the possibility of clinical heterogeneity across the studies.

Most of the included studies involved more men than women, although peripheral artery disease is a condition that affects women as often, if not more often, than men. Sex‐related differences in pathophysiology and risk factors may be accompanied by a late start and often atypical symptoms in women with peripheral artery disease, in addition to the overall disease burden (Pabon 2022; Savji 2013). Therefore, the reader should consider that the results may not generalise to women.

Agreements and disagreements with other studies or reviews

The findings slightly disagreed with the original version of this review (Sommerfield 2004). The original review found that participants with intermittent claudication may have a slight increase in the triglyceride and LDL levels. These findings were not replicated in this updated version of the review, which found no impact of omega‐3 FAs on these measures when we included data from the new studies. However, our results are in agreement with results from the two older versions of the review, for these, and other outcomes (Campbell 2013; Sommerfield 2007).

The review results are also in line with two other reviews (Abdelhamid 2020; Enns 2014). A Cochrane review assessed the effects of LCn‐3 FA on the primary and secondary prevention of cardiovascular disease. They included 86 RCTs that recruited 162,796 participants at high risk of cardiovascular disease, and found little or no effect on peripheral artery disease by increasing the LCn‐3 FA intake. They also reported little or no effect of LCn‐3 on the risk of all‐cause mortality, cardiovascular events, and deaths, stroke, lipoides, and blood pressure (moderate‐ and high‐certainty evidence (Abdelhamid 2020)). Another systematic review and meta‐analysis included five studies and 396 participants. They found insufficient evidence to suggest a beneficial effect of omega‐3 polyunsaturated fatty acid supplementation in adults with peripheral artery disease in connection with cardiovascular events and other serious clinical outcomes (Enns 2014).

Authors' conclusions

Implications for practice.

In total, we included 15 randomised controlled trials (RCTs), with 1830 participants with intermittent claudication, which evaluated the effects of omega‐3 fatty acids. The evidence is very uncertain about the effect of omega‐3 fatty acid on this population on quality of life, walking distance (pain‐free or maximal), or ankle‐brachial index. Omega‐3 fatty acid supplementation has very uncertain effects on the incidence of revascularisation procedures or rate of amputation in the lower limbs. The evidence suggests that omega‐3 may have little to no effect on lipid levels, systolic and diastolic blood pressure, mortality, and vascular events (fatal and non‐fatal). The evidence suggests that omega‐3 results in little to no difference in adverse events.

Implications for research.

This update demonstrates that more research is required in this area, as currently, there is generally a poor body of evidence available for the primary outcomes. In particular, an adequately powered RCT is needed to compare the therapeutic dose of omega‐3 fatty acid with inactive placebos in people with peripheral artery disease. Research comparing different formulations and doses of omega‐3 fatty acid would be useful. New studies should adequately evaluate quality of life, walking distance, ankle‐brachial index, vascular events (fatal and non‐fatal), and adverse effects. Long‐term follow‐up of the participants is also essential. Studies examining the effects of dietary manipulations would also be valuable, as people might prefer to consume the omega‐3 fatty acids in a more natural form, rather than taking capsules.

Further research is needed to fully evaluate the short‐and long‐term effects of omega‐3 fatty acids on people living with intermittent claudication, to more fully understand the benefits and harms.

What's new

Date	Event	Description
29 October 2024	New citation required but conclusions have not changed	updated review finalized; conclusions remain unchanged with addition of new data
29 October 2024	New search has been performed	literature search updated; six new studies included (16 records); five ongoing studies identified

History

Protocol first published: Issue 4, 2002
Review first published: Issue 3, 2004

Date	Event	Description
20 March 2023	New search has been performed	New search run. Six new studies included, seven new studies excluded, and five new ongoing studies identified.
20 March 2023	New citation required but conclusions have not changed	New search run. Six new studies included, seven new studies excluded, and five ongoing studies identified. New author team involved. Text updated to reflect current Cochrane recommendations, including the addition of a Summary of findings table and GRADE assessment. Conclusions unchanged.
6 November 2012	New citation required and conclusions have changed	New author (AC) joined review team. One author (TS) stepped down from author team. Review updated with three new included studies four excluded studies and one ongoing study. Risk of bias tables added. New sections added in line with review format changes. Change in conclusion for effect of omega‐3 fatty acids: different findings from previous review versions for serum total cholesterol and LDL cholesterol levels.
6 November 2012	New search has been performed	Searches re‐run. Review updated with three new included studies, four excluded studies and one ongoing study.
6 August 2008	Amended	Converted to new review format.
21 August 2007	New citation required and conclusions have changed	Substantive amendment.Two new included studies; adverse effects added to conclusions. Dr Jackie Price added as new co‐author. Search strategy and dates updated. Copy edits performed and text updated to reflect new trials
12 August 2004	Amended	Synopsis added, date of searches amended. Minor copy edits.

Appendices

Appendix 1. Sources searched and search strategies

Source	Search strategy	Hits retrieved
1. VASCULAR REGISTER IN CRSW [Date of most recent search: 19 April 2024]	#1 Omega 3 fatty acids AND INREGISTER #2 Fats, Unsaturated AND INREGISTER #3 #1 OR #2	June 2021: 40; March 2023: 1; April 2024: 0
2. CENTRAL via CRSO [Date of most recent search: 19 April 2024]	#1 MESH DESCRIPTOR Intermittent Claudication EXPLODE ALL TREES 928 #2 MESH DESCRIPTOR Peripheral Vascular Diseases EXPLODE ALL TREES 3262 #3 MESH DESCRIPTOR Intermittent Claudication EXPLODE ALL TREES 928 #4 MESH DESCRIPTOR Peripheral Arterial Disease EXPLODE ALL TREES 1104 #5 MESH DESCRIPTOR Arterial Occlusive Diseases EXPLODE ALL TREES 12417 #6 MESH DESCRIPTOR Iliac Artery EXPLODE ALL TREES 163 #7 MESH DESCRIPTOR Popliteal Artery EXPLODE ALL TREES 361 #8 MESH DESCRIPTOR Femoral Artery EXPLODE ALL TREES 1007 #9 MESH DESCRIPTOR Tibial Arteries EXPLODE ALL TREES 41 #10 (PVD or PAOD or PAD):TI,AB,KY 4276 #11 ((arter* or vascular or vein* or veno* or peripher*) adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)):TI,AB,KY 14132 #12 (peripheral adj3 dis*):TI,AB,KY 6965 #13 claudic*:TI,AB,KY 2529 #14 arteriopathic:TI,AB,KY 6 #15 dysvascular*:TI,AB,KY 28 #16 (leg adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)):TI,AB,KY 177 #17 (limb adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)):TI,AB,KY 341 #18 (lower adj3 extrem* adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)):TI,AB,KY 137 #19 ((iliac or femoral or popliteal or femoro* or fempop* or crural or tibial) adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)):TI,AB,KY 2324 #20 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 35713 #21 MESH DESCRIPTOR Fatty Acids, Omega‐3 EXPLODE ALL TREES 3158 #22 MESH DESCRIPTOR Dietary Fats, Unsaturated EXPLODE ALL TREES 4155 #23 (fatty adj3 acid):TI,AB,KY 10088 #24 (omega adj3 acid):TI,AB,KY 2318 #25 eicosapentanoic:TI,AB,KY 101 #26 docosahexanoic:TI,AB,KY 113 #27 docosapentanoic:TI,AB,KY 1 #28 alpha‐linolenic:TI,AB,KY 509 #29 eicosapentaen*:TI,AB,KY 2740 #30 icosapentaenoic:TI,AB,KY 618 #31 docosahexaeno*:TI,AB,KY 3104 #32 (PUFA or EPA or E‐EPA or DHA or DPA or ALA):TI,AB,KY 6522 #33 linseed:TI,AB,KY 303 #34 rapeseed:TI,AB,KY 304 #35 canola:TI,AB,KY 231 #36 soy:TI,AB,KY 2526 #37 alga:TI,AB,KY 101 #38 hemp:TI,AB,KY 67 #39 (fish oil):TI,AB,KY 2841 #40 (mackerel or sardine or trout or herring or salmon):TI,AB,KY 770 #41 #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 20371 #42 #20 AND #41 468	June 2021: 468; March 2023: 41; April 2024: 17
3. Clinicaltrials.gov [Date of most recent search: 19 April 2024]	Interventional Studies | Intermittent Claudication OR Peripheral Vascular Diseases OR Peripheral Arterial Disease OR Arterial Occlusive Diseases | Fatty Acids, Omega‐3 OR Dietary Fats, Unsaturated	June 2021: 34; March 2023: 1; April 2024: 2
4. ICTRP Search Portal [Date of most recent search: 19 April 2024]	Intermittent Claudication OR Peripheral Vascular Diseases OR Peripheral Arterial Disease OR Arterial Occlusive Diseases | Fatty Acids, Omega‐3 OR Dietary Fats, Unsaturated	June 2021: 2; March 2023: 0; April 2024: 0
5. MEDLINE (Ovid MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE) 1946 to present [Date of most recent search: 19 April 2024]	1 Intermittent Claudication/ 2 exp Peripheral Vascular Diseases/ 3 exp Peripheral Arterial Disease/ 4 exp Arterial Occlusive Diseases/ 5 exp Leg/bs 6 Iliac Artery/ 7 Popliteal Artery/ 8 Femoral Artery/ 9 Tibial Arteries/ 10 (PVD or PAOD or PAD).ti,ab. 11 ((arter* or vascular or vein* or veno* or peripher*) adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 12 (peripheral adj3 dis*).ti,ab. 13 claudic*.ti,ab. 14 arteriopathic.ti,ab. 15 dysvascular*.ti,ab. 16 (leg adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 17 (limb adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 18 (lower adj3 extrem* adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 19 ((iliac or femoral or popliteal or femoro* or fempop* or crural or tibial) adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 20 or/1‐19 21 exp Fatty Acids, Omega‐3/ 22 exp Dietary Fats, Unsaturated/ 23 (fatty adj3 acid).ti,ab. 24 (omega adj3 acid).ti,ab. 25 eicosapentanoic.ti,ab. 26 docosahexanoic.ti,ab. 27 docosapentanoic.ti,ab. 28 alpha‐linolenic.ti,ab. 29 eicosapentaen*.ti,ab. 30 icosapentaenoic.ti,ab. 31 docosahexaeno*.ti,ab. 32 (PUFA or EPA or E‐EPA or DHA or DPA or ALA).ti,ab. 33 linseed.ti,ab. 34 rapeseed.ti,ab. 35 canola.ti,ab. 36 soy.ti,ab. 37 alga.ti,ab. 38 hemp.ti,ab. 39 fish oil.ti,ab. 40 (mackerel or sardine or trout or herring or salmon).ti,ab. 41 or/21‐40 42 20 and 41 43 randomized controlled trial.pt. 44 controlled clinical trial.pt. 45 randomized.ab. 46 placebo.ab. 47 drug therapy.fs. 48 randomly.ab. 49 trial.ab. 50 groups.ab. 51 or/43‐50 52 exp animals/ not humans.sh. 53 51 not 52 54 42 and 53	June 2021: 1016; March 2023: 155; April 2024: 47
6. EMBASE via OVID 1972‐present [Date of most recent search: 19 April 2024]	1 intermittent claudication/ 2 exp peripheral vascular disease/ 3 exp peripheral occlusive artery disease/ 4 iliac artery/ 5 popliteal artery/ 6 femoral artery/ 7 tibial artery/ 8 (PVD or PAOD).ti,ab. 9 ((arter* or vascular or vein* or veno* or peripher*) adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 10 (peripheral adj3 dis*).ti,ab. 11 claudic*.ti,ab. 12 arteriopathic.ti,ab. 13 dysvascular*.ti,ab. 14 (leg adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 15 (limb adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 16 (lower adj3 extrem* adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 17 ((iliac or femoral or popliteal or femoro* or fempop* or crural or tibial) adj3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)).ti,ab. 18 or/1‐17 19 exp omega 3 fatty acid/ 20 exp edible oil/ 21 fatty acid.ti,ab. 22 (omega adj2 acid).ti,ab. 23 eicosapentanoic.ti,ab. 24 docosahexanoic.ti,ab. 25 docosapentanoic.ti,ab. 26 alpha‐linolenic.ti,ab. 27 eicosapentaen*.ti,ab. 28 icosapentaenoic.ti,ab. 29 docosahexaeno*.ti,ab. 30 PUFA.ti,ab. 31 linseed.ti,ab. 32 rapeseed.ti,ab. 33 canola.ti,ab. 34 soy.ti,ab. 35 alga.ti,ab. 36 hemp.ti,ab. 37 fish oil.ti,ab. 38 (mackerel or sardine or trout or herring or salmon).ti,ab. 39 or/19‐38 40 18 and 39 41 randomized controlled trial/ 42 controlled clinical trial/ 43 random$.ti,ab. 44 randomization/ 45 intermethod comparison/ 46 placebo.ti,ab. 47 (compare or compared or comparison).ti. 48 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 49 (open adj label).ti,ab. 50 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 51 double blind procedure/ 52 parallel group$1.ti,ab. 53 (crossover or cross over).ti,ab. 54 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 55 (assigned or allocated).ti,ab. 56 (controlled adj7 (study or design or trial)).ti,ab. 57 (volunteer or volunteers).ti,ab. 58 trial.ti. 59 or/41‐58 60 40 and 59	June 2021: 4114 March 2023: 559; April 2024: 325
7. CINAHL via Ebsco [Date of most recent search: 19 April 2024]	S56 S41 AND S55 S55 S42 OR S43 OR S44 OR S45 OR S46 OR S47 OR S48 OR S49 OR S50 OR S51 OR S52 OR S53 OR S54 S54 MH "Random Assignment" S53 MH "Triple‐Blind Studies" S52 MH "Double‐Blind Studies" S51 MH "Single‐Blind Studies" S50 MH "Crossover Design" S49 MH "Factorial Design" S48 MH "Placebos" S47 MH "Clinical Trials" S46 TX "multi‐centre study" OR "multi‐center study" OR "multicentre study" OR "multicenter study" OR "multi‐site study" S45 TX crossover OR "cross‐over" S44 AB placebo* S43 TX random* S42 TX "latin square" S41 S19 AND S40 S40 S20 OR S21 OR S22 OR S23 OR S24 OR S25 OR S26 OR S27 OR S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 OR S37 OR S38 OR S39 S39 TX mackerel or sardine or trout or herring or salmon S38 TX fish oil S37 TX hemp S36 TX alga S35 TX soy S34 TX canola S33 TX rapeseed S32 TX linseed S31 TX PUFA S30 TX docosahexaeno* S29 TX icosapentaenoic S28 TX eicosapentaen* S27 TX alpha‐linolenic S26 TX docosapentanoic S25 TX docosahexanoic S24 TX eicosapentanoic S23 TX omega N3 acid S22 TX fatty N3 acid S21 (MH "Dietary Fats+") S20 (MH "Fatty Acids, Omega‐3+") S19 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 S18 TX (iliac or femoral or popliteal or femoro* or fempop* or crural or tibial) N3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*) S17 TX ((lower extrem*) N3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*)) S16 TX limb N3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*) S15 TX (leg N3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*) S14 TX dysvascular* S13 TX arteriopathic S12 TX claudic* S11 TX peripheral N3 dis* S10 TX (arter* or vascular or vein* or veno* or peripher*) N3 (occlus* or steno* or obstruct* or lesio* or block* or harden* or stiffen* or obliter*) S9 TX PVD or PAOD S8 (MH "Tibial Arteries") S7 (MH "Femoral Artery") S6 (MH "Popliteal Artery") S5 (MH "Iliac Artery") S4 (MH "Leg/BS") S3 (MH "Arterial Occlusive Diseases+") S2 (MH "Peripheral Vascular Diseases+") S1 (MH "Intermittent Claudication")	June 2021: 401; March 2023: 25; April 2024: 12
TOTAL before deduplication		June 2021: 6075; March 2023: 482; April 2024: 403
TOTAL after deduplication		June 2021: 5219; March 2023: 701; April 2024: 369

Data and analyses

Comparison 1. Omega‐3 fatty acids (FA) versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Walking distance	3		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.1.1 Pain‐free walking distance	3	147	Mean Difference (IV, Random, 95% CI)	1.01 [‐34.23, 36.24]
1.1.2 Maximum walking distance	3	164	Mean Difference (IV, Random, 95% CI)	‐4.18 [‐37.10, 28.74]
1.2 Ankle brachial index	3	168	Mean Difference (IV, Fixed, 95% CI)	‐0.02 [‐0.08, 0.04]
1.2.1 Post‐intervention	3	168	Mean Difference (IV, Fixed, 95% CI)	‐0.02 [‐0.08, 0.04]
1.3 Cholesterol	5		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.3.1 Total cholesterol	5	206	Mean Difference (IV, Random, 95% CI)	0.10 [‐0.48, 0.67]
1.3.2 High‐density lipoprotein cholesterol	5	206	Mean Difference (IV, Random, 95% CI)	‐0.07 [‐0.16, 0.02]
1.3.3 Low‐density lipoprotein cholesterol	4	186	Mean Difference (IV, Random, 95% CI)	0.28 [‐0.28, 0.85]
1.3.4 Very low‐density lipoprotein cholesterol	1	33	Mean Difference (IV, Random, 95% CI)	0.08 [‐0.47, 0.63]
1.4 Triglycerides	4	165	Mean Difference (IV, Random, 95% CI)	‐0.14 [‐0.59, 0.30]
1.4.1 Post‐intervention	4	165	Mean Difference (IV, Random, 95% CI)	‐0.14 [‐0.59, 0.30]
1.5 Blood pressure	2		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
1.5.1 Systolic blood pressure	2	118	Mean Difference (IV, Fixed, 95% CI)	‐4.36 [‐11.37, 2.65]
1.5.2 Diastolic blood pressure	2	118	Mean Difference (IV, Fixed, 95% CI)	‐3.54 [‐7.14, 0.06]
1.6 Vascular events	3		Odds Ratio (IV, Fixed, 95% CI)	Subtotals only
1.6.1 Non‐fatal coronary events	2	141	Odds Ratio (IV, Fixed, 95% CI)	0.59 [0.13, 2.60]
1.6.2 Non‐fatal stroke	2	110	Odds Ratio (IV, Fixed, 95% CI)	0.95 [0.13, 6.77]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Carrero 2005.

Study characteristics
Methods	Design: single centre, 2‐arm, RCT (500 mL/day of skimmed milk containing 200 mg EPA, 130 mg DHA vs 500 mL/day of skimmed milk), 12 months Summary risk of bias*: unclear risk of bias
Participants	Country: Spain Setting: hospital Number of centres: 1 Number of participants: 60 (30 in each group) Age (mean ± SD): intervention group 62.4 ± 1.6 years, control group 65.6 ± 1.7 years Sex: male Inclusion criteria: all males diagnosed with PAD (ABI < 0.70) and presenting with IC (Fontaine Stage IIb, claudication distance < 200 m) Exclusion criteria: eligible for revascularisation surgery, endocrine or metabolic disturbances, cardiac history, resident outside the Granada region of Spain
Interventions	Treatment: 500 mL/day of skimmed milk containing 200 mg EPA, 130 mg DHA plus oleic acid, folic acid and vitamins A, B6, D, E Control: 500 mL/day semi‐skimmed milk plus vitamins A and D Duration: 12 months
Outcomes	PFWD ABI Cholesterol (total, LDL, HDL) Triglycerides
Study funding sources	Supported in part by a PhD educational grant from the University of Granada (J.J.C.). Milk products, reagents, and kits were supplied by Puleva Food S.L., Granada, Spain
Declarations of interest	One of the authors was an employee of Puleva Biotech.
Contact	Authors' names: Carrero JJ, Lopez‐Huertas E, Salmeron LM, Baro L, Ros E Institution: Department of Biochemistry and Molecular Biology, University of Granada, Granada, Spain Address: University of Granada, Granada, Spain
Notes	We judged this study to have an unclear summary ROB due to unclear selection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A table of random numbers was used
Allocation concealment (selection bias)	Unclear risk	No description of allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	From correspondence: "dairy drinks in white tetrapak containers, of the same appearance with the key A or B, performed by the supplier"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	From correspondence: "The teams dealing with the analysis of samples only received numbered samples. The medical team collecting data did not have indication of the type of drink assigned. Only the team dealing with the production of the dairy drinks would know the meaning of keys A and B. The meaning of keys A or B was disclosed once all the data was obtained"
Incomplete outcome data (attrition bias) All outcomes	High risk	Details are provided on request for all missing outcome data; however, the withdrawals and exclusion led to an imbalance between the study groups 107 eligible candidates were originally identified. 34 then excluded (11 living outside Granada, 12 previous MI, 6 statins prescription before time of inclusion, 5 refused to participate). A further 13 were excluded due to budget restrictions, leaving 60 to be randomised 4 were excluded from the control group post randomisation (2 due to change of residence, 1 due to depression and 1 due to statin prescription). 0 were excluded post randomisation from the supplement group. This meant that 30 were analysed from the supplemented group but only 26 were analysed from the controlled group
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported in the prespecified way
Other bias	Unclear risk	Intervention preparation containing omega‐3 fatty acids, contained oleic acid and certain vitamins (not included in the placebo preparation). However, the omega‐3 fatty acid dose was very small, but it would be impossible to attribute any observed effects to omega‐3 fatty acids.

Carrero 2006.

Study characteristics
Methods	Design: single centre, 4‐arm, single blinded, RCT (receiving 500 mL/day of a fortified dairy product containing fish oil, oleic acid, folic acid, and other vitamins vs the same fortified product plus 20 mg/day of simvastatin vs 500 mL/day of semi‐skimmed milk vs semi‐skimmed milk plus simvastatin), 12 months Summary risk of bias*: unclear risk of bias
Participants	Country: Spain Settling: hospital Number of centres: 1 Number of participants: 40 (10 in each group) Age (mean): 66:66:65:63 years per group Sex: male Inclusion criteria: participants diagnosed with PAD (ABI < 0.7) and IC (Fontaine stage IIb, claudication distance < 200 m) Exclusion criteria: eligibility for revascularisation surgery; endocrine or metabolic disturbances (such as hypothyroidism or obesity, BMI > 30 kg/m2; history of cardiac episodes (such as angina pectoris) or previous acute MI; living outside of Granada
Interventions	Treatment 1: 500 mL/day of a fortified dairy product containing fish oil, oleic acid, folic acid, and other vitamins Treatment 2: same fortified dairy product plus 20 mg/day of simvastatin Control 1: 500 mL/day of semi‐skimmed milk Control 2: semi‐skimmed milk plus simvastatin Duration: 12 months
Outcomes	ABI PFWD Cholesterol (total, HDL, LDL) Triglycerides
Study funding sources	Study supported by a Ph.D. educational grant from the University of Granada and by Puleva Biotech SA, Granada Spain
Declarations of interest	No details provided
Contact	Authors' names: Carrero JJ, Lopez‐Huertas E, Salmeron LM, Ramos VE, Baro L, Ros E Institution: Department of Biochemistry and Molecular Biology, University of Granada, Granada, Spain Address: Department of Biochemistry and Molecular Biology, University of Granada, Granada 18071, Spain
Notes	Groups very small and poorly matched for age at baseline We judged this study to have an unclear summary ROB due to unclear selection bias, performance and detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A table of random numbers was used
Allocation concealment (selection bias)	Unclear risk	Allocation concealment not described
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Study states that "Products were packaged in white 500ml Tetra Bricks to perform a blind study." Unclear whether investigators were blinded or not
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear whether assessors were blinded or not. Detection bias is unlikely to have affected the objective outcomes (ABI, total cholesterol, HDL, LDL, triglycerides) but may have had an effect on PFWD, which is slightly more subjective
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Number of exclusions post randomisation not stated.
Selective reporting (reporting bias)	Low risk	Prespecified outcomes were reported in the prespecified way
Other bias	Unclear risk	Claudication distance for baseline (when participants were recruited) and T0 (the start of the study) were very different with no obvious explanation for this. Very small groups poorly matched for age at baseline. Intervention preparation containing omega‐3 fatty acids, contained oleic acid and certain vitamins (not included in the placebo preparation). However, the omega‐3 fatty acid dose was very small but it would be impossible to attribute any observed effects to omega‐3 fatty acids.

Conway 2005.

Study characteristics
Methods	Design: 2‐arm, parallel RCT, double‐blind (capsules containing 1.7 g EPA and 1.15 g DHA daily vs placebo), 16 weeks follow up Summary risk of bias*: unclear risk of bias
Participants	Country: UK Setting: not reported Number of settings: not reported Number of participants: 50 Age (mean (range)): 66.1 years (49 ‐82 years) Sex: 35 female; 15 male Inclusion criteria: participants with IC who have been referred to a nurse‐led exercise programme Exclusion criteria: critical ischaemia, warfarin therapy, haemorrhagic disorders, DM, aspirin‐sensitive asthma
Interventions	Treatment: 10 g fish oils per day (capsules containing 1.7 g EPA, 1.15 g DHA) Control: capsules containing 10 g of mixed oils Duration: 16 weeks
Outcomes	QoL PFWD MWD ABI Adverse effects
Study funding sources	No details
Declarations of interest	No details
Contact	No details
Notes	We judged this study to have an unclear summary ROB due to unclear performance and detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	From correspondence: "Twenty five patients were randomised by computer software (Microsoft® Excel 2000 Number Generator) to each arm of the trial"
Allocation concealment (selection bias)	Low risk	From correspondence: "Following randomisation, allocations were placed in order, in sealed opaque envelopes in the pharmacy"
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Claims to be "double‐blinded" and in correspondence author states that: "Trial packs containing either medication or placebo were selected sequentially by the pharmacist in accordance with the randomisation code." However, no further information is provided and it is unclear whether there was a visible difference between the medication or placebo packs
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear whether assessors were blinded or not. Detection bias is unlikely to have had an effect on the QoL or ABI outcomes; however, may have had an impact on assessment of PFWD and MWD
Incomplete outcome data (attrition bias) All outcomes	Low risk	10 participants did not complete the study. 2 participants had an MI (1 from the treatment group and 1 from the control group), 4 withdrew due to nausea or an inability to swallow the capsules (3 from the intervention group, 1 from the control group), 2 were lost to follow‐up and 2 were unable to complete treadmill retesting. Power calculation and intention‐to‐treat analysis was carried out
Selective reporting (reporting bias)	Low risk	The prespecified outcomes in the study protocol were reported in the prespecified way
Other bias	Low risk	No other obvious source of bias

FLAX‐PAD 2011.

Study characteristics
Methods	Design: single centre, 2‐arm, RCT, double‐blind (a diet supplemented with 30g of milled flaxseed vs placebo), 12 months Summary risk of bias*: low risk of bias
Participants	Country: Canada Setting: hospital Number of centres: 1 Number of participants: 58 intervention, 52 control Age (Mean ±SD): intervention 66.8 ± 7.7; control 66.1 ± 9.3 Sex: 79% males in intervention; 71% males in control Inclusion criteria: male or female over 40 years old with PAD for more than 6 months or IC secondary to lower extremity atherosclerotic arterial disease that confirmed with ABI ≤ to 0.9 in one or both legs or who have had a previous intervention for PAD; able to comply with protocol requirements and informed consent; participants taking lipid‐lowering or anti‐platelet drugs must be on a stable dose for 3 months prior and during the study Exclusion criteria: patients with ischemia, rest pain in limbs, ulceration, or gangrene; any condition that prevents walking on a treadmill; history of major bleeding; have a bowel disease (e.g., Crohn's disease, celiac disease, peptic ulcer disease, irritable bowel syndrome and diverticulosis); patients with an estimated life expectancy less than 2 years and with high baseline cardiac risk; moderate to severe renal failure; taking other supplements; fish limitations (no more than 2 fish meals/week); gluten allergy; allergies to any ingredient in the study product or placebo; plan to undergo surgery during the trial
Interventions	Treatment: food containing 30 g of milled flaxseed provided to the participants included buns, snack bars, muffins, bagels, pasta, and tea biscuits. Control: placebo (the products containing 30 g of milled wheat provided to the participants included buns, snack bars, muffins, bagels, pasta, and tea biscuits) Duration: 12 months
Outcomes	PFWD MWD ABI Cholesterol (total, HDL, LDL) Triglycerides DBP, DBP All‐cause and vascular mortality Non‐fatal vascular events
Study funding sources	Supported by grants from Flax2015, the Canola Council of Canada, the Agri‐Food Research and Development Initiative, and the Canadian Institutes for Health Research (CIHR). ALE was a recipient of the Banting and Best Canada Graduate Scholarship Doctoral Award from CIHR as well as a recipient of a Doctoral Research Scholarship from the Heart and Stroke Foundation of Canada. SPBC was a recipient of the Patient‐Oriented Research Doctoral Scholarship from CIHR. Support in the form of creating and producing the bagels for the study was provided by Canada Bread. The flaxseed or wheat was provided gratis by Glanbia Nutritionals, Inc., and The Canadian International Grain Institute, Winnipeg, Manitoba, Canada, and the Food Development Center, Portage la Prairie, Manitoba, Canada, created and produced the buns, bars, tea biscuits, muffins, and pasta for this study. Indirect research support was obtained from St. Boniface Hospital Foundation.
Declarations of interest	The authors declare that they have no conflict of interest
Contact	Authors' names: Edel AL, Rodriguez‐Leyva D, Maddaford TG, Caligiuri SPB, AustriaN JA, Weighell W, Guzman R, Aliani M, and Pierce GN Institution: University of Manitoba, Winnipeg, Canada Address: Department of Human Nutritional Sciences, Faculty of Agricultural and Food Sciences, University of Manitoba, Winnipeg, Canada
Notes	We judged this study has low summary ROB as there was no bias in the main domains
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Subjects are randomized to one of two possible groups by a computer generated randomization schedule." ...."Randomization of patients was performed, according to the procedure of minimization, by a blinded person at the coordinating center using dedicated software."
Allocation concealment (selection bias)	Low risk	Quote: "Allocation was concealed. The person who determined if a subject was eligible for inclusion in the trial was unaware, when this decision was made, of which group the subject would be allocated."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Participants, personnel administering the intervention and those assessing the outcomes were blinded to group assignment." Quote: "The intervention group receives products which contain thirty grams of milled flaxseed. The control group receives similar products made with wheat. Product color and texture were similar to disguise the composition of the product."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Participants, personnel administering the intervention and those assessing the outcomes were blinded to group assignment."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Dropout is low, balanced across groups (21 ‐ 22%) and reasons for dropout reported. Flaxseed: 43/58 assessed, placebo: 41/52 assessed
Selective reporting (reporting bias)	Unclear risk	NCT00781950 study started in October 2008 and completed in March 2011. Appears to be discrepancies in outcomes between trial registry information and publications
Other bias	Low risk	No other bias identified

Gans 1990.

Study characteristics
Methods	Design: 2‐arm, RCT, double‐blind (daily dose of 1.8 g EPA + 1.2 g DHA vs placebo with 3 g linoleic acid per day), 4 months Summary risk of bias*: unclear risk of bias
Participants	Country: The Netherlands Setting: not reported Number of centres: not reported Number of participants: 32 (16 patients in each group) Age (mean ± SD): fish oil group 65.5 ± 10.4; corn oil group 66.7 ± 5.3 Sex: male 22; female 10 Inclusion criteria: symptoms of IC due to atherosclerotic disease (Fontaine IIa and IIb). Stable for 1 year according to ABI and claudication distance on standardised testing Exclusion criteria: unstable angina or MI in past 3 months; any illness with rapid evolution; rest pain; gangrene; DBP > 100 mmHg; poorly controlled DM; (HbA1c > 12%); vasculitis, thromboangiitis obliterans; platelet count > 500,000x109/L or < 90,000x109/L; haematocrit > 55%; fish allergy; lipid‐lowering or platelet active drugs
Interventions	Treatment: 6 capsules of fish oil per day (daily dose of 1.8 g EPA + 1.2 g DHA) Control: placebo 6 capsules of corn oil per day (3 g linoleic acid per day) Duration: 4 months
Outcomes	PFWD MWD ABI Cholesterol (total, HDL, LDL) Triglycerides BP Adverse effects
Study funding sources	No details provided
Declarations of interest	No details provided
Contact	Authors' names: Gans RO, Bilo HJ, Weersink EG, Rauwerda JA, Fonk T, Popp‐Snijders C, Donker AJ Institution: Free University Hospital, Amsterdam, The Netherlands Address: Free University Hospital, Amsterdam, The Netherlands
Notes	We judged this study to have an unclear summary ROB due to unclear selection, performance and detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Study states that "Patients were randomly assigned"; however, no further details are provided.
Allocation concealment (selection bias)	Low risk	Allocation is carried out by a "numbered sealed envelope system"
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Claims to be "double‐blind" study but no information given regarding a method of blinding Poor blinding is unlikely to have had an effect on the majority of outcomes which were objective, i.e. BP, ABI, blood and plasma viscosity, cholesterol, triglycerides, but may have had an impact on the more subjective outcomes, i.e. PFWD and MWD
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear whether assessors were blinded or not. Detection bias is unlikely to be an issue for the objective outcomes listed but may potentially have had an effect on the more subjective outcomes
Incomplete outcome data (attrition bias) All outcomes	Low risk	Data provided on all missing outcome data following request to study authors, 3 participants (2 from the intervention group and 1 from the control group) were excluded from the study post‐randomisation due to concurrent medical conditions (2 were diabetic and 1 required eye surgery). 2 participants in the fish oil group later dropped out; 1 sustained a spinal fracture and 1 developed unstable ischaemia in his non‐claudicated leg. 16/18 (89%) of the intervention group and 16/19 (84%) of the control group, therefore, completed the study
Selective reporting (reporting bias)	Low risk	Prespecified outcomes were reported in the prespecified way
Other bias	Low risk	No other obvious sources of bias

Hammer 2019.

Study characteristics
Methods	Design: single centre, 2‐arm, parallel RCT, double‐blind (capsules n‐3 EPA + DHA 4 g/d vs placebo), 6 months follow up Summary risk of bias*: low risk of bias
Participants	Country: Austria Setting: The Vascular Medicine outpatient clinic Number of centres: 1 Number of participants: 35 intervention, 35 control Age (median (IQR)): Omega group 66.3 (12.6); Placebo group 66.7 (7.1) Sex: male 55; female 15 Inclusion criteria: history of IC (Rutherford categories 2 – 3), exercise tolerance limited by IC during a screening treadmill test and an ABI at rest < 0·9 Exclusion criteria: rest pain or ischaemic ulcers, exercise tolerance limited by factors other than claudication (e.g. coronary artery disease, dyspnoea, poorly controlled blood pressure, any kind of restriction of the musculoskeletal system) and planned revascularisation procedure within the next 3 months (including angioplasty and bypass surgery)
Interventions	Treatment: 4 g daily of n‐3 PUFA (high‐puritycn‐3 PUFA (850–882 mg EPA and DHA as ethyl esters in the average ratio of EPA/DHA 1·2:1 per 1 g soft gelatine capsules marketed as Omacor Control: placebo was produced with an identical appearance to Omacor® consisting of a gelatine capsulated nutrition supplement containing fatty acids, which typically occur in the Austrian diet Duration: 3 months
Outcomes	PFWD MWD ABI Cholesterol (total, HDL, LDL) Triglycerides Non‐fatal vascular events SBP and DBP
Study funding sources	Vascular Medicine outpatient clinic of the Department of Internal Medicine II, Medical University of Vienna, Austria; no external funding was provided
Declarations of interest	None of the authors has any conflicts of interest to declare
Contact	Authors' names: Hammer A, Moertl D, Schlager O, Matschuck M, Seidinger D, Koppensteiner R, Steiner S Institution: Vascular Medicine outpatient clinic of the Department of Internal Medicine II, Medical University of Vienna, Austria Address: Medical University of Vienna, Austria
Notes	We judged this study to have low summary ROB as no bias was detected in the main domains
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Participants were electronically randomised (block‐wise randomisation with random block size)"
Allocation concealment (selection bias)	Low risk	Quote:"After providing informed consent, the study physician contacted the Academic Studies Support Office for randomisation. Packages were provided with unique numbers, which were also used as participant identification numbers. Only persons from the Academic Studies Support Office, which were not otherwise involved in the study, had access to the coding"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Physicians and patients were supplied with medication packages of equal appearance and weight containing an active drug or placebo produced by the local hospital pharmacy. Packages were provided with unique numbers, which were also used as participant identification numbers. Only persons from the Academic Studies Support Office, which were not otherwise involved in the study, had access to the coding"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "All measurements were performed by one experienced ultrasound specialist blinded to the treatment strategy"
Incomplete outcome data (attrition bias) All outcomes	Low risk	24/35 n‐3 PUFA participants analysed and 31/35 placebo participants analysed ‐ but it is conducted. Note the figure on the flow chart in the right‐hand column is the placebo group but is incorrectly named (both groups are named 'n‐3 PUFA 4 g/d)
Selective reporting (reporting bias)	Low risk	The registered protocol was checked, and all outcomes were reported.
Other bias	Low risk	No other bias identified

Leng 1998.

Study characteristics
Methods	Design: 2‐arm, parallel RCT, double‐blind (capsules 280 mg GLA and 45 mg EPA daily vs placebo), 2 years follow up Summary risk of bias*: unclear risk of bias
Participants	Country: UK Setting: Peripheral Vascular Clinic at the Royal Infirmary of Edinburgh Number of centres: 1 Number of participants: 120; 60 in each group Age (mean (SE)): treatment group 65 years (0.94); placebo group 66.48 years (0.90) Sex: male 84; female 36 Inclusion criteria: Participants with IC on the Edinburgh Claudication Questionnaire, ABI < 0.9 in at least 1 limb, stable > 6 months Exclusion criteria: clinical evidence of critical ischaemia (rest pain, ulcer, gangrene); previous or impending arterial surgery or angioplasty (within 3 months); unstable angina or MI within previous 3 months; severe intercurrent illnesses; concurrent treatment with anticoagulants, other oils, lithium or phenothiazines; and pregnancy
Interventions	Treatment: 2 capsules evening primrose oil and fish oil bd for 2 weeks then 3 capsules bd for 2 years (280 mg GLA and 45 mg EPA per day) Control: identical placebo capsules containing 500 mg sunflower oil (linoleic acid)
Outcomes	PFWD ABI Cholesterol (total, HDL and LDL) BP All‐cause and vascular mortality Non‐fatal vascular events Adverse effects
Study funding sources	No details provided
Declarations of interest	No details provided
Contact	Authors' names: Leng GC, Lee AJ, Fowkes FG, Jepson RG, Lowe GD, Skinner ER, Mowat BF Institution: Department of Primary Care and Population Sciences, Royal Free Hospital Address: Department of Primary Care and Population Sciences, Royal Free Hospital, Rowland Hill Street, London, UK
Notes	We judged this study to have an unclear summary ROB due to unclear selection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Performed in blocks of 4 using a minimisation method taking into account age, sex, smoking habits and regular aspirin consumption
Allocation concealment (selection bias)	Low risk	Code number allocated by telephone
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Treatment and placebo groups were given identical capsules
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear whether assessors were blinded or not. Unlikely to have affected most outcomes as they were objective measurements analysed in laboratories. However, it may have affected PFWD, which entails a little more subjectivity
Incomplete outcome data (attrition bias) All outcomes	High risk	Most missing outcome data were documented (45 in total). 6 of the intervention group and 11 of the control group chose not to complete the study. In addition, 15 of the intervention group and 13 of the control group were withdrawn by the trial organisers. However, some additional missing data for some outcomes were not fully reported. For cholesterol and lipoprotein levels, only data from 37 of the 39 people that completed the trial in the treatment group were included in the results. For blood viscosity, data from 37 out of 39 people from the treatment group and 33 out of 36 from the placebo group were included in the results. For plasma viscosity, 37 out of 39 people from the treatment group and 35 out of 37 from the placebo group were included in the results. No reasons for this missing data were provided. Questionnaires were sent to participants and their general practitioners 2 years after recruitment to identify CV events in those who withdrew from the trial
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported in the prespecified way
Other bias	High risk	The authors evaluated the effect of both omega‐3 and omega‐6 fatty acid supplementation. This makes it difficult to attribute any particular effects to omega‐3 fatty acids alone, since omega‐6 fatty acids have also been described as having a total cholesterol‐lowering effect and a slight positive effect on modification of TG, HDL and LDL levels in the blood. They used linoleic acid as a placebo, even though this is a precursor compound of GLA. The authors acknowledge in the paper that the linoleic acid given to the control group may have been (at least partly) metabolised into GLA.

Mackay 2012.

Study characteristics
Methods	Design: 2‐arm, crossover RCT, double‐blind (capsules n‐3 EPA + DHA 4 g/d vs placebo), 24 weeks follow up Summary risk of bias*: unclear risk of bias
Participants	Country: UK Setting: clinic and claudication database Number of centres: 1 Number of participants: 77 intervention, 73 control Age (Mean (SD)): group A 68.5 (9.3); group B 68.6 (9.4) Sex: men 104; women 46 Inclusion criteria: ABI less than 0.8 and receiving statin and aspirin therapy Exclusion criteria: patients unable to give informed consent and those who ate over two portions of oily fish per week, already taking fish oil supplements, patients with rest pain or ulceration, liver impairment, abnormal platelet count, diabetes and taking clopidogrel, warfarin or non‐steroidal anti‐inflammatory drugs
Interventions	Type: supplement (capsule) Treatment: one capsule per day of n‐3 PUFA (850–882 mg EPA and DHA) marketed as Omacor (6 weeks of OMACOR supplementation followed by 12‐week washout followed by 6 weeks of placebo) Control: placebo (an 80:20 blend of palm and soybean oils) Duration: 6 weeks treatment We used data from first phase only and study results were adjusted for order of treatment
Outcomes	Adverse effects
Study funding sources	British heart foundation and omega‐3 capsules supplied by Solvay healthcare
Declarations of interest	No details provided
Contact	Authors' names: Mackay I, Ford I, Thies F, Fielding S, Bachoo P, Brittenden J Institution: Division of Applied Medicine, C/o Vascular Unit Address: Aberdeen Royal Infirmary, Foresterhill, AB25 2ZN, United Kingdom
Notes	We judged this study to have an unclear summary ROB due to unclear detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Pharmacists dispense placebo or active drug packs according to a computer‐generated randomisation process."
Allocation concealment (selection bias)	Low risk	Quote: "The code was held by the trial drugs pharmacy department and was only revealed to the researchers once recruitment, data collection and laboratory analyses were complete."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "All study personnel and participants were blinded to treatment assignment for the duration of the study." ...." The placebo and supplement were indistinguishable and independently packaged."
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	17/73 group A and 11/77 group B were lost to follow‐up, presumed not included in analyses = 23 vs 14%, so dropouts are not balanced
Selective reporting (reporting bias)	Low risk	The trial was registered (ISCRTN: 54802970). All outcomes reported in the trial registry were reported
Other bias	Low risk	No other bias identified

Mori 1992.

Study characteristics
Methods	Design: 2‐arm, parallel RCT, double‐blind (capsules 2.8 g EPA and 1.8 g DHA daily vs placebo), 4 weeks follow up Summary risk of bias*: unclear risk of bias
Participants	Country: Australia Setting: 1 Number of centres: hospital Number of participants: 32 Age: range 47–71 years; fish oil group 61.9 ± 1.2; olive oil group 61.9 ± 1.7 (Mean +/‐ SEM) Sex: male Inclusion criteria: symptomatic and angiographically demonstrated PAD Exclusion criteria: renal impairment, generalised inflammatory disorders
Interventions	Treatment: 5 capsules of fish oil 3 times per day with meals (2.8 g EPA, 1.8 g DHA per day) Control: 5 capsules of olive oil 3 times per day (11.2 g oleic acid per day) Duration: 4 weeks
Outcomes	Cholesterol (total, HDL, LDL) Triglycerides
Study funding sources	Supported by the National Heart Foundation of Australia. The fish oil and olive oil capsules were donated by Reckitt & Colman Pharmaceuticals, Sydney, and RP Scherer Pty Ltd, Melbourne, Australia
Declarations of interest	No details provided
Contact	Authors' names: Mori TA, Vandongen R, Mahanian F, Douglas A. Institution: University Department of Medicine, School of Medicine, Royal Perth Hospital Address: University Department of Medicine, School of Medicine, 35 Victoria Square, Perth, Western Australia
Notes	No data on severity of PAD. Some participants underwent peripheral vascular surgery at least 3 months before the start of the study. 1 IDDM in the intervention group, 1 NIDDM in the control group Controls matched for age and BMI We judged this study to have an unclear summary ROB due to unclear selection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Table of random numbers
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not stated
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Claims to be a "double‐blind" study. Although there are no further details about the blinding process, poor blinding is unlikely to have impacted on the outcomes
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear whether assessors were blinded or not. However, all outcomes are objective measurements, so unlikely to have had an impact
Incomplete outcome data (attrition bias) All outcomes	Low risk	3 people withdrew post randomisation but before commencement of the study (1 from the fish oil group; 2 from the olive oil group). All 29 participants who commenced the study completed the 10‐week trial
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported in the prespecified way
Other bias	Unclear risk	No data on severity of PAD

OMEGA‐PAD I 2015.

Study characteristics
Methods	Design: 2‐arm, parallel RCT, double‐blind (capsules 2.6 g of EPA and 1.8 g of DHA daily vs placebo), 1 month follow up Summary risk of bias*: unclear risk of bias
Participants	Country: USA Setting: vascular surgery clinic Number of centres: 1 Number of participants: 40 intervention, 40 control Age (mean ± SD): omega‐3 group 68 ± 7; Placebo group 69 ± 9 Sex: male 78; female 2 Inclusion criteria: history of IC; ABI at rest or exercise < 0.9; age 50 and more Exclusion criteria: critical limb ischaemia; hypersensitivity/allergies to fish or seafood; already on omega‐3 fatty acids or equivalent; significant renal, hepatic, and inflammatory disease; concurrent severe infections; acute illness (MI, stroke, major surgery within 30 days); receiving immunosuppressive medications or steroids
Interventions	Type: supplement (capsule) Treatment: 4.4 g daily of n‐3 PUFA (2.6 g of EPA and 1.8 g of DHA daily); capsules marketed as Pro‐Omega; Nordic Naturals, Watsonville, CA Control: placebo capsules containing an inactive substance (soybean; Nordic Naturals) Duration: 1 month
Outcomes	ABI Cholesterol (total, HDL, LDL) Triglycerides SBP and DBP Adverse effects
Study funding sources	This work was funded by University of California San Francisco and the Northern California Institute for Research and Education. The project was supported by an award from the National Center for Research Resources and an award from the National Institute of Health/NHLBI, and a Society for Vascular Surgery Seed Grant and Career Development Award. The publication was also supported by the National Center for Advancing Translational Sciences, National Institutes of Health
Declarations of interest	None
Contact	Authors' names: Grenon SM, Owens CD, Nosova EV, Hughes‐Fulford M, Alley HF, Chong K, Perez S, Yen PK, Boscardin J, Hellmann J, Spite M Institution: Department of Surgery, University of California Address: Department of Surgery, University of California, San Francisco Surgical Services, Veterans Affairs Medical Center, Clement St, San Francisco, USA
Notes	We judged this study to have an unclear summary ROB due to unclear detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Individuals were randomized by a block randomization process, with the study pharmacist maintaining the key"
Allocation concealment (selection bias)	Low risk	Quote: "Prior to the randomization visit, project staff will contact the assigned research pharmacist who will generate the block randomization"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "The placebo capsules are designed to be the same colour and shape as the treatment capsule." Study described as 'double‐blinded'
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Double‐blinded design but it is unclear exactly who was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	The dropout was low (10%) and balanced across groups with reasons reported. It is unclear if the analysis is based on all participants or excludes dropouts.
Selective reporting (reporting bias)	Unclear risk	Some of the outcomes that were not mentioned in the registered protocol were added to the published article, such as improvement in lipid profile (L‐LD, triglycerides, H‐DL), BP, patient‐perceived walking impairment
Other bias	Low risk	No other bias identified

OMEGA‐PAD II 2019.

Study characteristics
Methods	Design: 2‐arm, parallel RCT, double‐blind (capsules 2.6 g of EPA and 1.8 g of DHA daily vs placebo), 3 months follow up Summary risk of bias*: unclear risk of bias
Participants	Country: USA Setting: outpatient vascular surgery clinic Number of centres: 1 Number of participants: 11 intervention, 13 control Age (mean ± SD): omega‐3 group 69 ± 8; placebo group 73 ± 7 Sex: male 21; female 0 Inclusion criteria: history of IC; age 50 and more; one of the following items (ABI at rest or exercise < 0·9 or toe pressure < 70 mm Hg or documentation on imaging of greater than or equal to 50% stenosis in segments of aortoiliac arteries, femoral arteries, or tibial arteries) Exclusion criteria: critical limb ischaemia; hypersensitivity/allergies to fish or seafood; already on omega‐3 fatty acids or equivalent; significant renal, hepatic, and inflammatory disease; concurrent severe infections; acute illness (MI, stroke, major surgery within 30 days); receiving immunosuppressive medications or steroids
Interventions	Type: supplement (capsule) Treatment: 4.4 g daily of n‐3 PUFA (2.6 g of EPA and 1.8 g of DHA daily); capsules marketed as Pro‐Omega; Nordic Naturals, Watsonville, CA Control: placebo was capsules containing an inactive substance (soybean; Nordic Naturals) Duration: 3 months
Outcomes	PFWD ABI Cholesterol (total, HDL, LDL) Triglycerides SBP and DBP Non‐fatal‐vascular events Adverse effects
Study funding sources	This project was supported by multiple awards and funds provided by the following institutions: the National Center for Advancing Translational Sciences, National Institutes of Health (NIH), United States, through University of California, San Francisco Clinical and Translational Science Institute; the Society for Vascular Surgery Student Research Fellowship Award and the American Heart Association Student Scholarship (J.L.R.); the University of California, San Francisco and the Northern California Institute for Research and Education; the National Center for Research Resources; the National Heart, Lung, and Blood Institute (NHLBI), United States, and a Society for Vascular Surgery Seed Grant and Career Development Award (SMG); National Research Service Award from the NIH/ NHLBI (F32‐HL136044; B.E.S.); NIH‐NIGMS P01‐GM095467 (CoreB; M.S.)
Declarations of interest	The authors have no conflicts of interest in reporting
Contact	Authors' names: Ramirez JL, Gasper WJ, Khetani SA, Zahner GJ, Hills NK, Mitchell PT, Sansbury BE, Conte MS, Spite M, Grenon SM Institution: Division of Vascular and Endovascular Surgery, Department of Surgery, University of California Address: Division of Vascular and Endovascular Surgery, Department of Surgery, University of California, San Francisco 400 Parnassus Avenue, A‐581, San Francisco, Canada
Notes	We judged this study to have an unclear summary ROB due to unclear detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Participants enrolled in the trial were randomized to one of two groups: fish oil or placebo. Patients were randomized by a block randomization with four subjects per block with a ratio of1:1 for each block. "
Allocation concealment (selection bias)	Low risk	Quote: "The randomization was done by research pharmacist who maintained the key until the end of the study."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"a double‐blinded" design. "The placebo group took the same number of capsules containing soybean (Nordic Naturals)that was designed to appear the same as the treatment capsules."
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Doule‐blind but not clear who was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "All analyses were based on intention to treat." Dropout 17% (4/24) balanced across groups and reasons reported
Selective reporting (reporting bias)	Unclear risk	NCT01979874‐ Study started in February 2014 and completed in September 2017. Some of the outcomes that were registered in the protocol were not published in the article or added new outcomes that were not registered
Other bias	High risk	Recruitment for the trial was slower than expected and was ended by the principal investigator before reaching target enrolment. Trials targeted 60 participants and stopped the trial with 24. Very low power.

ORIGIN 2019.

Study characteristics
Methods	Design: multicenter, 2 x 2 factorial design, parallel RCT, double‐blind (capsules containing EPA 465 mg and DHA acid 375 mg daily vs placebo), 6.2 years follow up Summary risk of bias*: unclear risk of bias
Participants	Country: Canada and 39 other countries Setting: hospitals and clinics Number of centres: multicenter, multinational, 578 clinical sites in 40 countries Number of participants: 485 intervention, 486 control Age (mean ± SD): omega‐3 group 66.2 ± 8; placebo group 68.8 ± 8.1 Sex: male 643; female 328 Inclusion criteria: age 50 years and more; a diagnosis of diabetes; 50% or more stenosis of a coronary, carotid, or lower‐limb artery on angiography;or an ABI of less than 0.9; a negative pregnancy test for all women Exclusion criteria: unwilling to discontinue use of a non‐study preparation of omega‐3; glycated haemoglobin level of 9% or more; had undergone coronary‐artery bypass grafting within the previous 4 years with no intervening cardiovascular event; severe heart failure; cancer that might affect survival
Interventions	Type: supplement (capsule) Treatment: 1 g daily of n‐3 PUFA (containing EPA 465 mg and DHA 375 mg); capsules marketed as Omacor 1 g (Pronova, BioPharma AS, Lysaker, Norway) Control: placebo was capsules containing approximately 1 g of olive oil (Pronova, BioPharma AS, Norway Duration: 6.2 years (compliance 83%)
Outcomes	ABI Cholesterol (total, HDL, LDL) SBP and DBP
Study funding sources	This project was funded by Sanofi Aventis and the omega‐3 supplement, Omacor, was provided by Pronova Biocare AS
Declarations of interest	Four of the authors have no conflicts of interest to declare. One author received a research grant from Sanofi Aventis
Contact	Authors' names: Dagenais GR, Jung H, Bogaty P, Bosch J, Yusuf S, Gerstein HC Institution: University Institute of Cardiology and Pneumology of Quebec Address: Laval University, 2725, Chemin Ste‐Foy, Quebec City, Quebec, Canada
Notes	12,537 participants with either diabetes or prediabetes were randomised to the addition of a daily injection of insulin glargine vs standard care, and to 1 g of omega‐3 supplements vs placebo. We used the results that focused on PAD and are limited to the 971 participants (8%) who had an ABI of 0.9 or less We judged this study to have an unclear summary ROB due to unclear selection and detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not reported
Allocation concealment (selection bias)	Low risk	Participants were randomised by an automated telephone randomisation system
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Trial described as "double blind" and placebo described as identical. Blinding of participants, investigators, adjudication committee members, local and central trials personnel described.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote:" To compare the time‐to‐first‐event curves, we used an intention‐to‐treat approach for all efficacy analyses." 6281/6319 was included in analyses for the omega‐3 group and 6255/6292 was included in analyses for the placebo group. Data from 75 persons at 3 sites (38 omega‐3 vs 37 placebo) were excluded while the trial was ongoing at the request of national regulatory agencies, after their audits of data from those sites.
Selective reporting (reporting bias)	Low risk	NCT00069784 – registered October 2003, trial started August 2003, final data collection December 2011. Outcomes appear to have been reported in various publications
Other bias	Low risk	No other bias identified

Schiano 2008.

Study characteristics
Methods	Design: single centre, 2‐arm, RCT, single‐blind (capsules containing 1 g EPA and DHA vs pre‐enrolment therapy), 3 months follow up Summary of risk of bias*: unclear risk of bias
Participants	Country: Italy Setting: vascular laboratory Number of centres: 1 Number of participants: 32 (16 in each group) Age: (median) 66 years for both groups Sex: 29 male, 3 female Inclusion criteria: participants referred to a vascular laboratory with a history of IC and resting ABI < 0.9 plus presence of ≥ 1 stenoses > 50% in ≥ 1 artery on duplex scanning Exclusion criteria: rest pain; trophic lesions in affected limb; acute coronary syndrome or CV event in previous 6 months; significant renal, hepatic or inflammatory disease; history of hypersensitivity; other contraindications to n‐3 PUFAs
Interventions	Treatment: n‐3 PUFAs 1 g bd* + "usual" (pre‐enrolment) therapy Control: placebo "usual" pre‐enrolment therapy Duration: 3 months *capsules containing at least 85% EPA and DHA as ethyl esters in mean ratio of EPA/DHA of 0.9:1.5
Outcomes	ABI Cholesterol (total, HDL, LDL) Triglycerides Adverse effects
Study funding sources	No details provided
Declarations of interest	The authors reported no conflicts of interest
Contact	Authors' names:Schiano V, Laurenzano E, Brevetti G, De Maio JI, Lanero S, Scopacasa F, Chiariello M Institution:Department of Clinical Medicine and Cardiovascular and Immunological Sciences, University of Naples ‘‘Federico II’’ Address:University of Naples ‘‘Federico II’’, Via Pirro Ligorio, 10, 80129 Napoli, Italy
Notes	We judged this study to have an unclear summary ROB due to unclear selection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No description of the randomisation process
Allocation concealment (selection bias)	Unclear risk	No description of allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Personnel assessing the outcome is stated to be "blinded". Participants were not blinded but outcomes were unlikely to be affected by this
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Personnel assessing the outcome is stated to be "blinded"
Incomplete outcome data (attrition bias) All outcomes	Low risk	32 consecutive participants referred to a vascular laboratory were randomised. There were no exclusions post randomisation or withdrawals during follow‐up
Selective reporting (reporting bias)	Low risk	All prespecified outcomes are reported in the prespecified manner
Other bias	Low risk	No other obvious sources of bias

Stricker 2008.

Study characteristics
Methods	Design: single centre, 2‐arm, parallel RCT, double‐blind (canola oil containing 2.24 g of alpha‐linolenic acid daily vs placebo), 8 weeks follow up Summary risk of bias*: unclear risk of bias
Participants	Country: Switzerland Setting: hospital Number of centres: 1 Number of participants: 40 (20 in each group) Age: 50+ years (mean 66.8 intervention; 63.7 control) Sex: 27 male, 13 female (canola oil group: 14 male, 6 female) (sunflower oil group: 13 male, 7 female) Inclusion criteria: chronic PAD defined as ABI < 0.9 plus > 50% stenosis or occlusion in a leg artery (on duplex or angiogram) Exclusion criteria: acute, intermittent illness; thromboangiitis obliterans; renal insufficiency (creatinine > 130 µmol/L); acute stroke or MI in previous 2 months; current oral anticoagulation medicine; liver cirrhosis; presence of a malignant tumour
Interventions	Treatment: 2 tablespoons (35 mL) a day of canola oil (= 2.24 g of alpha‐linolenic acid, C18 3n‐3 omega‐3 fatty PUFA) Control: placebo was 2 tablespoons (35 mL) a day of sunflower oil (= 16.24 g of linolenic acid, C18 2n‐6 omega‐6 PUFA) Duration: 8 weeks
Outcomes	Cholesterol (total, HDL, LDL) Triglycerides
Study funding sources	The study was partly funded by the Fondo Balli, Locarno, Switzerland
Declarations of interest	No details provided
Contact	Authors' names: Stricker H, Duchini F, Facchini M, Mombelli G Institution: Department of Medicine, Switzerland Address: Department of Medicine, Via all’Ospedale, CH‐6600 Locarno, Switzerland
Notes	We judged this study to have an unclear summary ROB due to unclear selection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised in blocks of 4 using a randomisation website (www.randomization.com)
Allocation concealment (selection bias)	Unclear risk	No description of allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Patients and doctors blinded". The bottles containing the treatment and placebo were packaged similarly and rendered visually indistinguishable
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Any lack of blinding of assessors is unlikely to have affected the outcomes
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	40 participants were randomised. 40 were analysed, so it is implied that there were no exclusions or withdrawals. However, there is no explicit statement to confirm this
Selective reporting (reporting bias)	Low risk	All prespecified outcomes are reported in the prespecified way
Other bias	Low risk	No other obvious source of bias

Woodcock 1984.

Study characteristics
Methods	Design: single centre, 2‐arm, parallel RCT, double‐blind (capsules containing 1.8 g EPA daily vs placebo), 7 weeks follow up Summary risk of bias*: unclear risk of bias
Participants	Country: UK Setting: vascular clinic Number of centres: 1 Number of participants: 19 (10 in intervention; 9 in control) Age: 56 to 75 years Sex: 15 male; 4 female. Fish oil group: 7 males; 3 females; corn/olive oil group: 8 males; 1 female Inclusion criteria: IC. Presence and site of disease confirmed by Doppler USS. Claudication distance 14 ‐1600 m, ABI 0.34 to 0.78. All participants attended a vascular clinic and were ineligible for surgery Exclusion criteria: none described
Interventions	Treatment: 5 capsules of fish oil bd (= 1.8 g EPA per day) Control: placebo was 5 (identical) capsules of corn oil (linoleic acid) and olive oil (oleic acid) bd Duration: 7 weeks
Outcomes	PFWD ABI Cholesterol Triglycerides
Study funding sources	Supported by research grants from the British Heart Foundation and the special trustees for the Former United Sheffield Hospitals.
Declarations of interest	No details provided
Contact	Authors' names: Woodcock BE, Smith E, Lambert WH, Jones WM, Galloway JH, Greaves M, Preston FE Institution: University Department of Haematology and Department of Surgery, Royal Hallamshire Hospital, Sheffield Address: University Department of Haematology, Royal Hallamshire Hospital, Sheffield
Notes	No participants with DM We judged this study to have an unclear summary ROB due to unclear selection and detection bias
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A table of random numbers was used
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not stated
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Treatment and control are both provided in identical capsules
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear whether the outcome assessors were blinded or not. This is unlikely to have affected the majority of outcomes which were subjective, but may have had a potential effect on the more subjective outcomes such as PFWD
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	The number of participants randomised is not stated. No description of withdrawals/losses to follow‐up
Selective reporting (reporting bias)	High risk	All prespecified outcomes are mentioned in the results section; however, for the majority of outcomes no data were provided to back up statements made in the text
Other bias	Low risk	No other obvious sources of bias

*The overall summary of risk of bias for each study was defined as:

low summary risk of bias for all key domains (at low risk of selection bias, performance bias and detection bias), unclear summary risk of bias for one or more key domains (at unclear risk of selection bias, performance bias and detection bias), high summary risk of bias for one or more key domains (at high risk of selection bias, performance bias and detection bias)

ABI: ankle‐brachial index;
bd: twice daily;
BMI: body mass index;
BP: blood pressure;
CV: cardiovascular;
DBP: diastolic blood pressure;
DHA: docosahexaenoic acid;
DM: diabetes mellitus;
EPA: eicosapentaenoic acid;
GLA: gamma‐linolenic acid;
HbA1c: haemoglobin A1c;
HDL: high‐density lipoprotein;
IC: intermittent claudication
IDDM: insulin‐dependent diabetes mellitus;
LDL: low‐density lipoprotein;
m: metres;
MI: myocardial infarction;
MWD: maximal walking distance;
NIDDM: non‐insulin‐dependent diabetes mellitus;
PAD: peripheral arterial disease;
PFWD: pain‐free walking distance;
PUFA: polyunsaturated fatty acid;
QoL: quality of life;
ROB: risk of bias;
SD: standard deviation;
SE: standard error;
USS: ultrasound scan.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Berrettini 1996	Only 5 participants were in the intervention group and 2 in the control group with PAD. Data was requested but not provided for this subset. Therefore, there is no means of identifying which participants had PAD
Caligiuri 2014	Eligible participants were diagnosed as hypertensive and unclear whether participants had IC. No relevant outcomes measured
Euctr2013‐004342‐42	Study terminated due to low recruitment rate
Ishikawa 2010	Unclear whether participants had IC
Madden 2007	Non‐ RCT (No appropriate control group)
Moller 1998	Non‐ RCT (Unclear whether participants were randomly assigned to groups)
Olsson 1984	Non‐ RCT (Deemed after translation to be a CCT and not a RCT)
Ramirez‐Tortosa 1999	Non‐ RCT (Non‐randomised, non‐blinded cross‐over study)
Ruiz‐Canela 2014	Some participants did not have PAD or IC before randomisation

CCT: controlled clinical trial 
IC: intermittent claudication
PAD: peripheral arterial disease
RCT: randomised controlled trial

Characteristics of ongoing studies [ordered by study ID]

Canola‐PAD 2010.

Study name	Effects of canola oil on blood vessel function in peripheral arterial disease
Methods	RCT, parallel assignment, double‐blinded
Participants	51 patients 40 years and older with PAD including those with claudication (ABI < 0.9) or asymptomatic carotid stenosis >50% or having had a previous intervention for PAD
Interventions	Intervention group: participants consumed food items containing conventional canola oil (25 g/day), for 8 weeks Control group: usual diet with oil mixture representing the typical Western diet (25 g/day), for 8 weeks
Outcomes	Primary: vascular function (ABI, arterial stiffness, and endothelial function) Secondary: physiological parameters (body weight, blood pressure), serum biochemistry (lipids and inflammatory biomarkers), and fatty acid composition
Starting date	Registered on trials registry: November 2010 Trial start date: September 2011 Trial completion date: February 2013
Contact information	Dr Carla G Taylor, Department of Human Nutritional Sciences, University of Manitoba, Winnipeg, Manitoba, Canada ctaylor@sbrc.ca
Notes	Dr Carla G Taylor contacted them in 2021: they decided to publish their data. Checked for updates and publications, none found, 1 September 2021

NCT01256320.

Study name	The impact of egg consumption on indices of vascular health in individuals with peripheral arterial disease
Methods	RCT, parallel assignment, double‐blinded
Participants	60 participants 40 years and older with PAD, including those with claudication as defined by an ABI of ≤ 0.90 or asymptomatic carotid stenosis of > 50%; or who have had a previous intervention for PAD
Interventions	Intervention group 1: classic egg group, consumption of 6 eggs/week (1 egg/day for 6 days with 1 day rest) of commercial classic eggs Intervention group 2: omega‐3 egg group, consumption of 6 eggs/week (1 egg/day for 6 days with 1 day rest) of commercial omega‐3 eggs Control group: no intervention, no shell egg consumption and usual dietary practices
Outcomes	Primary: establish quantitative and qualitative data on the dietary patterns, including egg consumption, in patients with PAD. The assessments included anthropometrics (weight, height; waist circumference, body mass index, dietary assessment (3‐day dietary food record, food frequency questionnaire, biochemical analyses, expression and metabolic profiling, ABI, pulse wave velocity, arterial vasodilation, and advanced glycation end products Secondary: determine the impact of consuming 6 eggs per week on primary end‐points and other modifiable risk factors in patients with PAD. The assessments included anthropometrics (weight, height; waist circumference, body mass index, dietary assessment (3‐day dietary food record, food frequency questionnaire, biochemical analyses, expression and metabolic profiling, ABI, pulse wave velocity, arterial vasodilation, and advanced glycation end products
Starting date	Registered on the trial registry: 18 June 2010 Trial start date: June 2010 Trial completion date: February 2013
Contact information	Dr. James House, University of Manitoba, IH Asper Clinical Research Institute, St. Boniface General Hospital Winnipeg, Manitoba, Canada, R2H 2A6. rguzman@sbgh.mb.ca
Notes	Dr. James House was contacted on 2 September 2021. There was no response. Checked for updates and publications, none found, 1 September 2021

NCT02096757.

Study name	The use of fish oil to reduce inflammation caused by a peripheral vascular intervention
Methods	RCT, parallel assignment, triple ‐blinded (participant, care provider, investigator)
Participants	30 participants 40 years to 90 years with PAD including disabling claudication or CLI as an indication for PVI (Rutherford Class II‐V) and objective evidence of PAD, e.g. ABI < 0.9, TBI < 0.6 or absent pedal pulses
Interventions	Intervention group: pro‐omega (fish oil) (Nordic Naturals, Watsonville, CA, USA). Each ProOmega capsule contains 325mg of EPA and 225mg of DHA. 4.4gm/day 2 weeks prior to surgery and continued for 4 weeks after. Control group: pro‐omega placebo soybean capsules; 4.4gm/day (Nordic Naturals, Watsonville, CA, USA) 2 weeks prior to surgery and continued for 4 weeks after
Outcomes	Primary: systemic inflammatory bio‐markers Secondary: safety, primary patency ( as demonstrated by ultrasound)
Starting date	Registered on the trial registry: 26 March 2014 Trial start date: June 2014 Trial completion date: March 2017
Contact information	Warren Gasper, M. D UCSF & SFVAMC University of California, San Francisco, United States, 94121
Notes	Warren Gasper contacted in 2021: no response Checked for updates and publications, none found, 1 September 2021

NCT02152930.

Study name	The effects of fish oil supplements during supervised exercise therapy in patients with Intermittent claudication
Methods	RCT, parallel, no masking (open label)
Participants	8 patients 18 years and older with IC
Interventions	Intervention group: omega‐3 fatty acids (fish oil supplements)1000 mg of omega‐3 fatty acid ethylesters, 2 times daily, for 4 weeks, followed by 12 weeks of supervised exercise therapy in combination with 2 times daily omega‐3 fatty acid supplementation Control group: no intervention
Outcomes	Primary: maximal walking distance with standardised treadmill test Secondary: haemorrhagic parameters, visceral fat mass, ex vivo cytokine production after stimulation with lipopolysaccharide, microcirculation with side stream dark field technology, blood pressure, heart rate, cholesterol, triglycerides, haemoglobin, leucocytes, C‐reactive protein, ABI
Starting date	Registered on trials registry: June 2014 Trial start date: December 2014 Trial completion date: February 2016
Contact information	Dr Alexander P.J. Houdijk, Medical Center Alkmaar, Alkmaar, Noord Holland, Netherlands, 1800AM, a.p.j.houdijk@mca.nl
Notes	Alexander P.J. Houdijk contacted in 2021: no response Checked for updates and publications, none found, 1 September 2021

UMIN000026607.

Study name	The efficacy of EPA on the clinical outcome in patients with arteriosclerosis obliterans undergoing endovascular treatment: prospective study
Methods	RCT, parallel
Participants	300 participants, 20 to 90 years with arteriosclerosis obliterans
Interventions	Intervention group: omega‐3 supplementation (EPA 1800mg/day), duration: 2 years Control group: No treatment
Outcomes	Primary: major lower limbs events (revascularisation, amputation) Secondary: Fontaine classification, Rutherford classification, 6MW, ABI, TBI, skin perfusion pressure, major cardiovascular events
Starting date	Registered on trials registry: 18 March 2017 Trial start date: 22 December 2014 Trial completion date: 31 December 2019
Contact information	Ashikaga Takashi E‐mail: sysh.cvm@tmd.ac.jp Address: Department of Cardiovascular medicine, Tokyo Medical and Dental University, 1‐5‐45 Yushima, Bunkyo ward, Tokyo, Japan
Notes	Ashikaga Takashi contacted in August 2021: no response Checked for updates and publications, none found, 1 September 2021

6MW: 6‐minute walking test
ABI: ankle‐brachial index
CLI: critical limb ischaemia
IC: intermittent claudication
PAD: peripheral arterial disease
PVI: peripheral vascular ischaemia
RCT: randomised controlled trial
TBI: toe brachial pressure index

Differences between protocol and review

2024

For this update, we removed some outcomes presented in the 2013 version, because we decided these outcomes have less clinical relevance to intermittent claudication than other outcomes (Campbell 2013). The removed outcomes were angiographic findings, and blood and plasma viscosity. Progression to amputation was redefined as amputation rate/frequency, and two other outcomes (non‐fatal myocardial infarction (MI) and non‐fatal cerebrovascular accident (CVA)) were redefined as non‐fatal vascular events. We changed the order of outcomes in the Methods. and created a new summary of finding table to reflect current clinical importance. We also clarified the types of studies, interventions, and comparisons considered for inclusion.

In the protocol, we stated that we planned to subgroup outcomes according to age, sex, severity of disease, and comorbidities, such as diabetes, but due to insufficient data, this was not done. The same issue happened with sensitivity analysis and funnel plot.

Contributions of authors

MM: identification of trials for inclusion; evaluation of methodological quality of included trials; contact with trial authors; meta‐analysis; interpretation of data; and writing of the draft review
TB: study selection, risk of bias assessment, interpreting the data, drafting the manuscript
MR: identification of trials for inclusion; and writing the review
ES: data extraction
JL: assessed risk of bias; and data extraction

Sources of support

Internal sources

• Community Health Research Center, Iran

  The review was registered and funded by the community health research center, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran

External sources

• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK

  Cochrane Vascular editorial base is supported by the Chief Scientist Office.

Declarations of interest

MM: none known
TB: none known
MR: none known
ES: none known. ES has declared that they work as a health professional (Dental clinic)
JL: none known

New search for studies and content updated (no change to conclusions)