Healthy grocery delivery in the usual care for adults recovering from an acute coronary event: protocol for a three-arm randomised controlled trial

Abstract

Introduction

Coronary heart disease is a major contributor to the global burden of disease. Appropriate nutrition is a cornerstone of the prevention and treatment of coronary heart disease; however, barriers including cost and access to recommended foods limits long-term adherence for many. We are conducting, in adults with coronary heart disease, a randomised controlled trial comparing usual care with two dietary interventions in which usual care is augmented by 12 weeks free delivered groceries.

Methods and analysis

Three hundred adults recovering from an acute coronary event will be recruited from outpatient cardiovascular services in three regions of Aotearoa New Zealand. Participants will be randomly allocated to three arms: usual care (control group), usual care and the free delivery of foods high in dietary fibre or usual care and the free delivery of foods high in unsaturated fats. Interventions duration is 12 weeks, with a further 12 months follow-up. The primary outcome measures are change in low-density lipoprotein (LDL) cholesterol concentration following the intervention, and a cost-effectiveness analysis of healthcare access and social costs in the year after the intervention. A broad range of secondary outcome measures include other blood lipids, anthropometry, glycaemia, inflammatory markers, gut microbiome, dietary biomarkers, food acceptability, dietary change and the facilitators and barriers to dietary change. The trial will determine whether the free provision of groceries known to reduce cardiovascular risk within usual care will be clinically beneficial and justify the cost of doing so. Results may also provide an indication of the relative benefit of foods rich in dietary fibre or unsaturated fats in coronary heart disease management.

Ethics and dissemination

This trial, The Healthy Heart Study, has Health and Disability Ethics Committee approval (20/NTB/121), underwent Māori consultation, and has locality authority to be conducted in Canterbury, Otago and Southland.

Trial registration number

ACTRN12620000689976, U1111-1250-1499.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This is a practical consideration of usual care compared with usual care augmented by free healthy groceries to comment on the possible benefits as well as the feasibility of free grocery delivery within usual care models.

• Our participants represent a globally relevant patient population. Evidence of benefit due to free healthy groceries in this trial may enable further consideration in other patient populations, such as those with type 2 diabetes.

• To fully consider the effects of free healthy grocery delivery, we are collecting information to enable comment on a wide range of secondary outcomes.

• We consider 12-week free grocery delivery the minimum duration to observe a change in cardiometabolic risk factors between interventions; however, it is not yet known if this is of sufficient duration to elicit habitual change in dietary intakes beyond the intervention period.

Introduction

Non-communicable diseases (NCDs), such as coronary heart disease (CHD), are the leading cause of global mortality, accounting for 71% (41 million) of deaths each year.^{1 2} These values are expected to increase further,³ presenting a clear need for new pathways in the prevention and management of CHD. Inappropriate nutrition and dietary choices are principal causes of NCDs, contributing to CHD, obesity, stroke, cancer and all-cause death.^{4 5} Findings from the Global Burden of Disease study indicate that dietary factors are the most important determinants of global morbidity and mortality.² Insufficient intakes of whole grains, fruits, nuts and seeds, vegetables, seafood, dietary fibre, polyunsaturated fats, legumes and excessive intakes of sodium, trans fats, and red and processed meats explain the dietary contribution to the disease burden.⁴ However, changing habitual dietary intakes towards healthier choices can be hard for many.⁶ Many factors within the microenvironment and macroenvironment influence dietary intakes. These include social, cultural and personal preferences as well as income, food affordability, food availability, and food and health knowledge.^{7 8} Some of the major barriers identified to improving dietary intakes are cost and access⁹ with healthy foods often considered more expensive.¹⁰ Furthermore, interventions that require high personal agency, such as those focused solely on nutrition education, are less likely to prove beneficial than interventions requiring low agency,¹¹ such as changes to the food environment around the individual.

Given the potential for dietary interventions to improve outcomes for those with established CHD,^{12 13} and the need to provide equitable and achievable interventions, we are assessing free healthy grocery delivery within usual care, which includes cardioprotective drug use, on immediate and long-term outcomes in those recovering from an acute coronary event. This trial (The Healthy Heart Study) examines whether including the free delivery of key foods as a component of usual care has the potential to achieve risk factor reduction over the duration of the intervention, and is cost-effective when considering a range of longer- term outcome measures. In addition, this trial is designed to compare the respective merits of two different dietary approaches, one emphasising foods high in fibre, the other foods high in mono and polyunsaturated fats.

Objectives

• To evaluate the effect of free healthy grocery delivery within usual care on cardiometabolic risk factors and more exploratory risk factors in those recovering from an acute coronary event.

• To evaluate the cost-effectiveness of 12-week free healthy grocery delivery within usual care on health and social outcomes at 1 year after intervention completion.

• To compare two different dietary approaches in terms of health benefit, uptake into habitual diets and economic feasibility in the usual care of CHD.

• To evaluate the effect of free healthy grocery delivery within usual care on immediate and habitual dietary intakes.

• To identify the barriers and enablers to dietary change when foods are provided within usual care for 12 weeks.

Methods

Study design

This is a randomised, three-arm parallel trial of interventions lasting 12 weeks with follow-up for 1 year following intervention cessation. This trial is being conducted in three regions of Aotearoa New Zealand (NZ) (Canterbury, Otago and Southland) with a catchment area of 830 000 people. Recruitment for this study commenced in February 2021. Complete data capture of the 12-week intervention is expected in December 2023, with complete 1-year follow-up data expected in December 2024. Figure 1 presents the flow of participants through the study. Standard Protocol Items: Recommendations for Interventional Trials reporting guidelines have been followed.¹⁴

Figure 1.

Flow of participants through the healthy heart study.

Setting

This is a free-living study of adults recovering from an acute coronary event. The trial is being conducted without direct contact between study staff and participants. The only required in-person contact for participants is for phlebotomy and anthropometric assessment at two time points, baseline and 12 weeks, which was undertaken by trained health professionals. All other assessments were undertaken remotely/online.

Participants and recruitment

Participants are recruited through outpatient cardiovascular services following survival from an acute coronary event. Study staff, collaborating health professionals and study materials such as leaflets and posters all refer potential participants to a website containing study information and contact details. Once provided with trial information and the contact details of study staff, potential participants then have the time to consider participation before they themselves express interest in participating through the website. Consent to participate in this study is recorded online. Participants also chose whether their data may be used for unspecified future research questions or not. Consent for any aspect of this or future studies may be withdrawn by the participant at any time, without disadvantage. Participant information sheet and consent documents are provided in online supplemental files 1 and 2.

Inclusion criteria

The trial includes adults aged 18–80 years who are capable of giving informed consent and provide consent to participate between 6 weeks and 6 months from their most recent acute coronary event. Eligible events are: ST elevated myocardial infarction, non-ST elevated myocardial infarction, coronary artery bypass grafting or a percutaneous coronary intervention. Event eligibility was confirmed through medical records. Multiple events or further comorbidities do not exclude participation. Participants must be able to complete the online surveys and attend a participating phlebotomy service.

Exclusion criteria

Stand-alone non-eligible coronary events included: aortic valve stenosis, valve replacement, angina, spontaneous coronary artery dissection and heart failure. Potential participants with allergies or special dietary needs which may prevent them from being able to consume foods provided in the trial are asked to consider if this study is for them. Potential participants who cannot provide a postal address that study foods can be delivered to in a timely manner will be excluded from participation due to food safety concerns.

Sample size

We used data from a NZ cohort of adults with acute coronary syndrome¹⁵ to inform an estimate of the sample size with an alpha of 0.05 and power of 0.80 to detect a 0.2 mmol/L greater improvement in low-density lipoprotein (LDL) cholesterol in one trial arm over either other arm. We require 80 participants to complete each intervention. We intend to over-recruit to allow for drop out and better consider the secondary outcomes by enrolling 100 participants per intervention. Given the three intervention arms, we are recruiting a total of 300 participants. The sample size was not adjusted for multiple outcomes.

Randomisation

A computer-generated balanced randomisation schedule determines which of the three trial arms consenting participants are placed into. The randomisation list was developed by a biostatistician using unstratified block randomisation with a block size of 15. This randomisation list was uploaded into a secure web application (REDCap). Allocation of participants to trial arms is undertaken once their consent to participate is obtained. Allocation is performed within REDCap without researchers being able to access the randomisation list. This ensures researchers have no influence over the allocation process.

Interventions

All participants receive usual care in the recovery from an acute coronary event, which includes evidence-based lifestyle and dietary advice, as well as pharmacological interventions relevant to their presentation and underlying risk factors. Typical pharmacological management would include antiplatelet therapy, statins, ACE inhibitor/angiotensin receptor blocker and in those with reduced ventricular function, beta-blockers. Usual care is provided to patients individually by health professionals and cardiovascular outpatient services, with mandatory national audit of secondary prevention medications. The study investigators augment the dietary advice that is provided in usual care with a small recipe and menu book developed by the Heart Foundation of New Zealand (Cheap Eats) mailed to every participant at the start of the intervention.

Control arm

One-third of participants are randomised to the control group who receive a NZ$100 grocery voucher to compensate them for their time after the 12-week intervention period is completed.

Intervention: trial arms receiving free healthy groceries

Two-thirds of participants are randomised to 12-week free delivery of healthy groceries. Participants of these interventions receive weekly grocery deliveries. Participants are asked to incorporate the study-supplied groceries into their meals. No further information on how to prepare or incorporate the foods is provided. Groceries are assembled and dispatched by a collaborating national meal kit delivery service (My Food Bag).

One hundred participants are randomised to receive foods high in dietary fibre. The other hundred participant are randomised to receive foods high in monounsaturated and polyunsaturated fats. These interventions were chosen as a distillation of current messages in the management of CHD: increase fibre intakes and replace saturated fat with unsaturated fats.12 13 16 17 Each weekly delivery for participants is the same 12 foods shown in table 1. These foods were chosen as being either high in fibre or unsaturated fats, or key sources of these nutrients in the NZ diet. For participants in households larger than four individuals, we double the amount of groceries sent each week in-line within a family-centred approach.

Table 1.

Foods, weights, energy and macronutrients provided in the intervention arms

High fibre foods	Weight (g)	Energy (kJ)	SFA (g)	MUFA (g)	PUFA (g)	Protein (g)	Total CHO (g)	Dietary fibre (g)
Baby spinach	321	240	–	–	–	8.0	13.4	7.1
Broccoli	741	1035	–	–	–	28.0	59.2	25.2
Cabbage	1013	1013	–	–	1.7	17.1	80.1	29.4
Carrots	722	1127	–	–	–	4.6	80.8	20.2
Apples	984	2094	–	–	–	2.1	143.4	20.1
Banana	1127	4469	1.1	–	–	11.6	270.5	20.3
White beans	480	1596	–	–	2.4	32.4	82.6	26.9
Chickpeas	240	1064	–	1.7	3.7	17.4	45.4	13.2
Wholegrain wheat bread	1500	14 524	7.2	15.7	3.66	143.9	715.4	76.5
Brown rice	500	7044	3.4	4.2	4.9	40.8	367.3	14.5
Wholegrain oats	750	11 987	9.8	20.0	19.3	107.1	589.2	89.25
Whole wheat flour	500	6948	1.2	–	3.9	60.4	398.2	56
Weekly sum	8.9 kg	53 352	22.7	41.6	51.5	473.4	2845.5	399.3

Foods high unsaturated fats
Tinned salmon	415	2513	4.7	13.3	14.5	60.6	5	–
Tinned tuna in oil	850	9571	27.0	88.4	70.1	169.0	11	–
Walnuts	100	2909	4.9	9.0	50.0	14.5	12	9
Extravirgin olive oil	235	7990	14.4	166.9	21.2	–	–	–
Avocados	708	4910	17.0	84.0	12.9	9.2	26.0	22.1
Peanut butter	235	6135	19.8	74.8	23.1	52.6	41.8	16.5
Sunflower seeds	300	5027	11.5	48.2	55.2	40.9	16.1	15.4
Pumpkin seeds	200	4767	16.4	33.4	45.1	62.3	27.1	18.2
Eggs	460	2456	9.7	17.4	3.2	60.4	1.4	–
Almonds	100	2491	3.8	33.1	15.2	20.1	23.1	11.6
Unsaturated fat Spread	200	4660	29.8	70.8	25.2	25.2	–	–
Kalamata olives	90	884	3.2	15.8	2.5	1.4	6.1	3.0
Weekly sum	3.9 kg	54 313	162.2	655.1	338.2	491.0	169.6	95.8

Values <1 g are not shown. Nutrient values taken from Foodworks (Xyris) unless all relevant values were stated on food packaging.

CHO, cholesterol; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; SFA, saturated fatty acid.

Outcome measures

We collect data from all participants at nine time points: baseline, every 2 weeks during the intervention (total of five times) at intervention completion, 12 weeks after intervention completion and 52 weeks after intervention completion (see figure 2). All participants receive a NZ$25 then NZ$40 grocery voucher when providing data at 12 weeks and 52 weeks postintervention.

Figure 2.

Assessment time points during the Healthy Heart Study.

Primary outcome

The primary outcome from the 12-week intervention is LDL cholesterol concentration. This outcome is measured at baseline before the intervention, and in the final week of the 12-week intervention. Fasted venous blood are drawn at any of eleven collaborating phlebotomy services (Awanui Labs) within the regions this trial is being conducted. Participant fasting status is queried by the phlebotomist at time of blood draw, recorded and passed on to study investigators. LDL cholesterol concentration is being assessed following standard protocols by the phlebotomy service. Participants are not advised of LDL concentration, or any other measurement recorded, until after their 12-week intervention is finished.

The primary outcome at 1 year after intervention completion is a cost-effectiveness analyses free healthy grocery delivery in those recovering from an acute coronary event. Cost-effectiveness will be considered from both the societal (including individual quality of life^{18 19}) and healthcare provider perspectives,²⁰ following standard reporting protocols.²¹ The healthcare provider perspective will be used as the most likely entity to implement grocery provision into usual care, and benefit from any savings in healthcare usage. Total cost (intervention delivery plus healthcare costs) and effectiveness (quality-adjusted life-years) outcomes will be calculated with data captured at follow-up. Incremental cost-effectiveness ratios will be calculated for both interventions and cost-effectiveness ratios below a willingness-to-pay (WTP) threshold equivalent to NZ gross domestic product (GDP)/capita (NZ$70 685 in 2021) used to quantify cost-effectiveness.²² Bootstrapping (1000 scenarios) will be used to determine the probability that each intervention was cost-effective at varying WTP thresholds.

Should either of the grocery interventions (high fibre or high in healthy-fats) attenuate the primary immediate outcome of the trial (LDL cholesterol), we will estimate projected cardiovascular event likelihood and reduction due to the interventions to inform analysis of longer-term cost-effectiveness. These should be a highly informative given that LDL cholesterol concentration meets all criteria for causality of atherosclerosis and cardiovascular disease.^{23 24} We will apply an annual discount rate of 3.5% to any analyses of longer-term outcomes to determine the net present values of costs and effects, with sensitivity analysis from 0% to 5%.²⁵

Secondary outcomes

Other blood lipid parameters

Blood lipid parameters (total cholesterol, high-density lipoprotein (HDL), total triglycerides, HDL/total cholesterol ratio, ApoB and lipid particle size) are being measured from fasted venous samples collected at baseline before the intervention, and in the final week of the 12-week intervention. The clinically accepted lipid parameters are assessed following standard protocols by the phlebotomy service, while ApoB, and lipid particle size will be measured from frozen (−80°C) samples after data collection is complete.

Anthropometry

Anthropometry measures body weight (kg), body mass index (BMI) and body fat mass (%) measured by bioelectrical impedance will be assessed at baseline before the intervention, and in the final week of the 12-week intervention. Assessments will be performed on fasted participants by trained staff at the collaborating phlebotomy services using BC-587 Tanita Scales. Height (cm) is self-reported by participants at baseline, with the same value used at both time points to calculate BMI.

Blood glucose control

Glycated haemoglobin is measured at baseline before the intervention, and in the final week of the 12-week intervention. It is assessed following standard protocols by the phlebotomy service from a fasted venous blood sample.

Markers of inflammation

Markers of inflammation are assessed in fasted venous blood samples taken at baseline before the intervention, and in the final week of the 12-week intervention. Aliquots of plasma from all participants are frozen at −80°C and will be batch-processed when all participants have completed the intervention. Samples are then assessed with a Bio-Plex Pro Human Inflammation magnetic bead array (Bio-Rad Laboratories). A Bio-Plex calibration kit will be run to standardise the fluorescent signal on the Bio-Plex system, samples will then be ran and data quantified as per manufacturer instructions.

Circulating biomarkers

Effect of healthy food habits on modulating the level of circulating microRNAs (miRNAs) will be assessed using the fasted venous plasma taken at baseline before the intervention, and in the final week of the 12-week intervention. Total RNA will be isolated using an RNA isolation kit. A known concentration of external spike-in will be added during the last step of RNA isolation to use as a control. Concentration of total RNA will be measured using Nanodrop Spectrophotometer. Next, 20 ng of total RNA will be reverse transcribed using stem loop specific primers for miR-33, −128–1, −30c, −143, −15b. The miRNAs chosen for this study are based on their established function^26–30 in regulation of lipid metabolism, regulation of HDL, regulation of LDL, insulin resistance and insulin synthesis. The resulting cDNA will be amplified using specific amplification probes in StepOne Real-time PCR system (Thermo Fisher Scientific). The spike-in external RNA will also undergo reverse transcription and amplification. For analysis of the results, first, the relative expression (Δ^Ct) of target miRNAs will be determined by subtracting the Ct value of a target miRNA vs the Ct value of the spike-in control miRNA. The expression of target miRNA will then be expressed as 2^-ΔCT according to the MIQE guidelines.³¹ All results will be expressed as relative expression in miRNAs at 12 weeks compared with baseline.

Self-reported dietary intake and diet satisfaction

Habitual dietary intake is assessed with a short-form Food Frequency Questionnaire³² (FFQ) of 57 questions. This FFQ reflects intakes of the preceding 3 months and is validated in the NZ population. FFQs are completed at baseline before the intervention, within the final week of the 12-week intervention, and at 12 and 52 weeks after intervention completion. Four-day food diaries reflecting actual intakes³³ are recorded by participants at the same four time points as the FFQ. Participants are provided with a video link explaining how to complete a food diary, and to record four of the next 7 days, making sure 1 day is on the weekend. Recorded days do not have to be consecutive, but we requested that one recorded day be the day before the fasted blood test at baseline and within the final week of the 12-week intervention. Participants are asked to provide weighed amounts (preferred) of a measure of volume for each food consumed. Diaries are checked for details when received by study staff. Participants choose between using an online food diary app (Xryis) or a paper-based food diary to use at the four time points during the study. Every 2 weeks of the 12-week intervention period, participants are asked if they received their study supplied groceries (for the intervention arms), how much their household spent on groceries, and which of the study supplied foods the participant consumed, the family consumed, they gave away, stored or disposed of. Participants in the intervention arms are also asked every fortnight how much of each of the provided study foods they personally consumed. Diet satisfaction is assessed at baseline, in the final week of the 12-week intervention, and at 12 weeks and 52 weeks postintervention with a recognised tool of 10 questions.³⁴

Biomarkers of dietary intake

Biomarkers of dietary intake will be measured from frozen plasma samples collected at baseline and in the final week of the 12-week intervention. Plasma samples will be processed and analysed for known dietary biomarker compounds using liquid-chromatography coupled to high resolution mass spectrometry (LC-HRMS). LC-HRMS will be carried out using a Shimadzu LCMS 9030 quadrupole-time of flight mass spectrometer (Shimadzu, Kyoto, Japan) using hydrophobic interaction liquid chromatography for polar biomarkers and C18 chromatography for semipolar biomarkers. Of relevance for this study, biomarkers for plant-based foods (epicatechin, ferulic acid, cinnamic acid, α-carotene, lutein), cruciferous vegetables (3,3’-diindolylmethane, 2-thiothiazolidine-4-carboxylic acid, sulphoraphane, S-allylcysteine), nuts (pipecolic acid), fruits (phloretin), wholegrain wheat (alkylresorcinols and alkylresorcinol metabolites), wholegrain oats (avenacoside), fish (3-carboxy-4-methyl-5-propyl-2-furanpropanoic acid) and eggs (choline) will be detected by this method. Relevant fatty acids will also be measured as potential biomarkers of dietary intake.

Medication load

Medication load is measured at baseline before the intervention, in the final week of the 12-week intervention, and at 12 weeks and 52 weeks postintervention completion. To facilitate drug utilisation analysis both within the study and within the research literature, medication use is calculated using the WHO defined daily dose system.^{35 36}

Bowel habits

Participants’ bowel habits are self-reported at each of the nine time points, assessing habits over the preceding 2 weeks. We used eight questions,³⁷ supplemented with a question on evacuation frequency with seven predefined answer categories (less than once a week to five or more times a day).

Outcomes assessed in a subset of the total study

Gut microbiome

All participants were notified that they might be asked to provide a faecal sample at two time points in the study, baseline and in the final week of the 12-week intervention. This is an optional assessment where all participants can pre-emptively opt out should they wish. In reality, we will only collect faecal samples preintervention and postintervention from the first 85 consenting participants in each of the two trial arms providing free healthy groceries. With these samples, we will measure microbiota diversity, density and function using a collaborating service (Microba). We will publish preliminary findings from these analyses, and then make these data available to other research groups to address their own research questions.

Facilitators and barriers to diet change

Participants are notified that they may be invited to an interview in the last week of the 12-week intervention. The purpose of these interviews is to understand factors influencing dietary change when foods are provided within usual care. Quantitative questions include closed and (optional) open questions sent to participants to complete in REDCap. Individual, semistructured interviews of 12–15 participants per trial arm will explore facilitators and barriers to incorporating healthy foods into everyday eating. Interviews will be held over phone calls/zoom and be audio recorded. Audiorecordings will be transcribed and coded using NVIVO software. Analysis will be regarded as complete once saturation of codes is reached, with new data not altering the analysis and code system in any meaningful way.³⁸

Data collection

Demographic and self-reported outcome data are submitted by participants using REDCap, a secure web application for building and managing online surveys and databases. Data in REDCap are stored on the University of Otago servers. REDCap enables multisite users (participants and study staff) to contribute online to a secure database with one manager to oversee all data entries. REDCap access is restricted and password protected. Study staff are allocated separate access levels for different parts or functionalities in REDCap to ensure data are preserved and blinding maintained. For example, only the researcher who uploaded the randomisation list is able to access that list for the duration of recruitment, staff involved in allocating participants to trial arms cannot. Anthropometry is assessed by trained health professionals blinded to intervention at the collaborating phlebotomy services. These health professionals are external to study staff and are not briefed on the full study design or aims, so any potential incidents of disclosure of randomisation group are not tracked, and not expected to differentially bias data collection. All assessments undertaken on blood or faecal samples will be by individuals without any direct contact or knowledge of study participants beyond an ID number. Analyses will be undertaken by a biostatistician or health economist blinded to intervention.

Safety monitoring

Participants are asked to report any serious negative effects due to the trial every 2 weeks during the interventions. Self-reported negative health events, hospitalisations and negative effects due to participation in the trial are recorded at intervention end, at 12 and 52 weeks after the interventions. All self-reported entries are reviewed by study staff when received, to consider potential unintended risk with trial participation. Anything of potential viable concern is escalated to an independent data and safety monitoring board put in place during the ethical consent process before the trial commenced. Abnormal results from the clinical assessment of blood samples are alerted to study staff by the collaborating phlebotomy service, with participants contacted immediately with the results and instructions to discuss with their medical team.

Statistical analysis

Initial analyses will be performed in accordance with intention-to-treat principles (based on group randomised and using all available data without inputting values missing at random). Continuous variables, such as primary outcome LDL cholesterol, will be considered within a generalised linear model with an identity link function. Categorical variables, such as secondary outcomes bowel habits and diet satisfaction, will be considered with an appropriate link function depending on the nature of the categorical variable, for example, binary variables will use a logit link function. Some outcomes may require log transformation prior to analyses. Missing data will be investigated to see if the pattern is different by intervention arm or participant baseline characteristics. Missing outcome data will be investigated with a best case/worst case analysis as a sensitivity analysis for the primary analysis (by randomised arm). For the statistical models, any important confounders (ie, body mass) that show substantial differences at baseline will be included in the analysis to adjust out any residual confounding. Given the challenges of adjusting for multiple outcomes particularly with potentially highly correlated outcomes, no adjustment will be made, however, all secondary outcomes will be interpreted with caution with a focus on the estimate with 95% CIs. It will also be important to interpret effects (and CIs) in terms of clinical importance rather than purely on the p value alone. More novel assessment methods are mentioned above per outcome. A per-protocol analysis will also be undertaken to investigate how adherence with the intervention impacts the results, as well as investigation of any dose–response relationship. Exploratory analyses may be conducted separately to generate future hypotheses or answer future research questions. Uninformative group codes will be used to ensure that all described analyses are performed with blinding. Reporting will adhere to all items in the Consolidated Standards of Reporting Trials (CONSORT) statement, including a CONSORT flow diagram showing participant flow and reasons for exclusion and lost to follow-up where available. Analyses will be undertaken in Stata V. 17.

Patient and public involvement

Māori and non-Māori members of the public provided feedback on the study web site, promotional pamphlet and poster. They also pretested the online questionnaires, with amendments made based on their suggestions.

Ethics and dissemination

This trial has Health and Disability Ethics Committee approval (20/NTB/121), underwent Māori consultation, and has locality authority to be conducted in Canterbury, Otago and Southland. This trial was prospectively registered with the Australian New Zealand Clinical Trials Registry (ACTRN12620000689976), and has a universal trial identifier (U1111-1250-1499).

Dissemination of findings

We will disseminate the findings of this trial through peer-reviewed publications and conference presentations. We will also present our findings to our participants and stakeholders or collaborating services.

Discussion

This trial has been designed to evaluate the acute effects of free healthy grocery delivery within usual care on cardiometabolic risk factors and more exploratory risk factors in those recovering from an acute coronary event. The longer-term aim of this trial is to evaluate the cost-effectiveness of 12 weeks free healthy grocery delivery within usual care on health and social outcomes at 1 year after intervention completion. With three trial arms, we are comparing two different dietary approaches that include grocery provision with usual care, as well as outcomes between the two active interventions in terms of health benefit, uptake into habitual diets, and economic feasibility in the usual care of CHD. Alongside quantitative outcomes, this trial will identify the barriers and enablers to dietary change when foods are provided within usual care for 12 weeks. We do not yet know if free grocery provision in acute coronary event recovery improves outcomes, or leads to a change in immediate or habitual dietary intakes. The trial is ongoing; therefore, results are not yet known, and no conclusion can be drawn at this time, however, higher dietary fibre and unsaturated fat intakes have been shown effective in the general population^{17 39} and for those with other NCDs.^{40 41} Undertaking this trial, however, is an opportunity to answer these research questions, and inform acute CHD management going forward.

Ethics statements

Patient consent for publication

Not applicable.