Skip to main content
NCBI home page
Advances in Nutrition logoLink to Advances in Nutrition
. 2022 Sep 3;13(6):2125–2135. doi: 10.1093/advances/nmac071

Saturated Fatty Acid Intake and Risk of Type 2 Diabetes: An Updated Systematic Review and Dose–Response Meta-Analysis of Cohort Studies

Zahra Gaeini 1, Zahra Bahadoran 2,✉, Parvin Mirmiran 3
PMCID: PMC9776642  PMID: 36056919

ABSTRACT

This systematic review and meta-analysis was conducted to pool findings of cohort studies that investigated hazards of type 2 diabetes mellitus (T2DM) in relation to intakes of SFAs. A systematic search was conducted in the PubMed, Scopus, and Embase databases up to June 2021 to find eligible studies. Review articles or commentaries, clinical trials, cross-sectional studies, studies on gestational or type 1 diabetes patients, animal studies, articles with no access to full-texts, articles published in non-English languages, and articles with missing critical data needed for the systematic review were excluded from the meta-analysis. A random-effects model was used to combine study-specific results. Thirteen cohort studies with 361,686 participants and 11,865 T2DM events were included. Dietary total SFA intake, as well as dietary palmitic acid (PA) or stearic acid (SA) were not associated with risk of T2DM when the highest was compared with the lowest intake category (HR = 0.99; 95% CI: 0.91, 1.09; n = 13 for total SFAs; HR = 0.96; 95% CI: 0.79, 1.15; n = 4 for PA; and HR = 1.08; 95% CI: 0.79, 1.49; n = 4 for SA). However, the risk of T2DM decreased by 11% in the highest compared with the lowest category of dietary lauric acid (HR = 0.89; 95% CI: 0.82, 0.97; n = 2), and by 17% in the highest compared with lowest category of dietary myristic acid (MA) (HR = 0.83; 95% CI: 0.74, 0.92; n = 3). There was evidence of publication bias among studies on dietary total SFAs and T2DM. Our results indicated no significant association between dietary total SFA and risk of T2DM. However, dietary intake of MA was negatively associated with developing T2DM.

Keywords: saturated fats, diabetes mellitus, lauric acid, myristic acid, palmitic acid, stearic acid

Statement of significance: The prior published meta-analysis in this field investigated the association between dietary total SFAs and risk of T2DM; no systematic review and meta-analysis has been conducted for the association between individual SFAs (lauric acid, myristic acid, palmitic acid, and stearic acid) and risk of T2DM. Moreover, dose–response associations of SFAs and T2DM remained undetermined. The results of our meta-analysis showed negative associations between dietary intake of LA and MA and risk of T2DM.

Introduction

Although dietary guidelines generally recommend reducing total fat and SFA intakes (1), controversy still surrounds the diabetogenic effect of SFAs. Some of the experimental studies support the notion that dietary fats, and SFAs in particular, are associated with the development of insulin resistance and type 2 diabetes mellitus (T2DM) (2–5).  On the other hand, the majority of more recent cohort studies have indicated no association between dietary SFAs and the incidence of T2DM (6–9). In the case of dietary SFAs with different chain lengths [i.e., lauric acid (LA), myristic acid (MA), palmitic acid (PA), stearic acid (SA)], the results are inconclusive. Whereas there was no significant association between intake of SFAs and T2DM in some studies (5, 7, 9, 10), a significant negative association was reported for LA and MA (6, 8), and a significant positive association was reported for SA (11). Due to these challenges, dietary recommendations to limit SFAs for T2DM prevention might need re-evaluation.

Although a prior meta-analysis investigated the association between dietary total SFAs and T2DM incidence (12), no systematic review and meta-analysis has been conducted for the association between individual SFAs (LA, PA, MA, and SA) and risk of T2DM. Moreover, dose–response associations of SFAs and T2DM have remained undetermined.

Therefore, this study aimed to conduct an updated systematic review and a dose–response meta-analysis of cohort studies to test the linear and potential nonlinear dose–response associations between intakes of total SFAs, LA, MA, PA, and SA and the risk of T2DM. Further subgroup analyses were also conducted to clarify the possible effects of confounding factors.

Methods

We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement as guidance for reporting this systematic review (13). We also followed the 12-item PRISMA extension for writing the abstract (14).

Search strategy

We searched all English-language papers in PubMed, Scopus, and Embase databases, using appropriate keywords, according to the following search formula: (“diabetes”[Title] OR “diabetes mellitus”[Title] OR “type 2 diabetes”[Title] OR “dysglycemia”[Title] OR “diabetic”[Title] OR “NIDDM”[Title] OR ((“non-insulin”[Title] OR “noninsulin”[Title]) AND “depend”[Title])) AND (“fatty acid”[Title] OR “dietary fat”[Title] OR “saturated fat”[Title] OR “butyric acid”[Title] OR “butanoic acid”[Title] OR “caproic acid”[Title] OR “hexanoic acid”[Title] OR “caprylic acid”[Title] OR “octanoic acid”[Title] OR “decanoic acid”[Title] OR “capric acid”[Title] OR “lauric acid”[Title] OR “octadecanoic acid”[Title] OR “myristic acid”[Title] OR “tetradecanoic acid”[Title] OR “palmitic acid”[Title] OR “hexadecanoic acid”[Title] OR “stearic acid”[Title] OR “carbohydrate-restricted”[Title]). Evidence not published in commercial or academic publications (gray literature) were also searched using Google Scholar. We checked the reference lists from reviews and original studies investigating the potential association between dietary SFAs and T2DM, to find other relevant articles that might have been missed in our searches.

Eligibility and study selection

Two authors (ZG and ZB) reviewed the title and abstract of all obtained articles. The exclusion criteria for study selection were: 1) articles published in non-English languages; 2) review articles or commentaries; 3) study designs other than cohort studies including clinical trials and cross-sectional studies; 4) studies on gestational or type 1 diabetes patients; 5) animal studies; 6) articles with no access to full texts; and 7) articles with missing critical data needed for the systematic review. Studies that reported dose of SFA and adjusted effect sizes across categories of exposures; and provided the numbers of participants or person-years and events across categories were included for dose–response meta-analysis. Included studies were published between 1992 and 2020.

Study eligibility was assessed based on initial inclusion and exclusion criteria; final relevant full-text articles were included in the meta-analysis and retrieved for data extraction (Figure 1).

FIGURE 1.

FIGURE 1

Open in a new tab

Flow chart of the literature search.

Quality assessment for each study was conducted using the Cochrane Risk of Bias in Non-Randomized Studies of Exposures (ROBINS-E) tool (15) (Supplemental Table 1). The tool includes 7 domains of bias: confounding bias, selection of participants bias, exposure assessment bias, misclassification of exposures, missing data bias, measurement of outcomes, and selective reporting of the results. Disagreements were solved by consulting the principal investigator (PM).

Data extraction

Data extraction was conducted by the first author (ZG) and double-checked by the last author (PM) to ensure that all data were extracted correctly. The following variables were extracted from eligible studies: the first author's name, publication year, country, sex of participants, follow-up duration, exposure, categories of exposure, number of participants and cases per category, the dose of SFAs, LA, MA, PA, and SA intake per category, risk estimates expressed as HRs, RRs, or ORs with corresponding 95% CIs per category, and confounding factors in the multivariable analysis (Supplemental Table 2). For all meta-analyses, we used the maximally adjusted effect sizes reported in primary studies.

Data synthesis and statistical analysis

Due to the high heterogeneity observed between studies in our meta-analysis (>50%) and considering the potential for high variation within and between the observational studies, we performed random-effects meta-analyses to calculate summary HRs and 95% CIs for a 1% increase in the percentage of fat intake from total energy (16). The reported RRs were considered equal to HRs (17). We conducted linear dose–response meta-analyses using the method of Greenland and Longnecker (18). For this method, the distribution of events and participants or person-years and adjusted risk estimates across categories of SFAs were needed. When studies reported the ranges of exposure categories, instead of the direct median of each category, we estimated approximate medians by the midpoint between the lower and upper limit. For open categories, we assumed the same range as the adjunct category. If a study did not report the numbers of participants or person-years in each category, if the exposures were defined as quantiles, the distribution of participants and person-years was estimated by dividing the total number of participants or person-years by the number of categories (19). If the dietary SFA intake was reported as grams per day, we converted them to energy percentage. For the 1 study (20) that did not consider the lowest category as a reference, the effect size was recalculated, assuming the lowest category as reference, using the method suggested by Hamling et al. (21).

We performed subgroup analyses by sex, geographical location, follow-up duration, number of participants, exposure assessment method, study quality, and adjustment for main confounders. Also, sensitivity analyses were conducted to evaluate the potential influence of each study on the results by re-estimating the HRs after excluding 1 study at a time. We evaluated between-study heterogeneity by using the I2 statistic (specific categories such as low = 25%, moderate = 50%, and high = 75%) (22). Publication bias was assessed using funnel plots and Egger regression test asymmetry (23) if sufficient studies existed (n ≥ 10) (24).

Moreover, we performed a Pnonlinearityr dose–response meta-analysis, using the Wald test, to determine whether there was departure from linearity, and to test the potential nonlinear association between dietary SFAs and risk of T2DM (25). For the nonlinear dose–response meta-analysis, we modeled curvilinear dose–response associations using a 1-stage weighted mixed-effects meta-analysis (26). The exposures were modeled using restricted cubic splines (27): a spline function set of smoothly joined curves that curve at knots. The knots were based upon Harrell's recommended percentiles for 3 knots at 10th, 50th, and 90th percentiles (27). The correlation within each category of published RRs was taken into account, and the study-specific estimates were combined by using a 1-stage weighted mixed-effects meta-analysis (26, 28). This method estimates the study-specific slope lines and combines them to obtain an overall average slope in a single stage (18, 25).

Finally, we repeated meta-analyses with the inclusion of 1 unpublished dataset in the full data (Supplemental Figures 1–4).

All analyses were conducted with Stata software, version 16 (Stata Corp). P < 0.05 was considered statistically significant.

Results

After removing duplicates, 167 publications were found through database searching. We reviewed the titles and abstracts of all articles, and 153 were removed. The full texts of the remaining studies were assessed for eligibility, and 1 study was excluded. Ultimately, 13 prospective cohort studies with 361,686 participants and 11,865 T2DM events were included in the final meta-analysis (5–11, 20, 29–33).

Characteristics of included studies

Of the 13 studies, 6 were from Europe, 5 from the United States and 2 from Australia. Follow-up durations were between 4 and 17 y (median follow-up duration was 8.0 y, with IQR = 5.1–13.0 y). Of the 13 studies, 12 studies reported the risk of T2DM for total dietary SFAs (5–11, 20, 29–31, 33), 2 studies reported the risk for dietary LA (6, 8), 3 studies for MA (6, 8, 32), 4 studies for PA (6, 8, 11, 32), and 4 studies for SA (6, 8, 11, 32). Four studies included only women (9, 10, 30, 33), 1 study included only men (5), and another 9 studies included both men and women (6–8, 11, 20, 29, 31, 32). T2DM diagnosis was based on self-report in 5 studies (6, 9, 30, 33, 34); however, 8 studies confirmed self-reports of T2DM by assessing the glucose concentrations in blood samples. The general characteristics of the included studies are presented in Supplemental Table 2.

Quality assessment of included studies, using the ROBINS-E tool, revealed that 9 studies were at serious risk of bias, and 5 studies had moderate risk of bias (Supplemental Table 1). Mostly, the high risk of bias was due to not controlling confounders, selection bias due to the selection of participants into the study, or bias due to changes in exposure during follow-up that were not measured.

Total saturated fats

Figure 2 summarizes the results of the meta-analysis for total dietary SFAs as a risk factor for T2DM. Dietary SFA intake was not associated with risk of T2DM when the highest was compared with the lowest intake category (HR: 0.99; 95% CI: 0.91, 1.09; n = 13), with high heterogeneity (I2 = 54.4%; Pheterogeneity = 0.010). In sensitivity analysis, summary results did not change when each study was sequentially excluded from main analysis (HRs = 0.96–1.01). In the subgroup analyses, we observed no significant differences in meta-analysis results between the categories of studies based on sex, geographical location, follow-up duration, diabetes diagnostic method, number of participants, study quality, and main adjustments (Table 1).

FIGURE 2.

FIGURE 2

Open in a new tab

Forest plot of cohort studies showing weighted mean differences in risk of type 2 diabetes between highest vs. lowest category of total saturated fat intake, for all eligible studies. Analysis was conducted using a random-effects model. DL, DerSimonian & Laird.

TABLE 1.

Subgroup analyses of total saturated fat intakes (energy percentage) and risk of type 2 diabetes

Characteristics n HR (95% CI) I 2, % P-heterogeneity P-interaction1
All studies 13 0.99 (0.90, 1.09) 54.4 0.010
Sex 0.513
 Women 5 0.95 (0.83, 1.07) 0 0.564
 Men 1 0.97 (0.79, 1.20) —
 Men and women 7 1.06 (0.91, 1.22) 74.2 0.001
Geographical region 0.584
 United States 5 0.97 (0.87, 1.09) 0 0.922
 Europe 6 1.05 (0.88, 1.25) 72.1 0.006
 Oceania 2 1.15 (0.44, 2.98) 89.5 0.002
Follow-up duration, y 0.175
 <10 7 1.15 (0.89, 1.49) 55.5 0.036
 >10 6 0.95 (0.88, 1.03) 54.4 0.085
Diabetes diagnosis method 0.075
 Self-report 6 0.93 (0.85, 1.01) 0 0.676
 Blood sampling 7 1.09 (0.93, 1.29) 73.4 0.001
Number of participants 0.160
 <10,000 5 1.26 (0.86, 1.84) 64.2 0.025
 >10,000 8 0.95 (0.89, 1.02) 34 0.157
Study bias 0.055
 Moderate 5 1.07 (0.98, 1.17) 0 0.519
 Serious 8 0.94 (0.85, 1.04) 47.5 0.064
Main confounders adjustments 0.681
 Yes 9 0.97 (0.90, 1.06) 46.8 0.058
 No 4 1.06 (0.73, 1.53) 72 0.013
Open in a new tab
1

P-interaction (or P-between) refers to the significance of the difference between subgroups.

The linear dose–response meta-analysis of the main 13 studies showed no linear association between increasing intake of SFAs and T2DM risk (HR: 0.93; 95% CI: 0.84, 1.03). From 13 cohort studies regarding the association between total SFAs and T2DM risk, 7 studies (5–7, 9, 31, 33, 34) reported sufficient data for the nonlinear dose–response analyses. There was no evidence of a U- or J-shaped association between total SFA intake and risk of T2DM (Pnonlinearity = 0.153; n = 7; Figure 3). Supplemental Figure 5 presents the results for publication bias. Overall, there was evidence of publication bias with the Egger test (P = 0.032).

FIGURE 3.

FIGURE 3

Open in a new tab

Dose–response association between dietary total saturated fat and risk of type 2 diabetes. The solid line and the dashed lines represent the estimated HRs and 95% CIs, respectively. The solid circles and the open circles represent the reference categories and other categories of dietary saturated fats intake, respectively. E%, percentage from total energy intake.

Individual SFAs

Figure 4 summarizes the results of the meta-analysis for dietary LA, MA, PA, and SA regarding T2DM. Risk of T2DM decreased by 11% in the highest compared with lowest category of dietary LA (HR: 0.89; 95% CI: 0.82, 0.97; n = 2), with no evidence of heterogeneity (I2 = 37.8%; Pheterogeneity = 0.205). Dietary MA intake was associated with a 17% lower risk of T2DM when the highest was compared with the lowest intake category (HR: 0.83; 95% CI: 0.74, 0.92; n = 3), with no evidence of heterogeneity (I2 = 0%; Pheterogeneity = 0.522). The risk of T2DM was not associated with categories of PA (HR: 0.96; 95% CI: 0.79, 1.15; I2 = 33.9%; n = 4), or SA (HR: 1.08; 95% CI: 0.79, 1.49; I2 = 78.2%; n = 4).

FIGURE 4.

FIGURE 4

Open in a new tab

Forest plot of cohort studies showing weighted mean differences in risk of type 2 diabetes between highest vs. lowest categories of different saturated fatty acids intake. Analysis was conducted using a random-effects model. DL, DerSimonian & Laird.

The linear dose–response meta-analysis of the association between dietary MA intake and risk of T2DM showed a 12% lower risk of T2DM with each 1% increase in MA intake (HR: 0.88; 95% CI: 0.82, 0.95). However, there was no evidence that increased intakes of PA and SA were associated with risk of T2DM (HR: 0.82; 95% CI: 0.63, 1.06 for PA; and HR: 0.75; 95% CI: 0.42, 1.36 for SA). Furthermore, there was no evidence of a U- or J-shaped association between each SFA intake and risk of T2DM (Pnonlinearity for MA = 0.395; Pnonlinearity for PA = 0.549; Pnonlinearity for SA = 0.795; Figure 5). Potential publication bias was not assessed (n < 10). We could not assess the dose–response association between LA and risk of T2DM, because only 1 study reported sufficient data for the dose–response analysis of LA and risk of T2DM.

FIGURE 5.

FIGURE 5

Open in a new tab

Dose–response association between different saturated fatty acids and risk of type 2 diabetes. The solid lines and the dashed lines represent the estimated HRs and 95% CIs, respectively. The circles represent the categories of dietary saturated fatty acids. E%, percentage from total energy intake.

We repeated the meta-analyses after inclusion of 1 unpublished dataset (Supplemental Figures 1–4). As with the analysis excluding the unpublished data, those including the unpublished results showed no significant associations between total SFAs, PA, and SA and risk of T2DM, whereas higher intakes of LA and MA were associated with lower risks of T2DM (published articles only: HR: 0.89; 95% CI: 0.82, 0.97, n = 2 for LA; HR: 0.83; 95% CI: 0.74, 0.92, n = 3 for MA; adding unpublished data: HR: 0.90; 95% CI: 0.84, 0.96; n = 3 for LA; HR: 0.85; 95% CI: 0.75, 0.96; n = 4 for MA). Because the unpublished data had sufficient data regarding dose–response analyses, we repeated dose–response meta-analyses for total SFAs, and all 4 individual SFAs. There was a 20% lower risk of T2DM with each 1% increase in LA intake (HR: 0.80; 95% CI: 0.70, 0.92). However, there was no evidence that increasing intakes of total SFAs, MA, PA, and SA were associated with risk of T2DM.

Discussion

In the present systematic review and dose–response meta-analysis, we pooled current data from available prospective cohort studies to present a relatively broad overview of the association between dietary SFAs, LA, MA, PA, and SA and risk of T2DM. Our findings showed no significant association between dietary SFA intake and the risk of developing T2DM; however, medium-chain SFAs (LA and MA) exerted protective association against the development of T2DM.

The lack of association between dietary total SFA intakes and risk of T2DM is consistent with findings from previous meta-analyses of cross-sectional studies (12, 35, 36). In a meta-analysis of randomized controlled feeding trials estimating the effects of isocaloric replacements between macronutrients, there was no significant change in fasting glucose concentration by replacing 5% energy from carbohydrate with SFAs (+0.02 mmol/L; 95% CI = −0.01, +0.04; n trials = 99) (37).

The included studies were heterogeneous; this high heterogeneity can be explained by differences in follow-up duration of studies or number of participants, varying population groups, or differences in dietary sources of SFAs that each population consumed. Processed and red meats, as the major sources of SFAs in the European and American populations’ diet, were associated with a higher risk of T2DM (38, 39). In contrast, total dairy products, low-fat dairy products, and cheese, which are also primary sources of SFAs, had an inverse association with the risk of T2DM (40). Intakes of dairy products (3 servings/d) in the context of a healthy diet could reduce the risk of T2DM (41).

Recently, particular attention has focused on the effects of dairy products on cardiometabolic health. Randomized controlled trials (RCTs) have supported the hypothesis that a complex food matrix such as milk, cheese, or yogurt can ameliorate the potential adverse effects of SFAs on metabolic health (42). More RCTs with a large sample and long follow-up duration are needed to fully determine the effects of dairy products and other sources of SFAs on cardiometabolic outcomes.

The potential effects of dietary SFAs on insulin sensitivity are under debate. Reducing dietary SFAs in subjects with metabolic syndrome did not affect insulin sensitivity (43). An SFA-rich diet induced whole-body insulin resistance after a 24-h period (44). Replacement of a MUFA-rich diet with an SFA-rich diet had favorable effects on insulin sensitivity (45).

In contrast to the epidemiological findings, experimental studies and high-fat-diet–induced animal models of T2DM and insulin resistance mostly support diabetogenic effects of dietary fats, and SFAs in particular (46–48). The result of an animal study that tested the impact of SFAs with different chain lengths on insulin resistance is notable. It reported that animals fed a high-fat obesogenic diet with high amounts of medium-chain SFAs, particularly LA, had greater insulin sensitivity compared with animals fed an obesogenic diet with high concentrations of long-chain SFAs, although both obesogenic groups still had greater insulin resistance than a control group fed a normal-fat diet (49).

In vitro studies suggested that exposure of cultured muscle cells to SFAs, specially PA, can induce insulin resistance due to their proinflammatory properties and induction of cytokines, such as TNFα, in blood or tissues, activation of cell stress pathways, elevation of SFA metabolites (such as ceramides), inhibition of the phosphorylation cascade downstream of the insulin receptor, and insulin-induced glucose uptake (50–53). Also, an animal study showed that the markers of adipose tissue inflammation and systemic insulin resistance were lower in mice fed a high-LA diet, compared with mice fed a high-PA diet (54). However, the studies revealed that some potential confounders in experimental conditions, such as SFA-solubilizing agents, might affect interpretation of the effects of SFAs on inflammation and other cellular responses, so re-evaluation of the results of the culture studies seems essential (52). However, because T2DM is a multifactorial and multistage metabolic disorder, misclassification and missing actual cases have been suggested as a reason for reduced magnitude of the RR for dietary fat intakes in population-based cohorts (55). Different case definitions (e.g., hyperinsulinemia with normal glucose tolerance compared with established T2DM) could be responsible for inconsistent findings of epidemiological studies (55). Therefore, the case definition might need to consider the T2DM stage because a large number of steps are involved in the progression of T2DM where nutrient influences might occur (55). The variables included in multiple regression analyses could be a source of diverse findings of population-based studies; overadjustment for confounders modifying the potential association between dietary fats and T2DM risk (e.g., genetic susceptibility and physical activity) could underestimate the strength of the fat-T2DM associations (55). On the other hand, the divergent results could be real because the association of dietary fat and T2DM might vary with population characteristics, such as age, BMI, and physical activity, that are associated with insulin sensitivity (5). Notably, the high-fat diets used to induce T2DM and insulin resistance in animal models are usually composed of 45% fats (approximately one-third SFAs and two-thirds unsaturated fats) (47), which differs from the high-fat diets prescribed in human studies to induce metabolic alterations.

To the best of our knowledge, this is the first meta-analysis of prospective cohort studies examining the associations of individual SFAs with different chain lengths (LA, MA, PA, and SA) with risk of T2DM. We found that higher dietary intake of SFAs with medium chain lengths (LA and MA) might have a protective association with the risk of T2DM, whereas there was no significant association between SFAs with longer chain lengths (PA and SA) and T2DM.

One remarkable finding of our meta-analysis was that each 1% increase in energy intake from MA was associated with a 12% lower risk of T2DM. However, due to the limited number of studies with sufficient data for the dose–response analysis of LA and risk of T2DM (n = 1), we could not assess the dose–response association between LA and risk of T2DM. Two large prospective cohort studies indicated that higher dietary intake of LA was associated with significantly lower risk of T2DM (6, 8). There are limited data regarding the association of LA intake and T2DM, although the effect of coconut oil, as a rich source of LA (56), has been examined in a number of studies. An RCT conducted in 48 subjects with metabolic syndrome reported that subjects who consumed 30 mL virgin coconut oil for 4 wk had a significantly lower concentration of fasting blood sugar, compared with the control group (56). Also, a case report of a 66-y-old man with T2DM who had insulin treatment, found that he began experiencing hypoglycemia within 1–2 d of starting the coconut oil supplementation (57). On the other hand, a protective association between dietary MA and T2DM was reported by most of the previous studies (6, 8, 58, 59). This finding is largely consistent with studies that examined circulating concentration of MA, which could more accurately reflect its concentrations from both dietary intake and endogenous synthesis (11, 60).

To sum up, considering the null association between dietary total SFAs and risk of T2DM and the protective association between dietary SFAs with medium-chain lengths and T2DM observed in this meta-analysis and previous studies, and given the importance of early prevention strategies for reducing the risk of T2DM, it seems necessary to reconsider the dietary recommendations regarding limitation of SFAs to prevent hyperglycemia and instead emphasize the types and quality of fats consumed within the context of a healthy dietary pattern.

Strengths and limitations

To the best of our knowledge, this is the first study that provides a comprehensive review and dose–response meta-analysis of the association between dietary SFAs with different chain lengths and the risk of T2DM. Also, dose–response associations of SFAs and T2DM have not been investigated before. However, our study has some limitations. The high level of heterogeneity among the studies can be attributed to the diversity of populations, their different cultural and genetic backgrounds, geographical variation, and different follow-up durations. However, we tried to detect potential sources of heterogeneity of the studies by conducting subgroup analyses. Moreover, we did not assess the dose–response association between LA and risk of T2DM, because only 1 study reported sufficient data for such analysis.

We had some limitations for selecting the eligible studies because we searched only for English-language databases. Due to the limited number of included studies, publication bias could only be assessed for total SFAs; there was evidence of publication bias, which can lead to downgrading of evidence. The number of studies in each subgroup was also limited, especially for studies with men participants and studies in Oceania, which could affect the results of the subgroup analysis. Also, all of the included studies relied on questionnaires to estimate typical fat intake, which is a significant source of measurement errors in estimating food and nutrient intakes. It should be noted that SFAs are consumed as components of foods and meals and they are not eaten in pure form, and composition of fatty acids, particularly from animal sources like dairy and beef, can vary greatly with management conditions such as breed and food source. Further, well-designed prospective cohort studies and clinical trials with appropriate controlling for potential confounders and validated dietary intakes with nutrient concentrations in blood and tissues are needed.

Conclusions

In conclusion, this study supports the null association between dietary total SFAs and risk of T2DM. However, dietary LA and MA seem to have a protective association with developing T2DM.

Supplementary Material

nmac071_Supplemental_Files
Click here for additional data file. (136.2KB, zip)

Acknowledgements

The authors’ responsibilities were as follows—ZG: designed the study; ZG, ZB: collected data; ZG, PM: analyzed data; ZG: wrote the manuscript; ZB: corrected the manuscript; and all authors: read and approved the final manuscript.

Notes

The authors reported no funding received for this study.

Author disclosures: The authors report no conflicts of interest.

Supplemental Figures 1–5 and Supplemental Tables 1 and 2 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/advances/.

Abbreviations used: LA, lauric acid; MA, myristic acid; PA, palmitic acid; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RCT, randomized controlled trial; ROBINS-E, Cochrane Risk of Bias in Non-Randomized Studies of Exposures tool; SA, stearic acid; T2DM, type 2 diabetes mellitus.

Contributor Information

Zahra Gaeini, Nutrition and Endocrine Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran.

Zahra Bahadoran, Nutrition and Endocrine Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran.

Parvin Mirmiran, Department of Clinical Nutrition and Dietetics, Faculty of Nutrition Sciences and Food Technology, National Nutrition and Food Technology Research Institute, Shahid Beheshti University of Medical Sciences, Tehran, Iran.

Data Availability

The datasets analyzed during the current study are available from the corresponding author on reasonable request.

References

  • 1. Institute of Medicine (US) Committee on Dietary Guidelines Implementation , Thomas PR, editors. Improving America's diet and health: from recommendations to action. Washington (DC): National Academies Press; 1991. [PubMed] [Google Scholar]
  • 2. Feskens EJ, Kromhout D. Habitual dietary intake and glucose tolerance in euglycaemic men: the Zutphen Study. Int J Epidemiol. 1990;19(4):953–9. [DOI] [PubMed] [Google Scholar]
  • 3. Feskens EJ, Virtanen SM, Räsänen L, Tuomilehto J, Stengård J, Pekkanen Jet al. Dietary factors determining diabetes and impaired glucose tolerance. A 20-year follow-up of the Finnish and Dutch cohorts of the Seven Countries Study. Diabetes Care. 1995;18(8):1104–12. [DOI] [PubMed] [Google Scholar]
  • 4. Marshall JA, Bessesen DH. Dietary fat and the development of type 2 diabetes. Diabetes Care. 2002;25(3):620–2. [DOI] [PubMed] [Google Scholar]
  • 5. van Dam RM, Willett WC, Rimm EB, Stampfer MJ, Hu FB. Dietary fat and meat intake in relation to risk of type 2 diabetes in men. Diabetes Care. 2002;25(3):417–24. [DOI] [PubMed] [Google Scholar]
  • 6. Ericson U, Hellstrand S, Brunkwall L, Schulz CA, Sonestedt E, Wallström Pet al. Food sources of fat may clarify the inconsistent role of dietary fat intake for incidence of type 2 diabetes. J Diabetes Investig. 2015;101(5):1065–80. [DOI] [PubMed] [Google Scholar]
  • 7. Guasch-Ferré M, Becerra-Tomás N, Ruiz-Canela M, Corella D, Schröder H, Estruch Ret al. Total and subtypes of dietary fat intake and risk of type 2 diabetes mellitus in the Prevención con Dieta Mediterránea (PREDIMED) study. Am J Clin Nutr. 2017;105(3):723–35. [DOI] [PubMed] [Google Scholar]
  • 8. Liu S, van der Schouw YT, Soedamah-Muthu SS, Spijkerman AMW, Sluijs I. Intake of dietary saturated fatty acids and risk of type 2 diabetes in the European Prospective Investigation into Cancer and Nutrition-Netherlands cohort: associations by types, sources of fatty acids and substitution by macronutrients. Eur J Nutr. 2019;58(3):1125–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Meyer KA, Kushi LH, Jacobs DR Jr, Folsom AR. Dietary fat and incidence of type 2 diabetes in older Iowa women. Diabetes Care. 2001;24(9):1528–35. [DOI] [PubMed] [Google Scholar]
  • 10. Alhazmi A, Stojanovski E, McEvoy M, Garg ML. Macronutrient intake and type 2 diabetes risk in middle-aged Australian women. Results from the Australian Longitudinal Study on Women's Health. Public Health Nutr. 2014;17(7):1587–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Hodge AM, English DR, O'Dea K, Sinclair AJ, Makrides M, Gibson RAet al. Plasma phospholipid and dietary fatty acids as predictors of type 2 diabetes: interpreting the role of linoleic acid. Am J Clin Nutr. 2007;86(1):189–97. [DOI] [PubMed] [Google Scholar]
  • 12. Neuenschwander M, Barbaresko J, Pischke CR, Iser N, Beckhaus J, Schwingshackl Let al. Intake of dietary fats and fatty acids and the incidence of type 2 diabetes: a systematic review and dose-response meta-analysis of prospective observational studies. PLoS Med. 2020;17(12):e1003347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew Met al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. [Internet]2015;4(1). Available from: 10.1186/2046-4053-4-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Beller EM, Glasziou PP, Altman DG, Hopewell S, Bastian H, Chalmers Iet al. PRISMA for abstracts: reporting systematic reviews in journal and conference abstracts. PLoS Med. 2013;10(4):e1001419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Morgan RL, Thayer KA, Santesso N, Holloway AC, Blain R, Eftim SEet al. Evaluation of the risk of bias in non-randomized studies of interventions (ROBINS-I) and the ‘target experiment’ concept in studies of exposures: rationale and preliminary instrument development. Environ Int. 2018;120:382–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1(2):97–111. [DOI] [PubMed] [Google Scholar]
  • 17. Symons MJ, Moore DT. Hazard rate ratio and prospective epidemiological studies. J Clin Epidemiol. 2002;55(9):893–9. [DOI] [PubMed] [Google Scholar]
  • 18. Greenland S, Longnecker MP. Methods for trend estimation from summarized dose-response data, with applications to meta-analysis. Am J Epidemiol. 1992;135(11):1301–9. [DOI] [PubMed] [Google Scholar]
  • 19. Jayedi A, Rashidy-Pour A, Parohan M, Zargar MS, Shab-Bidar S. Dietary antioxidants, circulating antioxidant concentrations, total antioxidant capacity, and risk of all-cause mortality: a systematic review and dose-response meta-analysis of prospective observational studies. Adv Nutr. 2018;9(6):701–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Mandalazi E, Drake I, Wirfält E, Orho-Melander M, Sonestedt E. A high diet quality based on dietary recommendations is not associated with lower incidence of type 2 diabetes in the Malmö diet and cancer cohort. Int J Mol Sci. 2016;17(6):901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Hamling J, Lee P, Weitkunat R, Ambühl M. Facilitating meta-analyses by deriving relative effect and precision estimates for alternative comparisons from a set of estimates presented by exposure level or disease category. Stat Med. 2008;27(7):954–70. [DOI] [PubMed] [Google Scholar]
  • 22. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Higgins JP. Cochrane handbook for systematic reviews of interventions version 5.0.1 [Internet]. The Cochrane Collaboration; 2008; [cited 2008]. Available from: https://training.cochrane.org/handbook. [Google Scholar]
  • 25. Orsini N, Li R, Wolk A, Khudyakov P, Spiegelman D. Meta-analysis for linear and nonlinear dose-response relations: examples, an evaluation of approximations, and software. Am J Epidemiol. 2012;175(1):66–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Crippa A, Discacciati A, Bottai M, Spiegelman D, Orsini N. One-stage dose-response meta-analysis for aggregated data. Stat Methods Med Res. 2019;28(5):1579–96. [DOI] [PubMed] [Google Scholar]
  • 27. Harrell FE Jr. Regression modeling strategies: with applications to linear models, logistic and ordinal regression, and survival analysis. Springer; 2015. [Google Scholar]
  • 28. Orsini N. Weighted mixed-effects dose–response models for tables of correlated contrasts. Stata J. 2021;21(2):320–47. [Google Scholar]
  • 29. Churuangsuk C, Lean MEJ, Combet E. Lower carbohydrate and higher fat intakes are associated with higher hemoglobin A1c: findings from the UK National Diet and Nutrition Survey 2008–2016. Eur J Nutr. 2020;59(6):2771–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Colditz GA, Manson JE, Stampfer MJ, Rosner B, Willett WC, Speizer FE. Diet and risk of clinical diabetes in women. Am J Clin Nutr. 1992;55(5):1018–23. [DOI] [PubMed] [Google Scholar]
  • 31. Lindström J, Peltonen M, Eriksson JG, Louheranta A, Fogelholm M, Uusitupa Met al. High-fibre, low-fat diet predicts long-term weight loss and decreased type 2 diabetes risk: the Finnish Diabetes Prevention Study. Diabetologia. 2006;49(5):912–20. [DOI] [PubMed] [Google Scholar]
  • 32. Ma W, Wu JH, Wang Q, Lemaitre RN, Mukamal KJ, Djoussé Let al. Prospective association of fatty acids in the de novo lipogenesis pathway with risk of type 2 diabetes: the Cardiovascular Health Study. Am J Clin Nutr. 2015;101(1):153–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Salmerón J, Hu FB, Manson JE, Stampfer MJ, Colditz GA, Rimm EBet al. Dietary fat intake and risk of type 2 diabetes in women. Am J Clin Nutr. 2001;73(6):1019–26. [DOI] [PubMed] [Google Scholar]
  • 34. Alhazmi A, Stojanovski E, McEvoy M, Garg ML. Macronutrient intake and type 2 diabetes risk in middle-aged Australian women. Results from the Australian Longitudinal Study on Women's Health. Public Health Nutr. 2014;17(7):1587–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Alhazmi A, Stojanovski E, McEvoy M, Garg ML. Macronutrient intakes and development of type 2 diabetes: a systematic review and meta-analysis of cohort studies. J Am Coll Nutr. 2012;31(4):243–58. [DOI] [PubMed] [Google Scholar]
  • 36. de Souza RJ, Mente A, Maroleanu A, Cozma AI, Ha V, Kishibe Tet al. Intake of saturated and trans unsaturated fatty acids and risk of all cause mortality, cardiovascular disease, and type 2 diabetes: systematic review and meta-analysis of observational studies. BMJ. 2015;351:h3978. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Imamura F, Micha R, Wu JHY, de Oliveira Otto MC, Otite FO, Abioye AIet al. Effects of saturated fat, polyunsaturated fat, monounsaturated fat, and carbohydrate on glucose-insulin homeostasis: a systematic review and meta-analysis of randomised controlled feeding trials. PLoS Med. 2016;13(7):e1002087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Kouvari M, Notara V, Kalogeropoulos N, Panagiotakos DB. Diabetes mellitus associated with processed and unprocessed red meat: an overview. Int J Food Sci Nutr. 2016;67(7):735–43. [DOI] [PubMed] [Google Scholar]
  • 39. Pan A, Sun Q, Bernstein AM, Schulze MB, Manson JE, Willett WCet al. Red meat consumption and risk of type 2 diabetes: 3 cohorts of US adults and an updated meta-analysis. Am J Clin Nutr. 2011;94(4):1088–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Aune D, Norat T, Romundstad P, Vatten LJ. Dairy products and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies. Am J Clin Nutr. 2013;98(4):1066–83. [DOI] [PubMed] [Google Scholar]
  • 41. Mitri J, Mohd Yusof BN, Maryniuk M, Schrager C, Hamdy O, Salsberg V. Dairy intake and type 2 diabetes risk factors: a narrative review. Diabetes Metab Syndr. 2019;13(5):2879–87. [DOI] [PubMed] [Google Scholar]
  • 42. Poppitt SD. Cow's milk and dairy consumption: is there now consensus for cardiometabolic health?. Front Nutr. 2020;7:574725. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Tierney AC, McMonagle J, Shaw DI, Gulseth HL, Helal O, Saris WHet al. Effects of dietary fat modification on insulin sensitivity and on other risk factors of the metabolic syndrome—LIPGENE: a European randomized dietary intervention study. Int J Obes. 2011;35(6):800–9. [DOI] [PubMed] [Google Scholar]
  • 44. Koska J, Ozias MK, Deer J, Kurtz J, Salbe AD, Harman SMet al. A human model of dietary saturated fatty acid induced insulin resistance. Metabolism. 2016;65(11):1621–8. [DOI] [PubMed] [Google Scholar]
  • 45. Vessby B, Uusitupa M, Hermansen K, Riccardi G, Rivellese AA, Tapsell LCet al. Substituting dietary saturated for monounsaturated fat impairs insulin sensitivity in healthy men and women: the KANWU study. Diabetologia. 2001;44(3):312–9. [DOI] [PubMed] [Google Scholar]
  • 46. Wang C-Y, Liao JK. A mouse model of diet-induced obesity and insulin resistance. Methods Mol Biol. 2012;821:421–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Stott NL, Marino JS. High fat rodent models of type 2 diabetes: from rodent to human. Nutrients. 2020;12(12):3650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Heydemann A. An overview of murine high fat diet as a model for type 2 diabetes mellitus. J Diabetes Res. 2016;2016:2902351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Žáček P, Bukowski M, Mehus A, Johnson L, Zeng H, Raatz Set al. Dietary saturated fatty acid type impacts obesity-induced metabolic dysfunction and plasma lipidomic signatures in mice. J Nutr Biochem. 2019;64:32–44. [DOI] [PubMed] [Google Scholar]
  • 50. Blackburn ML, Ono-Moore KD, Sobhi HF, Adams SH. Carnitine palmitoyltransferase 2 knockout potentiates palmitate-induced insulin resistance in C(2)C(12) myotubes. Am J Physiol Endocrinol Metab. 2020;319(2):E265–75. [DOI] [PubMed] [Google Scholar]
  • 51. Griffin ME, Marcucci MJ, Cline GW, Bell K, Barucci N, Lee Det al. Free fatty acid-induced insulin resistance is associated with activation of protein kinase C theta and alterations in the insulin signaling cascade. Diabetes. 1999;48(6):1270–4. [DOI] [PubMed] [Google Scholar]
  • 52. Ono-Moore KD, Blackburn ML, Adams SH. Is palmitate truly proinflammatory? Experimental confounders and context-specificity. Am J Physiol Endocrinol Metab. 2018;315(5):E780–94. [DOI] [PubMed] [Google Scholar]
  • 53. Szendroedi J, Yoshimura T, Phielix E, Koliaki C, Marcucci M, Zhang Det al. Role of diacylglycerol activation of PKCθ in lipid-induced muscle insulin resistance in humans. Proc Natl Acad Sci USA. 2014;111(26):9597–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Saraswathi V, Kumar N, Gopal T, Bhatt S, Ai W, Ma Cet al. Lauric acid versus palmitic acid: effects on adipose tissue inflammation, insulin resistance, and non-alcoholic fatty liver disease in obesity. Biology. 2020;9(11):346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. Marshall JA, Bessesen DH. Dietary fat and the development of type 2 diabetes. Diabetes Care. 2002;25(3):620–2. [DOI] [PubMed] [Google Scholar]
  • 56. Nikooei P, Hosseinzadeh-Attar MJ, Asghari S, Norouzy A, Yaseri M, Vasheghani-Farahani A. Effects of virgin coconut oil consumption on metabolic syndrome components and asymmetric dimethylarginine: a randomized controlled clinical trial. Nutr Metab Cardiovasc Dis. 2021;31(3):939–49. [DOI] [PubMed] [Google Scholar]
  • 57. Malaeb S, Spoke C. The glucose-lowering effects of coconut oil: a case report and review of the literature. Case Rep Endocrinol. 2020;2020:8841781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58. Takato T, Iwata K, Murakami C, Wada Y, Sakane F. Chronic administration of myristic acid improves hyperglycaemia in the Nagoya-Shibata-Yasuda mouse model of congenital type 2 diabetes. Diabetologia. 2017;60(10):2076–83. [DOI] [PubMed] [Google Scholar]
  • 59. Wada Y, Sakiyama S, Sakai H, Sakane F. Myristic acid enhances diacylglycerol kinase δ-dependent glucose uptake in myotubes. Lipids. 2016;51(8):897–903. [DOI] [PubMed] [Google Scholar]
  • 60. Hodson L, Skeaff CM, Fielding BA. Fatty acid composition of adipose tissue and blood in humans and its use as a biomarker of dietary intake. Prog Lipid Res. 2008;47(5):348–80. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

nmac071_Supplemental_Files
Click here for additional data file. (136.2KB, zip)

Data Availability Statement

The datasets analyzed during the current study are available from the corresponding author on reasonable request.

Articles from Advances in Nutrition are provided here courtesy of American Society for Nutrition

RESOURCES