A comprehensive systematic review of 17 randomized trials involving 66,337 participants reveals that interventions to reduce saturated fatty acid (SFA) intake produce markedly different outcomes based on baseline cardiovascular risk—demonstrating little benefit for low-risk patients while showing important mortality reductions in high-risk populations over 5 years.
The analysis, published in Annals of Internal Medicine, stratified results across 4 cardiovascular risk levels and found that for patients at low baseline cardiovascular risk (<5%), reducing saturated fat intake resulted in fewer than 5 events per 1000 patients over 5 years—below the researchers' predetermined threshold for clinical importance. However, among those at high baseline risk (≥20%), the intervention yielded 6 fewer deaths per 1000 participants for all-cause mortality and demonstrated potentially important reductions in cardiovascular mortality, nonfatal myocardial infarction, and stroke.
"For persons at low cardiovascular risk, reducing or modifying saturated fat intake has little or no benefit over a period of 5 years. Among persons at high cardiovascular risk, low- to moderate-certainty evidence was found for important reductions in mortality and major cardiovascular events, particularly for MI, with respect to replacing saturated fat with polyunsaturated fat," wrote lead study author Jeremy P. Steen, BHSc (Hons), of the University of Toronto, and colleagues.
The researchers employed risk-stratified evidence synthesis using baseline risks derived from the Cholesterol Treatment Trialists' Collaboration meta-analysis of statin trials. The review included trials with intended dietary intervention durations of at least 24 months and assessed outcomes including all-cause mortality (risk ratio [RR], 0.96), cardiovascular mortality (RR, 0.93), nonfatal myocardial infarction (RR, 0.86), and fatal and nonfatal stroke (RR, 0.83).
Macronutrient Replacement Patterns
Subgroup analyses revealed differential effects based on replacement macronutrients. When SFA was replaced primarily with polyunsaturated fat (PUFA), researchers observed a more pronounced benefit for nonfatal myocardial infarction (RR, 0.75).
For high-risk patients, replacing SFA with PUFA resulted in 21 fewer nonfatal myocardial infarctions per 1000 participants over 5 years, compared with 2 fewer events per 1000 in low-risk patients. The certainty of evidence for PUFA replacement was rated as moderate, while evidence for overall SFA reduction was rated as low to moderate certainty.
The researchers also assessed carbohydrate replacement, analyzing data from 4 trials, though these analyses showed no credible subgroup effects when comparing carbohydrate to other macronutrient replacements. Notably, data were insufficient to evaluate replacement with monounsaturated fatty acids or protein, representing a significant evidence gap.
Cholesterol Effects and Compliance
At last reported follow-up (mean, 29 months), 14 randomized controlled trials with 9795 participants demonstrated that lower SFA intake probably resulted in an important reduction in total cholesterol (mean difference, −0.34 mmol/L; moderate-certainty evidence). The pooled estimate of −13 mg/dL surpassed the researchers' minimally important difference threshold of 0.26 mmol/L (10 mg/dL).
For low-density lipoprotein cholesterol (LDL-C), 6 randomized controlled trials (3630 participants) with mean follow-up of 24 months showed that lower SFA intake resulted in an important reduction (mean difference, −0.15 mmol/L; high-certainty evidence), equivalent to −6 mg/dL, exceeding the 0.10 mmol/L (4 mg/dL) minimally important difference threshold.
The observed LDL-C reduction of 0.15 mmol/L (6 mg/dL) equates to approximately 3% energy from SFA being substituted by cis-PUFA or approximately 5% energy from SFA being substituted by carbohydrate, based on metabolic ward trials. This modest effect suggests that included trials achieved only moderate success in reducing SFA intakes. Meta-regression analyses demonstrated that greater LDL-C reduction was significantly associated with greater benefit for nonfatal myocardial infarction, though this was based on only 3 trials.
Risk of Bias and Study Limitations
Of the 17 eligible trials, researchers judged 4 to be at low risk of bias and 13 at high risk of bias. However, subgroup analyses found no credible subgroup effects when considering pooled estimates for trials at low vs high risk of bias. All trials employed random assignment judged to be at low risk of bias.
The review identified substantial heterogeneity in dietary interventions. Fifteen trials advised participants to modify dietary fat intake, with 4 trials (24%) providing supplements such as oils or other foods. Two trials supplied all food within residential institutions. The Women's Health Initiative trial, which contributed the largest number of participants and observed events, aimed to reduce total fat intake and increase fruits, vegetables, and grains without specific guidance on reducing SFA intake. Food frequency questionnaire-derived estimates indicated this intervention reduced SFA intakes by 4% energy at year 1 and 3% energy at year 6, while simultaneously reducing monounsaturated fatty acids and PUFAs.
"Efforts to assess the effect of SFA lowering and replacement on disease end points have largely been relegated to systematic reviews of observational analyses, wherein SFA intakes have no independent association with total coronary heart disease, ischemic stroke, CVD mortality, and all-cause mortality. Such nonrandomized analyses are subject to residual confounding; bias from memory-based dietary recalls; and overadjustment for causal intermediates in analyses, including LDL-C," wrote senior author Bradley C. Johnston, PhD, of Texas A&M University, and colleagues.
Historical Context and Contemporary Relevance
The entire literature base spanned trials published between 1965 and 2006, preceding contemporary understanding of omega-3 fatty acids, trans fatty acids, and modern cardiovascular therapies including statins, ezetimibe, and proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors. Only 2 trials explicitly reported allowing statin use: the Lyon Diet Heart Study (8% of participants) and the Women's Health Initiative (12% at baseline). Seven trials were conducted before 1980, when statins had not yet become common therapy.
Many trials lacked comprehensive dietary intake assessments at baseline and throughout the intervention, precluding detailed analyses of the degree of SFA reduction and co-nutrient changes including PUFA, omega-3, or omega-6 fatty acids. This limitation is particularly relevant given that trials with estimated concomitant reductions in trans fatty acid intake showed possible cardiovascular mortality reduction compared with trials without known trans fatty acid differences.
In an accompanying editorial, Ramon Estruch, MD, of the University of Barcelona, and Rosa M. Lamuela-Raventós, PhD, of the University of Barcelona, contextualized the findings: "For decades, dietary intake of SFAs has been regarded as harmful to cardiometabolic health. Almost simultaneous with Dylan's song, the so-called 'diet-heart hypothesis' emerged with force and spread the concept that SFA causes heart disease by raising serum cholesterol. This hypothesis, despite being based on weak evidence of association not causation, remains today with nearly all dietary guidelines worldwide recommending SFA intake of less than 10%."
Implications for Guidelines and Practice
The researchers established minimally important difference thresholds of 5 fewer cases per 1000 followed over 5 years for fatal outcomes and 10 fewer cases per 1000 for nonfatal outcomes. These thresholds informed their risk-stratified conclusions that reducing SFA intake has "little or no benefit" for low-risk patients but "probably" or "may" have important benefits for high-risk patients, depending on the specific outcome and certainty of evidence.
Current dietary guidelines generally advocate for saturated fat intake of less than 10% of total daily caloric intake, with some guidelines suggesting limits around 5% to 6%. The Minnesota Coronary Survey, which tested an 18% vs 9% saturated fat diet in 9057 participants followed for 5 years, identified no statistically significant reduction in cardiovascular events, cardiovascular death, or all-cause mortality.
"The findings of this review align with the current emerging recognition that dietary SFA per se are unlikely deleterious for cardiometabolic health for general population but may be deleterious for people at high risk for CV events," the editorialists noted. They emphasized that the impact of SFA on human metabolism varies markedly according to carbon chain length.
The study authors reported no conflicts of interest.