Re: Intake of individual saturated fatty acids and risk of coronary heart disease in US men and women: two prospective longitudinal cohort studies
This is the third paper reaching the same conclusion by this research group in the last 18 months (1-3), all based on the same data. Despite some strengths (such as large sample size, length of follow-up and the use of hard end-points) there are some important limitations.
A major issue is the conflict of interest. Three of the authors are employees of, and one received funding from, a food production company whose portfolio includes products high in polyunsaturated fats (PUFA); including margarine, which it is the world’s biggest producer of. Half of the authors therefore were linked to a company that benefits financially from people choosing PUFA over SFA, which the study concludes we should do. Although this in itself is not evidence of issues with the research there is an unacceptable risk of bias, with previous research demonstrating that industry funded studies produce more favourable results than independent research (4, 5). Indeed, an independent meta-analysis of observational studies does not support their findings (6).
The usual pattern of research is for randomised controlled trials (RCTs) to be performed to confirm or deny trends identified in observational studies. In the case of SFAs there have been a number of RCTs performed; and a number of systematic reviews and meta-analyses pooling their results (e.g. 6, 7-11). This raises the question, why are observational studies still being produced on the subject? It is notable that the evidence from meta-analyses does not support the Harvard Group’s conclusions. For example, a recent Cochrane review found no statistical difference in all-cause or cardiovascular disease mortality in participants with higher SFA intake (12). This review did conclude reducing SFA intake corresponds to a 17% reduction in cardiovascular events, but there was high-heterogeneity in this analysis and the difference became non-significant when only studies that significantly reduced SFA intake were included in sensitivity analyses.
The limitations of observational research are well documented, for example it is difficult to isolate the effects of a single factor due to a vast number of potentially confounding variables. Observational studies are also unable to prove cause and effect (correlation doesn’t equal causation). Any attempt to infer whether there is a causal relationship can be judged against Bradford Hill’s criteria (13). This study fails to achieve a number of these: including that it is non-experimental, the assumption of temporality is not met (the study does not demonstrate an increase in SFA occurs before any increase in factors leading to mortality or CHD), there is little evidence of a dose response (for example there is no consistent increase in hazard ratios across quintiles of SFA consumption) and the size of the effect is relatively small (reduction in risk is only 6-8% when SFA was replaced with MUFA, PUFA or wholegrains).
The use of food frequency questionnaires, the weaknesses of which are well understood, is another limitation; whilst the attempt at validating their use is also flawed. Correlations are not a suitable statistical method for assessing agreement, but rather alternatives such as Bland-Altman plots would be more appropriate (14). The use of isocaloric replacement of SFA for other macronutrients is another flaw, as it fails to consider differences in individual’s responses to food (e.g. 15, 16) and how changes may impact upon overall dietary patterns. It should also be noted that this statistical analysis is NOT equal to performing an intervention study where foods were actually substituted, thus the conclusions over-extend the findings.
The authors acknowledge that the high correlation for intake of different SFA chain lengths reduced the ability to consider their individual effects. This is an important limitation as the length of the SFA chain may have an important influence. Short chain SFA have been suggested to be beneficial for the gut microbiome (17) whilst medium chain fatty acids are metabolised rapidly in the liver and have been associated with less body weight gain (18). Longer-chain SFA (>20:0) have been shown to have an inverse association with Type 2 diabetes, for example, perhaps due to having a different metabolic pathway (19).
There is also evidence for a differential effect of odd and even chains. An inverse association has been observed between odd chain SFA (15:0 and 17:0) and Type 2 diabetes (19). This is opposed to the positive association seen with even chain SFA (14:0, 16:0, and 18:0). It should be noted these associations were for plasma rather than dietary SFA. The underlying cause of increased circulating even chain SFA has been shown to be de novo lipogenesis in the liver, thus this association may better reflect the impact of dietary carbohydrate rather than exogenous SFA (19). Nevertheless the differential effects observed mean the failure to consider odd chain SFA’s is another limitation. More research is needed to fully elucidate the effects of different chain lengths, but there is enough evidence to suggest this plays a part. As such it is incorrect to conclude that all SFA is harmful based on studies that have not adequately differentiated between different forms.
Failure to consider the source of SFA is another limitation. There are a wide variety of foods containing SFA, each of which includes a variety of fatty acids (no food contains a single form of fat, for example the fat content of extra virgin olive oil is ~15% SFA and ~10% PUFA alongside the MUFA it’s renowned for). In each of these the SFA is also accompanied by different combinations of macronutrients, vitamins and minerals. To assume that each will have a comparable impact on health is erroneous.
There are limitations to what studies are practical. As such it is important to be pragmatic in seeking the best available evidence. However, in an area where there are multiple randomised controlled trials, meta-analyses, and an abundance of observational research the addition of another epidemiological paper, particularly a highly flawed one, adds no value. High quality evidence from RCT’s and meta-analyses does not support the assertion that SFA intake should be restricted, and this most recent addition does nothing but unnecessarily muddy the waters.
1. Zong G, Li Y, Wanders AJ, Alssema M, Zock PL, Willett WC, et al. Intake of individual saturated fatty acids and risk of coronary heart disease in US men and women: two prospective longitudinal cohort studies. Bmj. 2016:i5796.
2. Wang DD, Li Y, Chiuve SE, et al. Association of specific dietary fats with total and cause-specific mortality. JAMA Internal Medicine. 2016.
3. Li Y, Hruby A, Bernstein AM, Ley SH, Wang DD, Chiuve SE, et al. Saturated Fats Compared With Unsaturated Fats and Sources of Carbohydrates in Relation to Risk of Coronary Heart Disease: A Prospective Cohort Study. J Am Coll Cardiol. 2015;66(14):1538-48.
4. Lexchin J, Bero LA, Djulbegovic B, Clark O. Pharmaceutical industry sponsorship and research outcome and quality: systematic review. BMJ : British Medical Journal. 2003;326(7400):1167-.
5. Chartres N, Fabbri A, Bero LA. Association of Industry Sponsorship With Outcomes of Nutrition Studies: A Systemtic Review and Meta-analysis. JAMA Internal Medicine. 2016.
6. Chowdhury R, Warnakula S, Kunutsor S, Crowe F, Ward HA, Johnson L, et al. Association of Dietary, Circulating, and Supplement Fatty Acids With Coronary Risk: A Systematic Review and Meta-analysis. Ann Intern Med. 2014;160(6):398-406.
7. Mente A, de Koning L, Shannon HS, Anand SS. A systematic review of the evidence supporting a causal link between dietary factors and coronary heart disease. Archives of internal medicine. 2009;169(7):659-69.
8. Siri-Tarino PW, Sun Q, Hu FB, Krauss RM. Saturated fat, carbohydrate, and cardiovascular disease. The American Journal of Clinical Nutrition. 2010;91(3):502-9.
9. Schwingshackl L, Hoffmann G. Comparison of long-term low-fat versus high-fat diets on blood lipids: a systematic review and meta-analysis. Proceedings of the Nutrition Society. 2012;71(OCE3):doi:10.1017/S0029665112003114.
10. Schwingshackl L, Hoffmann G. Comparison of the long-term effects of high-fat v. low-fat diet consumption on cardiometabolic risk factors in subjects with abnormal glucose metabolism: a systematic review and meta-analysis. The British journal of nutrition. 2014;111(12):2047-58.
11. Harcombe Z, Baker JS, DiNicolantonio JJ, Grace F, Davies B. Evidence from randomised controlled trials does not support current dietary fat guidelines: a systematic review and meta-analysis. Open Heart. 2016;3(2).
12. Hooper L, Martin N, Abdelhamid A, Davey Smith G. Reduction in saturated fat intake for cardiovascular disease. The Cochrane database of systematic reviews. 2015;6:CD011737.
13. Bradford Hill A. The Environment and Disease: Association or Causation? Proceedings of the Royal Society of Medicine. 1965;58(5):295-300.
14. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. The Lancet. 1986;8476:307-10.
15. Zeevi D, Korem T, Zmora N, Israeli D, Rothschild D, Weinberger A, et al. Personalized Nutrition by Prediction of Glycemic Responses. Cell. 2015;163(5):1079-94.
16. Matthan NR, Ausman LM, Meng H, Tighiouart H, Lichtenstein AH. Estimating the reliability of glycemic index values and potential sources of methodological and biological variability. Am J Clin Nutr. 2016;104(4):1004-13.
17. Bugaut M. Occurrence, absorption and metabolism of short chain fatty acids in the digestive tract of mammals. Comparative biochemistry and physiology B, Comparative biochemistry. 1987;86(3):439-72.
18. St-Onge MP, Jones PJ. Physiological effects of medium-chain triglycerides: potential agents in the prevention of obesity. The Journal of nutrition. 2002;132(3):329-32.
19. Forouhi NG, Koulman A, Sharp SJ, Imamura F, Kröger J, Schulze MB, et al. Differences in the prospective association between individual plasma phospholipid saturated fatty acids and incident type 2 diabetes: the EPIC-InterAct case-cohort study. The Lancet Diabetes & Endocrinology. 2014;Oct;2(10):810-8. doi: 10.1016/S2213-8587(14)70146-9. Epub 2014 Aug 5.
Competing interests: Trudi Deakin is the author of the X-PERT Health programme and Eat Fat: Step by step guide to low carb living. X-PERT Health is a registered charity, and all proceeds from sales go to the charity. Sean Wheatley, Trudi Reeves and Matthew Whitaker are all employees of X-PERT Health.