Replacing saturated fat by unsaturated fat decreases CVD risk. Or maybe not!

Background: Recently, a few systematic reviews were published in which the effect of replacing saturated fats by polyunsaturated fats on coronary heart disease (CHD) was examined (1, 2). Both articles showed that this might lead to a small reduced risk of CHD, but all-cause mortality rates did not change by modifying dietary fats. The authors did not conclude that lowering intakes of saturated fat may reduce CHD risk. And the authors from one of the articles showed that results were strongly biased by potential confounders (2).
Last year another systematic review on this topic was published (3). Based on their results, this was the first time that authors adviced to lower intakes of saturated fat in order to reduce risk of cardiovascular diseases (CVD):

The findings are suggestive of a small but potentially important reduction in cardiovascular risk on modification of dietary fat, but not reduction of total fat, in longer trials. Lifestyle advice to all those at risk of cardiovascular disease and to lower risk population groups, should continue to include permanent reduction of dietary saturated fat and partial replacement by unsaturates. The ideal type of unsaturated fat is unclear.

Let's take a look if this advice is valid based on the study results.

Study results:
Table 1/Comparison 1 shows the primary results found by the authors. Effects on total mortality, CVD mortality and CVD events were published. Three types of studies were included:

• Modified fat intake (replacing saturated fat by unsaturated fats).
• Reduced fat intake.
• Both reduced fat intake and modified fat intake.

As stated before, 3 effects were examined. The top effects in table 1 are based on all studies combined. Below that, results can be seen from studies in which effects from fat modification only were examined. When all types of studies were combined, a small significant protective effect (- 14% risk) against CVD risk was found. But fat modification did not lead to a significantly lower risk of CVD: RR = 0.82 (0.66-1.02). No significant effects on CVD mortality, or all-cause mortality were found. Subjects in the fat modification group did not live a day longer than subjects in the control group.

Effects from fat modification on CVD:
Replacing saturated fats by unsaturated fats apparently lead to a non-significant 18% lower risk of CVD events, but effects on CVD mortality were even weaker (- 8%). This suggests that risk of nonfatal CVD events might slightly decrease. But what specific type of events was lowered by fat modification?
If we take a look at other results published by the authors, we see no significant effects on myocardial infarctions (- 9% risk), non-fatal myocardial infarctions (- 14% risk), or stroke (- 30% risk).
It is noteworthy to mention that risk of total, or non-fatal myocardial infarctions did not change by fat modification, since replacing butter by margarine is supposed to lead to a lower risk of this disease.
The authors define cardiovascular events as follows:

Combined cardiovascular events included any of the following data available from a trial: cardiovascular deaths, cardiovascular morbidity (non-fatal myocardial infarction, angina, stroke, heart failure, peripheral vascular events, atrial fibrillation) and unplanned cardiovascular interventions (coronary artery bypass surgery or angioplasty).

So which type of nonfatal CVD was reduced by fat modification? Did the experimental groups have a smaller risk of angina?

The article also shows how the non-significant protective effect against CVD mortality (- 8% risk) was possibly compensated to lead to absolutely no protective effect against all-cause mortality (+ 2% risk): Fat modification lead to a non-significantly increased risk of cancer death (+ 46% risk).

Did the studies included only examine the effect of replacing saturated fat by unsaturated fat?
At least part of the studies should never have been included in the systematic review. In addition to a reduced intake of saturated fats and dietary cholesterol, also the intakes of trans fat were reduced in the experimental groups from the studies included in the review. In some of the studies, the dietary intervention included a complete mediterranean dietary pattern. Bias from other (non)dietary factors is discussed on my other site (4) and in an article of mine that was published recently (5).

Can other types of bias be responsible for the results found by the authors?
The possibility of publication bias was examined by including a socalled "funnel plot". This plot can be seen below. The squares/plots represent the results from the individual studies. The "1" on the x-axis represents zero effect. Plots on the left of the "1" value represent a "protective effect" and plots on the right of the "1" value represent an increased risk of CVD events. The larger trials are plotted higher in the figure. The smaller studies are shown lower in the figure. We see that most larger trials are plotted around the "1"value indicating little or no effect on CVD events. But we can also see that most of the smaller trials are plotted left from the "1" value, showing protective effects. This suggests that small studies showing that the dietary intervention lead to increased risk of CVD events may not have been published. This would indicate that the small protective effect is actually an overestimation of the true effect. The authors conclude the same:

A funnel plot did not suggest severe small study (or publication) bias, but it is likely that a few small studies with more cardiovascular events in the intervention groups may be missing from the review (Figure 4).

Was there a specific effect from saturated fat?
Because intakes of both saturated- and unsatured fats changed in the dietary interventions, it is impossible to assess which change is responsible for the effects found. To examine this, the researchers made an additional analysis, shown in the table below. None of the effects was significant. None of the different types of fat could be held responsible for the effects found.

Conclusion: Dietary interventions in the systematic review by Hooper et al included replacing saturated fat by unsaturated fats, sometimes accompanied by a large amount of other (non)dietary changes which were not mentioned. These interventions may lead to a slightly lower risk of nonfatal CVD events, while risks of fatal CVD and all-cause mortality do not seem to change. It is not clear which types of nonfatal CVD events may be influenced beneficially by these interventions, since risk of non-fatal myocardial infarctions did not significantly change. Saturated fat intake could not be helt responsible for the effects found. It is possible that publication bias lead to an overestimation of the true effects of the dietary interventions on CVD events.


References:
1) Mozaffarian D. Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: a systematic review and meta-analysis of randomized controlled trials. PLoS Med. 2010 Mar 23;7(3):e1000252. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2843598/?tool=pubmed
2) Ramsden CE. n-6 fatty acid-specific and mixed polyunsaturate dietary interventions have different effects on CHD risk: a meta-analysis of randomised controlled trials. http://www.ncbi.nlm.nih.gov/pubmed/21118617
3) Hooper L. Reduced or modified dietary fat for preventing cardiovascular disease. http://www.ncbi.nlm.nih.gov/pubmed/22592684
4) Hoenselaar R. Randomized trials substituting polyunsaturated fat for saturated fat and their effect on coronary heart disease (CHD). A closer look. http://canceranddiet.nl/cardiovascular_disease/randomized-saturated-polyunsaturated-fat.html
5) Hoenselaar R. Further response from Hoenselaar. Br J Nutr. 2012 Mar 1:1-4. [Epub ahead of print]. http://www.ncbi.nlm.nih.gov/pubmed/22377374

Posted in: cardiovascular disease