Reader Comments

Post a new comment on this article

Invalid Information Published in Figure 3

Posted by E_Roehm_MD on 13 May 2010 at 11:01 GMT

Dear Sirs:

Dr. Mozaffarian et al (1) published a meta-analysis of eight trials analyzing the consumption of polyunsaturated fat (PUFA) intake in the setting of a reduced saturated fat (SFA) intake in regards to the incidence of coronary heart disease.

Unfortunately, in this paper by Mozaffarian et al, as part of the discussion regarding monounsaturated fat intake (MUFA), information from a recent pooled analysis of 11 observational studies by Jakobsen et al (2) is published as Figure 3. The publication of Figure 3 in the article by Dr. Mozaffarian et al results in the second publication of the erroneous data from the prior recent study by Jakobsen et al. This same data also plays a role in the discussion section of the Mozzafarian article (1) in comments regarding MUFA intake.

Specifically, in Figure 3, information from the pooled analysis by Jakobsen et al (2) is shown purportedly giving data about the effects of monounsaturated fat (MUFA) replacing saturated fat (SFA) in regards to the relative risk of coronary heart disease. There actually was no valid information on MUFA replacing SFA in the Jakobsen study, which was a pooled analysis of 11 cohort studies (2) .

The source of the monounsaturated fat (MUFA) intake in the populations in the studies analyzed by Jakobsen et al (2) was primarily the result of meat and dairy fat intake. This means that MUFA intake was invariably linked to saturated fat (SFA) intake.

Ground beef, for example, contains approximately 5g of saturated fat (SFA) and 6g of monounsaturated fat (MUFA) for every 200 calories of beef (3). It is not possible for any statistical technique to create valid information for groups of individuals who are eating ground beef, in regards to the effect of monounsaturated fat replacing saturated fat intake. A statistical technique can not transform information on monounsaturated fat (MUFA) intake from food sources where saturated fat and monounsaturated fat are invariably linked, and magically create valid information on what happens when monounsaturated fat (MUFA) replaces saturated fat (SFA) for each 5% increase in MUFA intake and corresponding 5% decrease in SFA intake.

The incorrect statistical analysis by Jakobsen resulted in false information that was not present in the data itself. The repeat publication of this invalid information in Figure 3 in the Mozaffarian paper is an implied validation of this erroneous information. It is requested that Mozaffarian et al address the information republished in their article from the recent Jakobsen study that through an erroneous statistical analysis purported to show the effects of MUFA intake replacing SFA intake. If this is not addressed, the publication of the information in Figure 3 is an implied validation of that information, which would be an unfortunate occurrence.

Eric Roehm, M.D.

1. Mozaffarian D, Micha R, Wallace S (2010) 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 Medicine (2010) 7(3): 1-10 e1000252.

2. Jakobsen MU, O Reilly EJ, Heitmann BL, Pereira MA, Balter K, Fraser GE, et al. (2009) Major type of dietary fat and risk of coronary heart disease: a pooled analysis of 11 cohort studies. Am J Clin Nutr 89: 1425-1432.

3. U.S. Department of Agriculture, Agricultural Research Service. 2009. USDA National Nutrient Database for Standard Reference, Release 22. Nutrient Data Laboratory Home Page, (cited 07 April 2010).

No competing interests declared.

RE: Invalid Information Published in Figure 3

D_Mozaffarian replied to E_Roehm_MD on 18 May 2010 at 19:13 GMT

Evaluating independent effects of correlated variables is challenging, but not impossible. While Dr. Roehms's concerns that intercorrelation between monounsaturated fat and saturated fat could limit their independent evaluation, the facts remain that (a) monounsaturated fat is not convincingly associated with lower CHD risk in prospective cohort studies, and that (b) no randomized controlled trials have tested the independent effects of increased monounsaturated fat consumption on CHD events. Additionally, primate and experimental studies raise concerns that monounsaturated fats may not produce benefits expected from changes in lipid concentrations alone. [ref: Degirolamo C, Shelness GS, Rudel LL. LDL cholesteryl oleate as a predictor for atherosclerosis: evidence from human and animal studies on dietary fat. J Lipid Res 50 Suppl:S434-439, 2009.] Thus, our discussion and conclusions remain valid. It has also been argued that health effects of monounsaturated fats may differ based on their source (e.g., animal vs. vegetable). If this were true, then these health effects would not be due to the monounsaturated fat itself, but to other factors in the animal or vegetable foods/oils.

No competing interests declared.