The American Journal of Clinical Nutrition, Volume 92, Issue 2, August 2010, Pages 458–459, https://doi.org/10.3945/ajcn.2010.29504
Issue Section: Letters to the Editor
Dear Sir:
The recent article by Siri-Tarino et al (1), which reported on a meta-analysis of prospective cohort studies evaluating the association of saturated fat with coronary heart disease (CHD), stroke, and cardiovascular disease (CVD) observed that “there is no significant evidence for concluding that dietary saturated fat is associated with an increased risk of CHD or CVD.” This finding has generated some interest in the media (2, 3). However, we believe that the interpretations of the results presented in this article are overstated and could be a result of flaws in the methodologic design of the study.
The meta-analysis involves data from 16 studies that evaluate the effect of saturated fat intake on CHD incidence or mortality and from 8 studies that evaluate the effect of saturated fat intake on stroke incidence or mortality. The results for CVD include any events for either CHD or stroke. The authors state that “wherever possible, risk estimates from the most fully adjusted models were used in the estimation of the pooled [relative risks].” It is well established that saturated fat intake is associated with increased concentration of serum cholesterol (4), and that serum cholesterol concentrations are associated with CHD and CVD (5). Therefore, serum cholesterol concentrations lie on the causal chain between saturated fat intake and CHD and CVD and to adjust for serum cholesterol concentrations in a meta-analysis will obscure the effect of saturated fat intake on these health outcomes.(以下略)
中間変数(媒介変数、中間変量とも)を加えて重回帰分析を行った結果、因果推定が低い方に歪められて算出されたと思われる例。
上記の投稿は、公表された論文に対して、その問題点を的確に指摘したものである。
対象となる論文にこのような不備があると、系統的レビューやメタ解析にも、その不備が反映されてしまうので、解釈においては、充分な注意が必要である。
しかし、こんな奇妙な分析結果が出たら、(プロの研究者なら)もう一度、データや分析方法をチェックしそうなものだけど､､､(←門外漢の無責任な野次なので、あまり気にしないで下さい)
