Epidemiology beyond its limits

Abstract

In 1995, journalist Gary Taubes published an article in Science titled "Epidemiology faces its limits," which questioned the utility of nonrandomized epidemiologic research and has since been cited more than 1000 times. He highlighted numerous examples of research topics he viewed as having questionable merit. Studies have since accumulated for these associations. We systematically evaluated current evidence of 53 example associations discussed in the article. Approximately one-quarter of those presented as doubtful are now widely viewed as causal based on current evaluations of the public health consensus. They include associations between alcohol consumption and breast cancer, residential radon exposure and lung cancer, and the use of tanning devices and melanoma. This history should inform current debates about the reproducibility of epidemiologic research results.

Fig. 1.. Selected meta-analytic estimates were strongest for associations deemed causal by both Taubes (1995) and our group (2021) and were more modest for associations deemed indeterminate by Taubes (1995) and causal by our group (2021).

Meta-analytic estimates by causal evaluation (Taubes’ 1995 Assessment/Updated 2021 Assessment) for (A) causal/causal, (B) indeterminate/causal, (C) indeterminate/indeterminate, (D) indeterminate/noncausal, and (E) noncausal/noncausal. Arrows indicate that estimate is beyond the y-axis range. Points indicate meta-analytic estimates, and segment end caps indicate minimum and maximum values of the 95% confidence intervals (CIs). Key: A, human papillomavirus and cancer; B, ionizing radiation and cancer; C, hepatitis and cancer; D, smoking and lung cancer; E, cigarette smoke and cancer; F, sunlight and skin cancer; G, alcohol and cancer; H, asbestos and cancer; I, occupational steel (coke-oven) exposure and lung cancer; J, early childbirth (maternal age) and breast cancer; K, obesity and esophageal cancer; L, cigarette smoking and pancreatic cancer; M, lengthy occupational dioxin (2,3,7,8-tetrachlorodibenzodioxin) and cancer; N, alcohol and breast cancer; O, residential radon and lung cancer; P, eating red meat and colon cancer; Q, birthweight and breast cancer; R, oral contraceptive use and breast cancer; S, sun lamp use and melanoma; T, eating processed meat and colon cancer; U, breastfeeding and brain cancer/leukemia in children; V, high-alcohol mouthwash and mouth cancer; W, electromagnetic fields and brain cancer/leukemia in children; X, traffic density and brain cancer/leukemia in children; Y, high-cholesterol diet and rectal cancer; Z, douching and cervical cancer; AA, occupational stress and colorectal cancer; BB, smoking and fatal breast cancer; CC, hair dyes and myeloma; DD, drinking chlorinated tap water and bladder cancer; EE, eating yogurt and ovarian cancer; FF, hair dyes and lymphoma; GG, electromagnetic fields and brain cancer; HH, hair dyes and leukemia; II, smoking and breast cancer; JJ, diet high in saturated fat and lung cancer (among nonsmokers); KK, electromagnetic fields and leukemia; LL, fat intake and breast cancer; MM, maternal smoking and brain cancer/leukemia in children; NN, eating red meat and breast cancer; OO, electromagnetic fields and breast cancer; PP, coffee and heart disease; QQ, olive oil and breast cancer; RR, coffee and pancreatic cancer; SS, vasectomy and prostate cancer; TT, breast self-examination and breast cancer mortality; UU, abortion and breast cancer; VV, dichlorodiphenyltrichloroethane and breast cancer; WW, saccharine and bladder cancer.