Epidemiology beyond its limits

Area of Science:

• Epidemiology
• Public Health
• Medical Research

Background:

• Gary Taubes' 1995 article in "Science" critically examined the value of nonrandomized epidemiologic studies.
• The article highlighted numerous research topics with questionable associations, sparking ongoing debate.
• Over the years, substantial research has accumulated on these specific associations.

Purpose of the Study:

• To systematically evaluate the current evidence for 53 associations mentioned in Taubes' 1995 article.
• To assess whether previously doubted associations are now supported by scientific consensus.
• To provide historical context for current discussions on the reproducibility of epidemiologic findings.

Main Methods:

• Systematic evaluation of existing scientific literature.
• Analysis of 53 specific epidemiologic associations previously questioned.
• Assessment based on current public health consensus and evidence accumulation.

Main Results:

• Approximately 25% of the initially doubted associations are now widely accepted as causal.
• Examples of re-evaluated associations include alcohol and breast cancer, residential radon and lung cancer, and tanning devices and melanoma.
• This re-evaluation demonstrates the evolving nature of scientific consensus in epidemiology.

Conclusions:

• The historical trajectory of these 53 associations supports the utility and progress of epidemiologic research.
• Findings suggest that initial skepticism can be overcome with accumulated evidence and rigorous evaluation.
• This historical perspective is crucial for informing contemporary debates on the reproducibility and reliability of epidemiologic research.