Peter's take on studies that claim that moderate carbohydrate intake may be best for health (AMA #3)

Partially SupportedWhen evaluating risk, it is essential to consider absolute risk, relative risk, and the time period studied relative to the natural course of the illness.

Not SupportedMeta-analyses on moderate carbohydrate intake are flawed due to epidemiological issues like confounding and low hazard ratios.

Not SupportedHazard ratios in epidemiological studies can be misleading when the absolute risk is low.

Not SupportedA 24% relative increase in breast cancer risk from hormone replacement therapy translates to a 0.1% absolute increase.