Nested randomized trials in large cohorts and biobanks: studying the health effects of lifestyle factors.

Most diseases are likely to result largely from the interplay of lifestyle and genetic factors. However, both observational studies and randomized trials have faced major limitations in trying to address the impact of lifestyle on health. As large cohorts and biobanks are being developed, we need to find novel, efficient ways to address the effects of lifestyle interventions. We propose that this could be done using multiple lifestyle factorial experimental designs that combine characteristics of randomized trials and epidemiologic studies. Randomized trials of simple lifestyle interventions can be nested within large cohorts linked to reliable registries of outcomes. Participants can choose from a long list of simple lifestyle randomization options and many interventions may be tested concurrently with factorial randomization. Participants can tailor their own personal trial choosing several items among long laundry lists of randomization options. Participants are citizen-scientists rather than passive subjects and this may be attractive in modern societies of health-conscious people. These trials can use the existing machinery of the cohort for data collection and outcome linkage at no or minimal additional cost. We discuss a number of issues on the implementation of multiple lifestyle factorial experimental designs, as compared with the usual observational studies and randomized trials. These include participation, the number of allowed randomizations per participant, compliance/adherence, power, false-negatives, false-positives, composite lifestyle effects, selection of outcomes, follow-up and monitoring, masking and allocation concealment, age of participants, confounding, and cost. The aim should be to combine carefully the strengths of both observational epidemiology and randomized research without compounding their limitations.