Research Designs for Individualized Health
Developing evidence for precision medicine, personalized nutrition, and individual health maintenance requires new research study designs that go beyond comparing the responses of a treated/intervention group versus a control group. A prime example of one such design was reported in Nature Medicine (https://tinyurl.com/y38gckgu) and based on longitudinal analysis over ~2.8 years of 109 individuals who were deeply phenotyped by measuring clinical (including oral glucose tolerance test), transcriptomic, proteomic (with immunome), metabolomic, microbiome, activity, stress, vascular ultrasound, cardiopulmonary exercise, echocardiographic, diet, and whole genome sequencing. Data were analyzed individually and by subgroups by standard methods including correlation network analysis and linear mixed models.
Among the many outcomes were the discovery of (i) 67 clinically actionable health issues, (ii) multiple molecular pathways associated with metabolic, cardiovascular, and oncological pathophysiologies, (iii) predictive models for insulin resistance, and (iv) changes in diet and activity for many of individuals who participated in the research. Translating such approaches to health care face the obvious challenge of cost but nevertheless demonstrate the use of biological (i.e., micro) scale big data to individualize health trajectories.