Nutrients are an essential part of building and maintaining optimal health. Certain nutrient exposure has been shown to be associated with many important health outcomes, although there is variability among studies. Despite the scientific efforts of many, it is unclear why some well-hypothesised nutrients lack sufficient evidence for clear association with health outcomes. One potential reason for conflicting results is that certain subgroups of patients benefit or are harmed more by adequate or inadequate exposure to certain nutrients. These subgroup-specific effects have historically not been studied, or if they are, it is often in a one-off type of approach where the investigator believes that a subgroup effect could exist based on limited previous data. In the era of big data, improvements can be made in efforts to generate new hypotheses for subgroups of patients and recommendations for precision nutrition can be made. In the present paper, we present a strategy for exploring subgroup-specific effects in nutrient-related studies. This data-driven method can be useful in secondarily exploring which subgroups are harmed/helped most by inadequate/adequate nutrient exposure and could suggest target groups for future clinical trials to test the identified hypotheses. We then present an example study utilizing the National Health and Nutrition Examination Survey (NHANES) data from the years 2001–2006. In this example, a limited selection of nutrients is protective in subgroups of participants with diabetes on their self-reported number of poor mental health days.