A retrospective cohort study of 329 US adults with obesity enrolled in a commercial digital food-as-medicine program found meaningful but statistically inflated blood pressure reductions when using raw pre/post comparisons. Among Stage-2 hypertensive members (≥135/85 mmHg), unadjusted systolic blood pressure (SBP) fell 10.17 mmHg at 6 months — but after correcting for regression to the mean (RTM) using an at-goal internal control group, the true program-attributable reduction shrank to 5.50 mmHg (95% CI −8.50 to −2.50). RTM alone explained roughly 46% of the apparent improvement in Stage-2 members and 24% in Stage-1. Both corrected estimates remained statistically significant.
The finding matters for two reasons. First, a genuine 5–6 mmHg SBP reduction is clinically meaningful — epidemiological data consistently link a 5 mmHg lower SBP to roughly 10% lower cardiovascular mortality risk. Second, the methodological message is arguably more important: most published digital health trials enroll based on elevated baselines and never account for RTM, likely overstating program efficacy by 25–50%. The study's internal negative-control design is a practical model for an otherwise difficult problem. Limitations include the single-arm design (no randomized control group), a commercially motivated population, and modest 6-month follow-up numbers. This is a preprint posted on medRxiv and has not yet been peer-reviewed; findings and effect sizes may change. Incremental but methodologically instructive work for anyone evaluating dietary or digital interventions on blood pressure.