Rebekah J. Walker

Objective: This study investigated food access as a potential pathway between structural racism and the prevalence of cardiometabolic disease. Methods: A total of 11,457 census tracts across 201 counties, within 38 states, were combined for a final analytic sample. Direct and indirect relationships between redlining and each cardiometabolic disease via food access were investigated using structural equation modeling run in Stata v17, controlling for the population of each census tract. Results: Mean prevalence of diabetes was 11.8%, high blood pressure 31.9%, coronary heart disease 6.0%, and obesity 31.8%. Redlining (0.22, p < 0.001) and lower food access (−0.11, p < 0.001) were directly associated with higher prevalence of diabetes. Redlining (0.10, p < 0.001) and lower food access (−0.10, p < 0.001) were directly associated with higher prevalence of high blood pressure. Redlining (0.12, p < 0.001) and lower food access (−0.04, p < 0.001) were directly associated with higher prevalence of coronary heart disease. Redlining (0.18, p < 0.001) and lower food access (−0.12, p < 0.001) were directly associated with higher prevalence of obesity. Redlining was also significantly indirectly associated with each condition via food access. Conclusions: Structural racism was associated with higher cardiometabolic disease prevalence via decreased food access. Interventions to improve access to healthy food at the community level may be needed to address the impact of structural racism on health.

Background: Artificial intelligence (AI)–driven lifestyle interventions are increasingly used for reducing diabetes risk. However, whether they produce clinical benefit beyond standard Diabetes Prevention Programs (DPPs) and whether effects vary by AI type, population, or delivery setting, remains uncertain. Objective: This study aimed to evaluate the effects of AI-driven lifestyle interventions on clinical, behavioral, and engagement outcomes among people with prediabetes or elevated diabetes risk and to assess whether effects differed by AI type, population, or delivery setting. Methods: PubMed, Embase, CINAHL, Web of Science, IEEE Xplore, ClinicalTrials.gov, and citation searches were completed on August 3, 2026, for English-language reports published from 2015. Eligible studies evaluated AI-driven digital lifestyle interventions among people with prediabetes or elevated diabetes risks, or healthy participants in feasibility studies. One reviewer (OIE) screened records and extracted data; a second reviewer (TN) independently verified exclusions and extraction. Study quality was assessed with National Heart, Lung, and Blood Institute tools appropriate to each design. Given substantial heterogeneity in designs, estimands, and outcome metrics, findings were synthesized narratively following the Synthesis Without Meta-Analysis guideline rather than pooled statistically. Results: Of 8032 records, 11 studies were included: 1 randomized, 1 nonrandomized comparative, 4 observational, and 5 uncontrolled pre-post studies. Seven evaluated AI lifestyle coaching systems, 2 predictive AI platforms, and 2 conversational AI chatbots. Ten of 11 studies used commercially developed platforms. Study quality was rated good for 5 studies, fair for 3, and poor for 3. The randomized controlled trial showed noninferiority of an AI-DPP versus human coach–led DPP for the combined 12-month outcome (23/183, 12.6% vs 23/185, 12.4%; risk difference 0.1%, 1-sided 95% CI –5.5%). Program initiation (171/183, 93.4% vs 153/185, 82.7%; risk difference 10.7%, 95% CI 4.2%-17.3%) and completion (117/183, 63.9% vs 93/185, 50.3%; risk difference 13.7%; 95% CI 3.6%-23.7%) were higher with AI, but this did not translate into superior clinical outcomes. The nonrandomized comparison found no advantage of AI-personalized messaging over static messaging in terms of BMI or physical activity. The remaining studies reported uncontrolled changes precluding causal inference. Evidence was insufficient to determine whether effects differed by population or delivery setting. Conclusions: One randomized trial supports noninferiority, not superiority, of a specific AI-driven DPP relative to human coaching; the remaining evidence was nonrandomized, predominantly commercially developed, and hypothesis-generating. To our knowledge, this is the first systematic review to stratify evidence on AI-driven diabetes prevention by AI mechanism—coaching, predictive, and conversational—and by outcome domain, extending prior scoping work that mapped apps without appraising study quality or comparative effectiveness. This structured mapping shows where evidence is the strongest (engagement and weight) and the weakest (glycemic causality and generalizability), giving health systems a concrete basis for deciding where AI-DPP tools may supplement, rather than replace, standard DPP delivery, pending independent, adequately powered randomized trials.

Objective: To examine the association between depressive symptoms and advanced cardiovascular-kidney-metabolic (CKM) syndrome among US adults and to determine whether this association differs by demographic, socioeconomic, and behavioral factors. Methods: This cross-sectional study used National Health and Nutrition Examination Survey data. Depressive symptoms were assessed using the Patient Health Questionnaire-9. CKM stages were defined using the American Heart Association framework and categorized as nonadvanced CKM, stages 0 to 2, versus advanced CKM, stages 3 to 4. Survey-weighted logistic regression estimated odds ratios (ORs) and 95% CIs for advanced CKM. Interaction terms tested effect modification by age, sex, race and ethnicity, poverty-income ratio, education, and smoking status. Results: The study included 30,244 adults. Overall, 16% of adults had advanced CKM, and 8% had depressive symptoms. In fully adjusted sex-stratified models, depressive symptoms were associated with higher odds of advanced CKM among women (OR=2.24; 95% CI=1.74-2.88) and men (OR=1.75; 95% CI=1.34-2.28). Interaction estimates indicated a stronger association among women (OR=1.58; 95% CI=1.12-2.23; p =.010) and a weaker association at age 65 years or older versus 30 to 49 years (OR=0.42; 95% CI=0.27-0.67; p <.001). Conclusions: Depressive symptoms were associated with higher odds of advanced CKM in both women and men, with a stronger association among women and a weaker association at ages 65 years or older versus 30 to 49. These findings support integrating mental health screening into cardiometabolic and kidney disease prevention strategies, emphasizing age- and sex-specific vulnerabilities.

Background: Racial and ethnic minority populations face disproportionate rates of uncontrolled blood pressure (BP) and hypertension-related mortality. Remote hypertension monitoring (RHM) with active clinician-led medication titration has shown promise for improving BP control, but real-world evidence in majority-minority primary care settings remains limited. Methods: This retrospective cohort study (January 2022–December 2024) enrolled adults with hypertension in a Bluetooth-integrated RHM program at a single urban academic primary care clinic. Of 550 patients enrolled, 503 with evaluable follow-up data were included. Patients transmitted daily home BP readings; clinicians reviewed readings monthly and titrated anti-hypertensive regimens per 2017 ACC/AHA guidelines. BP control was assessed at baseline and 3, 6, and 9 months. Factors associated with longitudinal BP control were examined using multivariable generalized estimating equations (GEE), with outcomes defined as strict control (<130/80 mmHg), at-least-moderate control (140/90 mmHg). Results: Among 503 participants (mean age 58.3 [SD 12.1] years; 63.6% African American; 52.9% male), BP control increased from 10.1% at baseline to 37.1% at 9 months. Each additional month of enrollment was associated with reduced odds of uncontrolled BP (adjusted odds ratio [aOR] 0.82; 95% CI, 0.80–0.85; P<.001). White race was associated with lower odds of uncontrolled BP versus African American race (aOR 0.57, at-least-moderate control; aOR 0.40, strict control; both P<.001). Male sex (aOR 1.46; P=.02) and congestive heart failure (aOR 2.09, strict control; aOR 2.05, at-least-moderate control; both P<.05) were associated with higher odds of uncontrolled BP. Conclusion: Bluetooth-integrated RHM with active clinician-led medication titration was associated with a nearly 4-fold increase in BP control over 9 months in a majority-minority primary care population. Persistent within-program racial disparities underscore the need for equity-centered strategies beyond technology adoption alone. Prospective studies with concurrent usual-care comparators are needed to establish causal inference. What is Known: Remote hypertension monitoring combined with active pharmacist- or clinician-led medication titration reduces blood pressure more effectively than passive monitoring or usual care alone, as demonstrated in randomized trials conducted predominantly in White, non-minority populations. Black Americans bear a disproportionate burden of hypertension, with earlier onset, greater severity, and higher rates of hypertensive end-organ damage compared with other racial and ethnic groups, and persistent disparities in BP control remain despite similar awareness and treatment rates. What the Study Adds: In one of the largest single-center cohort studies of physician-led remote hypertension monitoring in a majority-African American urban primary care population, BP control rates increased nearly 4-fold (10.1% to 37.1%) over 9 months, with each month of enrollment independently associated with an 18% reduction in the odds of uncontrolled BP. Despite overall program efficacy, significant within-program racial disparities persisted — White patients had 43%–60% lower odds of uncontrolled BP compared with African American patients — and male sex and congestive heart failure were independently associated with failure to achieve BP control targets, identifying high-priority subgroups for augmented intervention strategies.

BACKGROUND Artificial intelligence (AI)–driven lifestyle interventions are increasingly used for reducing diabetes risk. However, whether they produce clinical benefit beyond standard Diabetes Prevention Programs (DPPs) and whether effects vary by AI type, population, or delivery setting, remains uncertain. OBJECTIVE This study aimed to evaluate the effects of AI-driven lifestyle interventions on clinical, behavioral, and engagement outcomes among people with prediabetes or elevated diabetes risk and to assess whether effects differed by AI type, population, or delivery setting. METHODS PubMed, Embase, CINAHL, Web of Science, IEEE Xplore, ClinicalTrials.gov, and citation searches were completed on August 3, 2026, for English-language reports published from 2015. Eligible studies evaluated AI-driven digital lifestyle interventions among people with prediabetes or elevated diabetes risks, or healthy participants in feasibility studies. One reviewer (OIE) screened records and extracted data; a second reviewer (TN) independently verified exclusions and extraction. Study quality was assessed with National Heart, Lung, and Blood Institute tools appropriate to each design. Given substantial heterogeneity in designs, estimands, and outcome metrics, findings were synthesized narratively following the Synthesis Without Meta-Analysis guideline rather than pooled statistically. RESULTS Of 8032 records, 11 studies were included: 1 randomized, 1 nonrandomized comparative, 4 observational, and 5 uncontrolled pre-post studies. Seven evaluated AI lifestyle coaching systems, 2 predictive AI platforms, and 2 conversational AI chatbots. Ten of 11 studies used commercially developed platforms. Study quality was rated good for 5 studies, fair for 3, and poor for 3. The randomized controlled trial showed noninferiority of an AI-DPP versus human coach–led DPP for the combined 12-month outcome (23/183, 12.6% vs 23/185, 12.4%; risk difference 0.1%, 1-sided 95% CI –5.5%). Program initiation (171/183, 93.4% vs 153/185, 82.7%; risk difference 10.7%, 95% CI 4.2%-17.3%) and completion (117/183, 63.9% vs 93/185, 50.3%; risk difference 13.7%; 95% CI 3.6%-23.7%) were higher with AI, but this did not translate into superior clinical outcomes. The nonrandomized comparison found no advantage of AI-personalized messaging over static messaging in terms of BMI or physical activity. The remaining studies reported uncontrolled changes precluding causal inference. Evidence was insufficient to determine whether effects differed by population or delivery setting. CONCLUSIONS One randomized trial supports noninferiority, not superiority, of a specific AI-driven DPP relative to human coaching; the remaining evidence was nonrandomized, predominantly commercially developed, and hypothesis-generating. To our knowledge, this is the first systematic review to stratify evidence on AI-driven diabetes prevention by AI mechanism—coaching, predictive, and conversational—and by outcome domain, extending prior scoping work that mapped apps without appraising study quality or comparative effectiveness. This structured mapping shows where evidence is the strongest (engagement and weight) and the weakest (glycemic causality and generalizability), giving health systems a concrete basis for deciding where AI-DPP tools may supplement, rather than replace, standard DPP delivery, pending independent, adequately powered randomized trials.

Cardiovascular disease remains the leading cause of death globally. Neighborhood characteristics such as food environments, green space access, and housing conditions shape cardiovascular health by influencing environmental exposures, resource access, and chronic stress, with disproportionate impacts on socially vulnerable populations. This scoping review aims to identify and evaluate neighborhood-level interventions designed to promote cardiovascular health and determine which strategies are most effective in improving health outcomes. A comprehensive and reproducible search was conducted across PubMed, ScienceDirect, Web of Science, and OVID Medline for studies published between 2015 and 2025. Keywords related to neighborhood-level interventions and cardiovascular health were used to identify relevant articles. Neighborhood-level interventions were defined as strategies that modify the physical, social, or economic environment of geographically defined areas. Outcomes included cardiovascular health behaviors, cardiometabolic risk factors, and clinical cardiovascular outcomes. Twenty studies were included, including 15 randomized controlled trials and 5 quasi-experimental studies. Interventions targeted built environments, food retail environments, housing mobility, or employed multicomponent approaches that coupled environmental intervention with individual or community-level support. Overall, 55% of studies demonstrated beneficial cardiovascular outcomes, 20% showed mixed effects, and 25% showed null effects. Dietary behavior outcomes showed the greatest benefit (73%), followed by physical activity (63%), while cardiometabolic risk factors showed limited improvement. Interventions were particularly effective when they integrated environmental changes with support across multiple levels, including education, peer support, and community resources. Multicomponent interventions coupling neighborhood environmental modifications with either individual or community-level support show promise for improving cardiovascular health behaviors, particularly physical activity and diet in disadvantaged populations. Future work should prioritize long-term studies on interventions that integrate these elements to understand sustainability and measure cost-effectiveness to facilitate implementation of strategies and policy changes that sustain long-term health improvements.

Objective: This study investigated structural drivers of neighborhood food access to identify possible pathways at a census tract level. Methods: Data from 9597 census tracts across 201 counties, within 38 states, tested potential pathways including the built environment, criminal justice, education, housing, social cohesion, transportation, wealth, and income/poverty. Historic residential redlining was defined using Home Owners' Loan Corporation residential security maps. Food access was defined using the modified retail food environment index. Direct and indirect relationships were investigated using structural equation modeling (SEM) run in Stata v18, controlling for the population of each census tract. Standardized estimates were used with p < 0.05 considered statistically significant. Results: Redlining was directly associated with higher disadvantage across all tested pathways. Higher disadvantage in the built environment (−0.06, p < 0.001), criminal justice (−0.11, p < 0.001), housing (−0.04, p < 0.01), and social cohesion (−0.18, p < 0.001) was associated with lower food access, while higher disadvantage in education was associated with higher food access (0.07, p < 0.001), and transportation, wealth, and income were not significant pathways. Conclusions: The strongest neighborhood‐level drivers were social cohesion, criminal justice, and education. Policies that focus on a holistic approach to the neighborhood environment may be needed to improve access to healthy food options at the neighborhood level.

Background: Limited literature exists on the relationship between postrelease social determinants of health and glycemic control among African Americans with a history of criminal legal involvement and Type 2 diabetes mellitus (T2DM). Methods: Participants were recruited from clinics, community centers, and places of worship in the midwestern United States. Stepwise regression with forward selection was used to evaluate the relationship between postrelease criminal justice system factors, social risk/environmental factors, and glycemic control among African American with history of criminal legal involvement and T2DM. Results: Retained variables from the stepwise regression included age, comorbidity, type of medication, length of incarceration, insurance, and housing insecurity. The analysis revealed statistically significant associations between comorbidity, type of medication, and insurance type with glycemic control. Compared with using only pills or oral hypoglycemic agents, the combination of pills/oral hypoglycemic agents and insulin ( β ‐coefficient: 1.44; 95% CI: 0.77, 2.12) and using insulin or injectables alone ( β ‐coefficient: 0.95; 95% CI: 0.16, 1.73) were associated with an increase in HbA1c. Additionally, each additional comorbidity ( β ‐coefficient: −0.25; 95% CI: −0.40, −0.11) was associated with a decrease in HbA1c. Compared with individuals with private/military/other insurance, those with Medicare ( β ‐coefficient: 1.40; 95% CI: 0.43, 2.37) and Medicaid ( β ‐coefficient: 1.11; 95% CI: 0.12, 2.10) were associated with higher HbA1c. Conclusion: Postrelease criminal justice factors and social/environmental risk factors were not statistically significant. Future studies should evaluate the relationship between insurance type and glycemic control among this population.