Social Risk Subgroups Predict Type 2 Diabetes Outcomes

Adults with type 2 diabetes can be grouped on the basis of differing levels of social risk to guide interventions, according to the results of a study published in JAMA Network Open.

Social determinants of health may account for 30% to 55% of health outcomes, as they increase the risk for poor outcomes among individuals with food insecurity, limited social support systems, housing instability, and poor access to health care. Although previous research findings suggest a link between social risks and diabetes, increasing numbers of social risk factors are not associated with glycemic control. Moreover, the identification of subgroups could facilitate intervention development.

To identify the associations between subgroups of social risk and clinical outcomes among adults with type 2 diabetes, researchers conducted a cross-sectional study using a latent profile analysis. Data was collected at 2 primary care clinics in the southeastern United States from 2013 to 2014.

After completing questionnaires for 26 social risk factors and psychological/behavioral risk, participants were categorized on the basis of 5 domains of social determinants of health (economic, neighborhood, food, social, and community).  The researchers regressed these subgroup profiles on outcomes of hemoglobin A1c (HbA_1c), blood pressure, and quality of life.

The study included 615 participants (mean [SD] age, 61.3 [10.9] years), of whom 61.6% were men. The participants identified as Black/African American (64.9%), White (33.0%), and another race/ethnicity (2.1%), whereas 41.6% of participants had an income of less than $20,000 per year.

The researchers identified 5 latent class profiles on the basis of Akaike Information Criterion and Bayesian Information Criterion scores:

• Group 1, lowest overall risk;
• Group 2, low economic risk and high neighborhood risk;
• Group 3, high economic and neighborhood risk;
• Group 4, high psychological and behavioral risk; and,
• Group 5, high economic, neighborhood, psychological, and behavioral risk.

Because it contained some risk, group 2 was identified as the reference group to better indicate whether the latent profiles could differentiate risk.

Compared with the reference group, the highest risk group (group 5) had a substantially higher HbA_1c level (β, 1.07; 95% CI, 0.50-1.63), and lower mental health-related quality of life (β, -2.15; 95% CI, -2.87 to -1.42).

Group 4, compared with the reference group, also had substantially higher HbA_1c levels (β, 0.47; 95% CI, 0.01-0.92) and a lower mental health-related quality of life (β, -1.83; 95% CI, -2.41 to -1.24).

Group 3 had considerably higher blood pressure (β, 8.08; 95% CI, 2.16-14.01) than the reference group. Group 1 had better mental health–related quality of life than another group with greater neighborhood risk (β, 1.11; 95% CI, 0.67-1.55).

Study limitations included the inability to infer causality, a lack of generalizability to other populations and regions of the country, and the potential that samples with different characteristics may yield different social risk profiles.

“These findings suggest that social risk profiles can be created identifying individuals with differing levels of economic, neighborhood, psychological, and behavioral risk and that these profiles could be used to identify ideal interventions for different risk groups,” the study report authors wrote. “This study also highlights the importance of capturing different domains of risk in screening tools.”