Abstract
Background: Type 2 diabetes is one of the most prevalent chronic diseases worldwide. People with type 2 diabetes have increased risk of cardiovascular disease (CVD), which is influenced by various medical, lifestyle, psychosocial, demographic, and socioeconomic factors. To understand CVD risk in type 2 diabetes within the whole-person paradigm of health, the aim of this study was threefold. First, we identified subgroups within the type 2 diabetes population based on lifestyle and psychosocial CVD risk factors. Second, we explored subgroup differences in demographic factors, socioeconomic factors and medical history that predict class membership. Third, we analyzed whether three-year CVD incidence differs between the identified subgroups.
Approach: To identify homogeneous latent classes, we performed a Latent Class Analyses (LCA), which is a data-driven, person-centered analysis technique. Such a data-driven approach can lead to new insights into subgroups within the population and relations between variables. The current LCA was performed based on data from the Dutch Public Health Services and Statistics Netherlands (2006-209). Within the type 2 diabetes population who participated in a national Health Survey (N = 29,086), latent classes were determined based on lifestyle and psychosocial CVD risk factors. The predictive value of demographic, socioeconomic and medical factors on class membership and class differences in CVD outcomes were examined.
Results: A 3-class model was considered statistically and clinically superior. The low riskclass was largest (7.2%). The mobility related risk class (9.3%) showed high probability of limitations in mobility (0.90) and not meeting the exercise norm (0.89). The psychosocial risk class (9.5%) had similar risks, with additional probabilities to be lonely (0.49) and have anxiety and depression (0.56). Strong predictors (OR 2.00) for membership of mobility related risk or psychosocial classes were: female gender, non-western immigrant status, lower education, financial difficulties, being unfit for work and/or prior heart failure. Three years after classification, fewer members of the low class had cardiology care and/or stroke diagnosis compared to the psychosocial and mobility related classes.
Implications: Classes with different lifestyle and psychosocial CVD risk factor patterns are associated with different CVD outcomes over time. This underlines the importance of a whole-person view in CVD prevention for the type 2 diabetes population. As such, in CVD prevention, these risks as well as demographic and socioeconomic backgrounds related to membership of the risk classes (e.g. low SES and financial difficulties) should be accounted for. A population health management approach can help to tailor (preventive) interventions to the specific class needs. This requires widening of the current medical focus as tailoring preventive CVD methods within a whole-person approach to health demands bridging gaps across healthcare, health promotion and prevention, social care and welfare.