The association of a lifestyle risk index with glycemic status in the German National Cohort (NAKO)

Abstract

This cross-sectional analysis examined the association of a Lifestyle Risk Factor Index (LSRI) with the presence of self-reported diabetes or HbA1c levels among participants of the German National Cohort (NAKO), while considering the role of body mass index (BMI) and visceral adipose tissue (VAT). Modified LSRI (mLSRI) scores were calculated from baseline questionnaires assigning points for adherence to alcohol intake, physical activity, smoking, and healthy diet recommendations. Of > 205,000 participants, 18,101 had VAT quantified by magnetic resonance imaging plus complete information on covariates. Diabetes status was self-reported and HbA1c levels ≥ 39 mmol/mol Hb (5.7%) were considered elevated. Polytomous logistic regression estimated prevalence odds ratios (POR) and linear regression computed mean HbA1c levels adjusting for age and sex in model 1 and adding BMI and VAT for model 2. The respective proportions for 0/1, 2, 3, and 4 mLSRI points were 6%, 25%, 51%, and 18%. Among the 17,457 (96.4%) participants without self-reported diabetes, 17.3% had elevated HbA1c levels. Compared to normoglycemia, the highest vs. lowest mLSRI was inversely related to abnormal HbA1c (POR 0.81; 95%CI 0.68–0.97) but the associations were attenuated after adding BMI and VAT. No significant association was seen for diabetes (POR 0.91; 95%CI 0.62–1.32) after full adjustment. In this large sample of predominantly healthy adults, a combination of lifestyles, foremost driven by not smoking, was weakly associated with lower HbA1c levels but not self-reported diabetes. Nevertheless, lifestyle factors may be relevant for glycemic control, possibly due to an influence on adiposity, but this relation cannot be confirmed due to the cross-sectional design.