High levels of lipoprotein(a) are associated with a lower prevalence of diabetes with advancing age: Results of a cross-sectional epidemiological survey in Gran Canaria, Spain
1 Section of Endocrinology and Nutrition, Hospital Universitario Insular, Avda. Marítima del Sur, s/n.,, 35016, Las Palmas de Gran Canaria, Spain
2 Department of Medical and Surgical Sciences, University of Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain
3 Department of Mathematics, University of Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain
Cardiovascular Diabetology 2012, 11:81 doi:10.1186/1475-2840-11-81Published: 2 July 2012
Recent data suggest that concentrations of lipoprotein(a) [Lp(a)] may be inversely associated with the risk of diabetes. This study analyzed the relationships between Lp(a) and both diabetes and insulin resistance in an adult cohort from the island of Gran Canaria, Spain.
Lp(a), homeostasis model assessment for insulin resistance (HOMA-IR) and conventional risk factors for diabetes were assessed in a sample of 1,030 adult individuals participating in a cross-sectional population-based epidemiological survey in the city of Telde. Diabetes was defined according to the WHO 1999 criteria, or as a previous diagnosis of diabetes. To identify patients at risk for diabetes, an Lp(a) cutoff level of 46 mg/dl was selected previously using classification and regression tree analysis. A multivariate logistic regression model with L2-regularization was used to assess the independent effect of Lp(a) on diabetes and its interactions with variables traditionally linked to the disease. Additionally, to investigate the effect of Lp(a) on insulin resistance, a parametric model was developed to describe the relationship between age and HOMA-IR values in subjects with levels of Lp(a) ≤46 or >46 mg/dl.
Along with variables known to be associated with diabetes, including age, mean blood pressure, serum triglycerides, and an interaction term between age and low HDL cholesterol, the logistic model identified a significant inverse association for diabetes and the interaction term between age and Lp(a) levels >46 mg/dl. According to the proposed parametric model, HOMA-IR was significantly lower in subjects of all ages who had Lp(a) levels >46 mg/dl.
These results suggest that the age-related increase in the probability of having diabetes is significantly lower in subjects with Lp(a) levels >46 mg/dl. This could be explained in part by a lower insulin resistance in this subset of the population.