Associations of age with serum insulin, proinsulin and the proinsulin-to-insulin ratio: a cross-sectional study

Background: Insulin responses and insulin levels seem to decline with age. However, the question of beta cell impairment attributable to ageing has been sparsely addressed in population-based studies. Non-fasting insulin levels are determined by the ambient degree of insulin resistance together with the capacity of beta cells to compensate by insulin secretion to prevent hyperglycaemia. A raised proinsulin-to-insulin ratio (proinsulin/insulin) due to impaired processing of proinsulin is an early marker of beta cell dysfunction. We hypothesised that in a general population, signs of beta cell failure with advancing age manifest not only by decreases in random insulin, but also with a corresponding increase in its precursor proinsulin. Methods: In the Tromso Study 1994-95 we measured insulin and proinsulin concentrations in random blood samples from 6212 persons without self-reported diabetes mellitus and plotted the levels as percentiles according to age. In regression analyses we assessed the relationships between age and insulin, proinsulin, and proinsulin/insulin, while adjusting for the concomitant measurements of glucose and other metabolic variables, and the time since the last meal. Results: Median insulin concentrations declined significantly with advancing age group in men, but not in women. Proinsulin levels and proinsulin/insulin increased across age groups in both genders. After adjustment, greater age was associated with lower log(10)(insulin) and higher log(10)(proinsulin) and log(10)(proinsulin/insulin) (p = 0.0001 for all). Conclusions: Negative associations of age with random insulin levels, together with positive associations of age with proinsulin and proinsulin/insulin, point towards a loss of beta cell function inherent in the ageing process.