Elevated systemic concentrations of interleukin (IL)-18, a potent proinflammatory cytokine, have recently been reported to increase the risk of developing type 2 diabetes in a prospective case-cohort study with a mean follow up of 10.8 years (1). IL-18 is secreted from adipocytes (2), and IL-18 levels were found to be increased in obesity and decreased after weight loss in obese women (3). However, we did not observe a direct relationship between BMI and IL-18 levels in a large population-based sample (1). Another recent study in 144 men surprisingly reported a significant association of IL-18 with fat-free mass (4). We now present a detailed analysis of the relationship of IL-18 serum concentrations with body composition and potential confounding effects by sex in 225 men and 230 women from the MONICA (Monitoring of Trends and Determinants in Cardiovascular Disease)/KORA (Cooperative Health Research in the Region of Augsburg) Survey 1994/1995 (1), for whom bioelectric impedance data were available.
Serum IL-18 concentrations did not differ between men and women (geometric means 154.4 and 163.9 pg/ml, respectively; P = 0.38). For men and women combined, log IL-18 was not significantly correlated (Pearson) with absolute fat mass (r = −0.02; P = 0.58), percent fat mass (r = 0.01; P = 0.82), and fat-free mass (r = −0.04; P = 0.38). Separate analyses for men and women yielded similar results. The strongest correlations were seen between log IL-18 and systolic blood pressure (r = 0.11; P = 0.01), diastolic blood pressure (r = 0.11; P = 0.01), log C-reactive protein (r = 0.09; P = 0.054), and log IL-6 (r = 0.10; P = 0.02), whereas correlations with BMI, waist circumference, cholesterol levels, and uric acid did not reach statistical significance.
Taken together, these data show that systemic IL-18 concentrations are independent from body fat composition including fat-free mass and support the hypothesis that adipose tissue is not a major contributor to circulating IL-18. This is not in contrast with data on IL-18 expression in adipocytes, since the reported amount of in vitro secreted IL-18 was rather low (2). This finding does not preclude a role for adipocyte-expressed IL-18 in adipose tissue since higher concentrations, which are relevant for the cross-talk between adipocytes and infiltrating immune cells, might be reached locally. Furthermore, we did not find a sex difference in the relation between IL-18 and markers of body fat composition, whereas this has been demonstrated for C-reactive protein and other inflammatory mediators (5). We conclude that the reported impact of weight loss on IL-18 levels (3) is indirect, as long-term caloric restriction (3) can be assumed to attenuate systemic immune activation and therefore also reduce IL-18 in the circulation. On the other hand, extensive weight loss is associated with a substantial increase in insulin sensitivity. The data linking increased IL-18 with reduced insulin sensitivity (4) and elevated type 2 diabetes risk (1) indicate that insulin resistance and not obesity per se may be the major determinant of circulating IL-18 concentrations.