Association of adiposity, telomere length and mortality: data from the NHANES 1999-2002

BACKGROUND/OBJECTIVES: Telomere shortening is associated with age and risk of medical comorbidity. We assessed the relationship between measures of adiposity, leukocyte telomere length, and mortality and whether it is modified by age. SUBJECTS/METHODS: Subjects with dual-energy X-ray absorptiometry measures were identified using the National Health and Nutrition Examination Survey 1999-2002. Obesity was categorized using two body fat definitions (BF1%: men >= 25%; females >= 35%; BF2% >= 28% and >= 38%, respectively), body mass index (BMI) and waist circumference (WC; men >= 102 cm; females >= 88 cm). Telomere length relative to standard reference DNA (T/S ratio) was assessed using quantitative PCR. Weighted multivariable regression models evaluated the association of telomere length with adiposity, both continuously and categorically (low/normal BF%, low/high WC and standard BMI categories). Differences in telomere length by age and adiposity were ascertained and subsequent models were stratified by age. Proportional hazard models assessed the risk of mortality by adiposity status. A telomere by adiposity interaction was tested in the entire cohort and by age category (<60 vs >= 60 years; <70 vs >= 70 years). RESULTS: We identified 7827 subjects. Mean age was 46.1 years. Overall telomere length was 1.05 +/- 0.01 (s. e.) that differed by BF1% (low/high: 1.12 +/- 0.02 vs 1.03 +/- 0.02; P < 0.001), BF2% (1.02 +/- 0.02 vs 1.11 +/- 0.02; P < 0.001), BMI (underweight 1.08 +/- 0.03; normal 1.09 +/- 0.02; overweight 1.04 +/- 0.02; and obese 1.03 +/- 0.02; P < 0.001) and WC (low/high 1.09 +/- 0.02 vs 1.02 +/- 0.02; P < 0.001). Adjusted beta-coefficients evaluating the relationship between telomere length and adiposity (measured continuously) were as follows: BF1% (beta = -0.0033 +/- 0.0008; P < 0.001), BF2% (-0.041 +/- 0.008; Po0.001), BMI (beta = -0.025 +/- 0.0008; P = 0.005) and WC (beta = -0.0011 +/- 0.0004; P = 0.007). High BF% (BF1%: beta = -0.035 +/- 0.011; P = 0.002; BF2%: beta = -0.041 +/- 0.008; P < 0.001) and WC (beta = -0.035 +/- 0.011; P = 0.008) were inversely related to telomere length (TL). Stratifying by age, high BF1% (-0.061 +/- 0.013), BF2% (-0.065 +/- 0.01), BMI-obesity (-0.07 +/- 0.015) and high WC (-0.048 +/- 0.013) were significant (all Po0.001). This association diminished with increasing age. In older participants, TL was inversely related to mortality (hazard ratio 0.36 (0.27, 0.49)), as were those classified by BF1% (0.68 (0.56, 0.81)), BF2% (0.75 (0.65, 0.80)), BMI (0.50 (0.42, 0.60)) and WC (0.72 (0.63, 0.83)). No interaction was observed between adiposity status, telomere length and mortality. CONCLUSIONS: Obesity is associated with shorter telomere length in young participants, a relationship that diminishes with increasing age. It does not moderate the relationship with mortality.