Dietary advanced glycation end-products (dAGEs) intake and its relation to sarcopenia and frailty – The Rotterdam Study

Studies on mice have shown a relationship between dietary intake of advanced glycation end-products (dAGEs) and deterioration of musculoskeletal health, but human studies are absent. We investigated the relationship between dietary intake of carboxymethyllysine (dCML) - an AGE prototype - and risk of sarcopenia at baseline and after 5 years of follow-up and a single evaluation of physical frailty in participants from the population-based Rotterdam Study. Appendicular lean mass (ALM) was obtained using insight dual-energy X-ray absorptiometry and hand grip strength (HGS) using a hydraulic hand dynamometer. Subjects with both low ALM and weak HGS were classified as having sarcopenia. Frailty (yes/no) was defined by presence of ≥3 and pre-frailty by presence of 1 or 2 components namely, exhaustion, weakness, slowness, weight loss or low physical activity. dCML was calculated using a food frequency questionnaire and dAGE databases. Logistic regression analysis was used to evaluate the odds of physical frailty and prevalent sarcopenia at baseline and follow-up and incident sarcopenia. 2782 participants with an age 66.4 ± 9.9 years and dCML intake 3.3 ± 1.3 mg/day, had data on sarcopenia at both time points. Of whom 84 had sarcopenia at baseline and 73 developed sarcopenia at follow-up. We observed an association of one SD increase in dCML intake with prevalent sarcopenia at baseline [odds ratio, OR = 1.27 (1.01-1.59)] and no association of dCML with incident sarcopenia at 5-year follow-up [OR = 1.12 (0.86-1.44)]. For frailty we analyzed 3577 participants, of whom 1972 were pre-frail and 158 were frail. We observed no association of dCML with either pre-frailty [OR = 0.99 (0.91-1.07)] or frailty [OR = 1.01 (0.83-1.22)] when non-frail subjects were used as reference. Our results show an association of dAGEs with sarcopenia cross-sectionally but not longitudinally where inconclusive findings are observed possibly due to a very low incidence of sarcopenia. There was no association with frailty cross-sectionally.