Statistical analysis

Sample characteristics are presented as mean ± SD, or percentage present, and differences across age groups were compared with t tests or chi-squared tests as appropriate. Spearman correlation coefficients were calculated to examine the association between FI and the level of metabolic syndrome for each age group. Kaplan-Meier curves with log-rank analysis were used to compare survival probability stratified by FI level and metabolic syndrome status for both younger (<65 years) and older people (≥65 years), with censoring for non-responders. Unadjusted Cox proportional hazard ratios for mortality risk were calculated for each of the components of the FI and the metabolic syndrome, FI score, metabolic syndrome status, education, race, age, and sex, for each age group. Adjusted Cox proportional hazard ratios for mortality were determined for FI and metabolic syndrome status with covariates of age, sex, education, and race for the whole sample and after stratifying by age. Data analysis was conducted using SPSS v20 and graphs were created in SigmaPlot v.11. A value of p < 0.05 was considered statistically significant.