Table 4.

Longitudinal relations between baseline circulating sex hormone levels and incident mobility limitation: multivariable logistic regression (n = 1111)

	Mobility limitation		Subjective health
	OR (95% CI)	P	OR (95% CI)	P
Continuous hormone levels
TT	0.90 (0.74–1.09)	0.28	1.09 (0.80–1.47)	0.59
FT	0.78 (0.62−0.97)	0.03	0.87 (0.62–1.22)	0.42
SHBG	1.11 (0.91–1.34)	0.31	1.29 (0.97–1.72)	0.08
Dichotomized sex hormone levels
Low TT	1.46 (0.91–2.34)	0.12	0.48 (0.19–1.24)	0.13
Low FT	1.57 (1.06–2.32)	0.03	1.23 (0.64–2.38)	0.53
High SHBG	1.29 (0.80–2.10)	0.30	1.59 (0.74–3.43)	0.24

OR values are for 1 sd change in hormone levels. For incidence, the sample excludes subjects at exam 7 with mobility limitation and poor subjective health. For progression, the sample excludes subjects at exam 7 with worst response choice for mobility limitation and subjective health. All models were adjusted for age, BMI, smoking, and comorbidities (cancer and cardiovascular disease). Sex hormones were defined as low or high vs. normal using healthy reference sample of FHS Generation 3 men. Low TT and FT levels were those below the 2.5th percentile of the referent sample (TT <348.3 ng/dl; FT <70.0 pg/ml), and SHBG levels above the 97.5th percentile of the referent sample (SHBG, 81.6 nmol/liter) represented high SHBG levels. Low FT levels (<70 pg/ml) at examination 7 were associated with increased risk (OR = 1.57) of developing mobility limitation at examination 8. Each sd increase in FT level at examination 7 was associated with a 22% decrease (OR = 0.78) in risk of reporting mobility limitation at examination 8.