Table 3.

Mediation Analysis of ln[25(OH)D] and SPPB Scores by ln[klotho] Among InCHIANTI Participants Aged ≥55 y

SPPB Outcome	ln[25(OH)D] Effect and Model	Beta Estimate*	95% Confidence Interval	p Value
Summary	Total	0.61	(0.35, 0.88)	<.001
	Direct (unmediated by klotho): Mediation Model 1	0.61	(0.34, 0.87)	<.001
	Mediation Model 2	0.63	(0.36, 0.90)	<.001
Balance	Total	0.29	(0.17, 0.42)	<.001
	Direct (unmediated by klotho): Mediation Model 1	0.29	(0.16, 0.42)	<.001
	Mediation Model 2	0.30	(0.17, 0.43)	<.001
Chair stand	Total	0.15	(0.06, 0.24)	.001
	Direct (unmediated by klotho): Mediation Model 1	0.15	(0.06, 0.24)	.001
	Mediation Model 2	0.15	(0.06, 0.25)	.001
Walk	Total	0.17	(0.09, 0.25)	<.001
	Direct (unmediated by klotho): Mediation Model 1	0.17	(0.09, 0.25)	<.001
	Mediation Model 2	0.18	(0.10, 0.26)	<.001

Notes: Total effect: adjustment for study visit, age, sex, visit-by-age interaction, eGFR, calcium intake, alcohol consumption, smoking, body mass index (continuous and quadratic terms), MMSE, years of education, physical activity level, history of falls, comorbid conditions (hypertension, congestive heart failure, peripheral arterial disease, stroke, diabetes, cancer, renal disease, and osteoporosis), and season of blood collection.

Mediation Model 1: additionally included ln(klotho).

Mediation Model 2: additional adjustment for PTH using weighting (see Supplementary Material). 25(OH)D = 25-hydroxyvitamin D; eGFR = estimated glomerular filtration rate; InCHIANTI = Invecchiare in Chianti, “Aging in Chianti”; MMSE = Mini-Mental State Examination; PTH = parathyroid hormone; SPPB = Short Physical Performance Battery.

*Mean difference in SPPB summary and component scores per ln[25(OH)D] in ng/mL.