Comparing associations of handgrip strength and chair stand performance with all-cause mortality—implications for defining probable sarcopenia: the Tromsø Study 2015–2020

Table 4 Cox proportional hazard regression models for all-cause mortality (N = 7838). The Tromsø Study 2015–2020

Variable	N	Deaths	MR^a	Model 1^b		Model 2^c		Model 3^c
				HR	95% CI	HR	95% CI	HR	95% CI
Handgrip strength
Not low	4191	62	3.13	1.00	Ref	1.00	Ref	1.00	Ref
Low	3269	132	8.64	1.30	0.94, 1.79	1.36	0.98, 1.91	1.22	0.87, 1.71
Very low	378	39	22.87	2.23	1.43, 3.47	2.29	1.43, 3.68	1.68	1.02, 2.75
5-CST performance
Not low	5592	88	3.35	1.00	Ref	1.00	Ref	1.00	Ref
Low	1842	97	11.22	2.09	1.55, 2.81	1.94	1.43, 2.64	1.88	1.38, 2.56
Very low	404	48	26.44	3.51	2.41, 5.11	3.01	2.02, 4.50	2.64	1.73, 4.03

MR, mortality rate; HR, hazard ratio; CI, confidence interval; 5-CST, 5-repetition chair stand test
^aMortality rate per 1000 person-years
^bTests of sex interaction: low handgrip strength (p = 0.987), very low handgrip strength (p = 0.156), low 5-CST performance (p = 0.099), very low 5-CST performance (p = 0.654)
^cEstimates are from models run across 30 imputed datasets combined using Rubin’s rules
Model 1: adjusted for age (as timescale) and sex
Model 2: adjusted for model 1 + height, BMI, leisure time physical activity, education, smoking status, cardiovascular disease, rheumatoid arthritis, respiratory disease
Model 3: adjusted for model 1 and model 2 + handgrip strength status or 5-CST status

ISSN: 1741-7015