Table 5. Multivariable logistic regression analysis of the predicted effect on functional status at the 6 and 12-month follow-up in groups divided on the basis of cognitive problems.

	ADL decline								IADL decline
	at 6 months				at 12 months				at 6 months				at 12 months
	Unadjusted		Adjusted		Unadjusted		Adjusted		Unadjusted		Adjusted		Unadjusted		Adjusted
Independent variables	RR	95% CI	RR	95% CI	RR	95% CI	RR	95% CI	RR	95% CI	RR	95% CI	RR	95% CI	RR	95% CI
Cognitive impairment only (B) vs non-cognitive problem (A)	1.44	0.62–3.36	1.50	0.52–4.39	1.04	0.43–2.50	0.99	0.32–3.06	1.54	0.73–3.20	1.95	0.79–4.78	1.5	0.71–3.14	1.88	0.76–4.66
Delirium superimposed on cognitive impairment (C) vs non-cognitive problem (A)	6.75&	1.84–24.78	12.80*	1.65–99.40	6.31&	1.89–21.11	7.96*	1.35–46.99	4.13*	1.27–13.46	13.68&	1.94–96.55	5.98&	1.79–20.03	30.61&	2.94–318.54

^*p<0.05, ^&p<0.01.

Covariates after adjusting for ADL: age, gender, admission type, type of surgery, hearing impairment, polypharmacy, GDS-15 scores, BMI, risk of malnutrition, CCI scores, pain VAS scores, ASA physical status, length of hospital stay, baseline IADL scores.

Covariates after adjusting for IADL: age, gender, admission type, type of surgery, hearing impairment, polypharmacy, GDS-15 scores, BMI, risk of malnutrition, CCI scores, pain VAS scores, ASA physical status, length of hospital stay, baseline ADL scores.