Skip to main content
. 2025 Apr 28;16(4):1389–1402. doi: 10.1007/s41999-025-01201-3

Table 5.

Association between STOPP/START and the outcome of functional decline using logistic regression (odds ratio (OR), 95% confidence intervals (CIs))

n (%)a Functional decline (n = 2507)
Adjusted OR (95% CI)b
Model 1c Model 2d
Any STOPPe (n = 917) 204 (22.25) 1.46 (1.11, 1.91), p = 0.007f
 Number of STOPP PIMS
0 STOPP (n = 2095) 230 (10.98) (Referent)
1 STOPP (n = 663) 116 (17.50) 1.27 (0.94, 1.72), p = 0.116
≥ 2 STOPP (n = 254) 88 (34.65) 2.02 (1.36, 2.99), p < 0.001
Any STARTe (n = 1067) 212 (19.87) 1.23 (0.95, 1.60), p = 0.114
 Number of START PPOs
0 START (n = 1945) 222 (11.41) (Referent)
1 START (n = 653) 120 (18.38) 1.28 (0.94, 1.73), p = 0.110
≥ 2 START (n = 414) 92 (22.22) 1.17 (0.83, 1.64), p = 0.364

aNumber and percentage who experienced functional decline at Wave 5, referent: none

bAdjusted for age group, sex, education, employment, insurance coverage, multimorbidity, polypharmacy, and functional impairment at Wave 4

cModel 1: PIP exposure assessed using binary variables for presence or absence of STOPP/START

dModel 2: PIP exposure assessed using categorical variables for presence of 0, 1 and ≥ 2 STOPP/START criteria

eReferent: none

fAssociations with p <0.05 are in bold