Skip to main content
. 2024 Mar 14;30(4):1127–1133. doi: 10.1038/s41591-024-02864-4

Table 3.

Secondary outcomes, including depressive symptomatology (PHQ-9), anxiety symptomatology (GAD-7), health-related quality of life (EQ-5D-5L), capability well-being in older adults (ICECAP-O) and perceived loneliness (3-item UCLA) at the 3-month and 5-month follow-ups

Outcome Baseline 3-month follow-up 5-month follow-up
Intervention Control Intervention Control Coefficient (95% CI)a,b Pc Intervention Control Coefficient (95% CI)a,b Pc
PHQ-9, n 298 305 257 270 250 260
Mean (s.d.) 16.39 (4.66) 16.68 (4.83) 11.72 (7.72) 12.92 (7.13) −1.14 (−2.33–0.05) 0.060 10.01 (7.07) 10.41 (7.01) −0.21 (−1.36–0.94) 0.716
GAD-7, n 285 283 256 268 250 260
Mean (s.d.) 14.46 (4.62) 14.36 (4.75) 10.30 (6.62) 11.07 (6.43) −0.89 (−1.88–0.11) 0.082 9.40 (6.79) 9.49 (6.66) −0.14 (−1.20–0.92) 0.793
EQ-5D-5L, n 296 305 256 266 247 259
Mean (s.d.) 0.789 (0.182) 0.794 (0.170) 0.812 (0.189) 0.801 (0.171) 0.025 (−0.015–0.064) 0.218 0.840 (0.164) 0.825 (0.171) 0.035 (−0.001–0.071) 0.059
ICECAP-O, n 295 300 254 266 246 259
Mean (s.d.) 0.614 (0.172) 0.621 (0.190) 0.655 (0.174) 0.642 (0.194) 0.018 (−0.009–0.045) 0.185 0.665 (0.165) 0.667 (0.172) 0.002 (−0.024–0.029) 0.856
3-item UCLA, n 298 305 255 267 250 259
Mean (s.d.) 6.04 (1.91) 6.16 (1.92) 5.68 (1.97) 5.54 (1.94) 0.19 (−0.09–0.48) 0.186 5.43 (1.97) 5.25 (1.97) 0.22 (−0.07–0.52) 0.140

PHQ-9 scores range from 0 to 27, with higher scores representing more severe depressive symptomatology.

GAD-7 scores range from 0 to 21, with higher scores representing more severe anxiety symptomatology.

EQ-5D-5L scores range from −0.264 to 1, with higher scores representing higher quality of life.

ICECAP-O scores range from 0 to 1, with higher scores representing greater levels of capability well-being.

The 3-item UCLA scores range from 3 to 9, with higher scores representing more severe levels of loneliness.

aDifferences in means were estimated using linear regression models, adjusted for the baseline assessment of the corresponding outcome and the stratified variable (sex, age group and baseline severity of the PHQ-9 score (categorical))

bAll estimates had missing data imputed separately, according to trial arm, using MICE models that included predictors of missingness (extended data), stratification (sex, age group and baseline severity of PHQ-9 score (categorical)) and any imbalances in outcome measure at baseline.

cAll P values reported are based on two-sided t-tests from the relevant linear regression model, with no adjustments for multiple tests.