Table 4.
Ordinal regression analysis for change in quality of life (dependent variable), results from a model including all covariates shown in the table
N | eβ | CI (95%) | |
---|---|---|---|
Age (years) | 1097 | 1.01 | 0.99 to 1.30 |
Menopausal transition status | |||
Pre-prea) | 225 | 1 (ref.) | |
Pre-peri, postb) | 513 | 0.93 | 0.77 o 1.14 |
Peri-peri/post.post-postc) | 359 | 0.88 | 0.70 to 1.11 |
Education | |||
primary | 284 | 1 (ref.) | |
secondary | 358 | 1.01 | 0.85 to 1.21 |
tertiary | 455 | 1.28** | 1.08 to 1.51 |
QoL at baseline | |||
poor | 19 | 14.18 | 8.38 to 24.00 |
moderate | 271 | 5.94 | 4.91 to 7.19 |
good | 383 | 1.95 | 1.67 to 2.28 |
excellent | 424 | 1 (ref.) | |
Physical activity change (MET/week) | |||
decrease | 232 | 1 | |
stable | 590 | 1.46*** | 1.24 to 1.73 |
increase | 275 | 1.49*** | 1.23 to 1.80 |
Weight change | |||
gainers | 208 | 1 | |
stable | 765 | 1.26** | 1.07 to 1.50 |
losers | 124 | 1.20 | 0.94 to 1.53 |
HRT Use in baseline | |||
Current | 403 | 1 | |
Before | 177 | 0.73 | 0.73 to 1.18 |
Never | 517 | 1.26* | 1.02 to 1.56 |
R2 (Nagelkerke) | 0.340 |
Interpretation of results: if eβ > 1, the odds of higher QoL are greater, whereas when eβ < 1 low QoL is more probable.
*p < 0.05
**p < 0.01
***p < 0.001
a) premenopause at both baseline and follow up
b) transition from premenopause to peri- or postmenopause
c) perimenopause or postmenopause both at baseline and follow-up; from perimenopause at baseline to postmenopause at follow-up