Table 5.

Logistic regressions for the association of 25th percentiles of MCS and PCS with psychosocial and health-related variables adjusted by age and gender

	MCS OR (95 % CI)	PCS OR (95 % CI)
Unable to read and write	5.25 (1.07–5.92)**	1.44 (0.31–6.74)
Education
<6 years of education	1.75 (0.76–4.03)	0.95 (0.46–1.96)
6–12 years of education	0.68 (0.33–1.37)	1.24 (0.69–2.23)
Living alone	0.39 (0.15–1.00)*	0.37 (0.17–0.80)**
Having social support	0.54 (0.30–0.97)**	1.09 (0.66–1.80)
Self-perceived symptoms of depression (GDS-15 ≥5)	26.89 (15.31–47.21)***	3.41 (2.13–5.45)***
Chronic diseases (≥2)	3.77 (1.41–10.06)***	4.23 (1.85–9.69)***
Limitations in ≥1 ADL	3.40 (0.80–14.45)*	17.99 (1.95–166.34)**
Limitations in ≥1 IADL	1.18 (0.47–2.94)	4.89 (2.54–9.44)***
Obesity	1.06 (0.63–1.79)	2.22 (1.48–3.34)***

Reference categories: able to read and write; >12 years of education; having 0–1 chronic diseases; nutritional state normal or overweight

OR odds ratios, CI confidence interval, MCS and PCS mental and physical health summaries. Odds ratios are from logistic regression related 25th percentiles of MCS and PCS with psychosocial and health-related variables adjusted by age and gender

* p < 0.10; ** p < 0.05; ***p < 0.01. Hosmer–Lemeshow (p = 0.635 and 0.488; respectively) indicated the goodness of fit of the models are satisfactory