Skip to main content
. 2016 Jul 18;14(2):101–109. doi: 10.1007/s10433-016-0392-1

Table 3.

Multinomial logistic regression models of change in loneliness by change in OIDPa between Waves 3 (2006–2007) and 5 (2010–2011), N = 4640, relative risk ratios (RRRs) 95 % confidence interval (95 % CI)

Becoming lonely (cases/n = 389/4640) Becoming less lonely (cases/n = 372/4640)
Change in OIDP No change Incident oral impact Recovery from oral impact No change Incident oral impact Recovery from oral impact
Cases/n 325/389 40/389 24/389 322/372 29/372 21/372
Model 1 (age-adjusted) 1.00 1.76 (1.23–2.50)** 1.62 (1.03–2.52)* 1.00 1.27 (0.85–1.90) 1.44 (0.90–2.30)
Model 2 (model 1 + socio-demographic factorsa) 1.00 1.76 (1.23–2.51)** 1.59 (1.02–2.49)* 1.00 1.22 (0.81–1.83) 1.35 (0.84–2.17)
Model 3 (model 2 + socio-economic factorsb) 1.00 1.74 (1.22–2.48)** 1.55 (1.00–2.42) 1.00 1.20 (0.80–1.81) 1.32 (0.82–2.13)
Model 4 (model 3 + health-related factorsc) 1.00 1.64 (1.15–2.36)** 1.40 (0.89–2.20) 1.00 1.11 (0.73–1.68) 1.17 (0.72–1.89)
Model 5 (model 4 + psychosocial factorsd) 1.00 1.56 (1.09–2.25)* 1.34 (0.85–2.11) 1.00 0.98 (0.65–1.49) 1.07 (0.65–1.75)

OIDP oral impacts on daily performances

p value < 0.05; ** p value < 0.01

aGender and cohabiting status

bEducational qualifications and wealth

cLimiting long-standing illness, depressive symptoms, smoking status, and edentulousness

dSocial participation and social support