Skip to main content

Table 3 Linear regression models in which baseline health behavior (HBSS2003) or its change (HBSSchange) predicts follow-up subjective well-being (SWBscore2012)

From: Changed health behavior improves subjective well-being and vice versa in a follow-up of 9 years

Model HBSS2003
β
p-value
(standard error)
HBSSchange
p-value
SWBscore2003
β
p-value
(standard error)
Age
p-value
Gender
p-value
Education
p-value
Diseases
p-value
Negative life events
p-value
AIC
Model 1.0: Crude model, no covariates  − 0.51 55,200
 < .001
(0.033)
Model 1.1: Model 1.0 + Age, gender, education, diseases  − 0.47  < 0.001 0.11  < 0.001  < 0.001 54,500
 < .001
(0.034)
Model 1.2: Model 1.1 + HBSSchange  − 0.64  < 0.001  < 0.001 0.038 0.003  < 0.001 51,100
p < .001
(0.040)
Model 1.3: Model 1.1 + SWBscore2003  − 0.24 0.44  < 0.001 0.26 0.023  < 0.001 51,800
 < .001  < .001
(0.031) (0.0086)
Model 1.4: Model 1.1 + HBSSchange + SWBscore2003  − 0.36  < 0.001 0.44  < 0.001 0.10 0.10  < 0.001 48,600
 < .001  < .001
(0.036) (0.0088)
Final model 1.5: Model 1.4 + Negative life events  − 0.35  < 0.001 0.42  < 0.001 0.45 0.038  < 0.001  < 0.001 48,300
 < .001  < .001
(0.036) (0.0088)
Model 1.6: HBSSchange + Age, gender, education, diseases .018  < .001 0.42  < 0.001 0.002 51,300
Model 1.7: HBSSchange + SWBscore2003 + Age, gender, education, diseases 0.001 0.45  < 0.001 0.88  < 0.001  < .001 48,700
 < .001
(0.0088)
  1. Results of Finnish population-based Health and Social Support study
  2. HBSS, Health behavior sum score indicating number of protective health behaviors; HBSSchange, Change in health behavior sum score during follow-up (categorical variable, c.f. Table 4); SWBscore, Subjective well-being score with reversed scoring (lower scores indicating better SWB)