Dependent variable

Hypertension
 
Cardiovascular Disease
 


dy/dx

Robust SE
 
dy/dx

Robust SE
 

Without mediators

t2012

0.0258

0.0132

*

0.0007

0.0026
 
BMI included

t2012

0.0224

0.0127

*

0.0004

0.0020
 
1 if underweight

0.0235

0.0961
 
^{ε}

^{ε}
 
1 if overweight

0.0606

0.0133

***

0.0002

0.0022
 
1 if obese

0.1710

0.0196

***

0.0032

0.0026
 
Alcohol included
 
0.0132
    
t2012

0.0238

0.0224

*

0.0002

0.0021
 
1 if daily drinker

0.0830

0.0158

***

^{σ}

^{σ}
 
1 if frequent drinker

0.0365

0.0150

*

0.0009 ^{σ}

0.0027 ^{σ}
 
1 if seldom drinker

0.0068

0.0166
 
0.0010

0.0022
 
1 if rare drinker

0.0185
  
0.0025

0.0025
 
Smoking included

t2012

0.0247

0.0133

*

0.0004

0.0020
 
1 if daily smoker

0.0358

0.0127

***

0.0020

0.0022
 
1 if weekly smoker

0.0367

0.0440
 
0.0053 ^{δ}

0.0028 ^{δ}

*

1 if seldom smoker

0.0385

0.0351
 
^{δ}

^{δ}
 
Perceived status in society included

t2012

0.0260

0.0134

*

0.0005

0.0019
 
1 if much better

0.0354

0.0443
 
0.0014

0.0054
 
1 if considerably better

0.0013

0.0179
 
0.0051

0.0023

**

1 if somewhat better

0.0153

0.0131
 
0.0025

0.0019
 
1 if somewhat worse

0.0112

0.0160
 
0.0028

0.0020
 
1 if considerably worse

0.0123

0.0228
 
0.0002

0.0031
 
1 if much worse

0.0048

0.0475
 
0.0037

0.0025
 
Stress included

t2012

0.0258

0.0135

*

0.0002

0.0020
 
PSS

0.0034

0.0029
 
0.0007

0.0004
 
Unemployment included

t2012

0.0256

0.0133

*

0.0004

0.0020
 
1 if unemployed

0.0190

0.0284
 
0.0110

0.0081
 
Income included

t2012

0.0247

0.0134

*

0.0009

0.0020
 
real income

0.0014

0.0033
 
0.0007

0.0006
 
Working hours included

t2012

0.0254

0.0135

*

0.0003

0.0021
 
working hours per workday

0.0033

0.0016

**

0.0004

0.0003
 
 Results are presented as marginal effects. Sample weights are applied. Covariates controlled for are age, age squared, number of children, marital status, residence, education, prescription medication, and shortterm crisis coefficient (t2009). ^{ε}Missing coefficient due to perfect predictability of underweight; hence optimal weight and underweight are combined in this estimation as a benchmark. ^{σ}Missing coefficient due to perfect predictability of daily drinker; hence daily drinker and frequent drinker are combined in this estimation. ^{δ}Missing coefficient due to perfect predictability of seldom smoker; hence seldom smoker and weekly smoker are combined in this estimation. *p < 0.1, **p < 0.05, ***p < 0.01