Skip to main content

Table 5 Pooled logit model estimations - Mediation analysis: males

From: Economic conditions, hypertension, and cardiovascular disease: analysis of the Icelandic economic collapse

Dependent variable Hypertension   Cardiovascular Disease  
  dy/dx Robust SE   dy/dx Robust SE  
Without mediators
 t2012 0.0029 0.0119   0.0009 0.0023  
BMI included ε ε   ε ε  
 t2012 0.0049 0.0112   0.0012 0.0023  
1 if underweight
 1 if overweight 0.0497 0.0118 *** 0.0005 0.0022  
 1 if obese 0.2080 0.0315 *** 0.0051 0.0046  
Alcohol included
 t2012 0.0021 0.0119   0.0009 0.0022  
 1 if daily drinker 0.0555 0.0362   0.0225 0.0156  
 1 if frequent drinker 0.0215 0.0165   0.0010 0.0030  
 1 if seldom drinker 0.0403 0.0173 ** 0.0003 0.0024  
 1 if rare drinker 0.0348 0.0212   0.0057 0.0040  
Smoking included
 t2012 0.0024 0.0120   0.0012 0.0022  
 1 if daily smoker -0.0089 0.0147   0.0051 0.0036  
 1 if weekly smoker -0.0836 0.0148 *** -0.0068 0.0043  
 1 if seldom smoker 0.0251 0.0440   -0.0058 0.0051  
Perceived status in society included
 t2012 0.0023 0.0120   0.0002 0.0022  
 1 if much better -0.0201 0.0322   -0.0072 0.0037 *
 1 if considerably better -0.0292 0.0137 ** 0.0047 0.0035  
 1 if somewhat better -0.0033 0.0123   -0.0021 0.0024  
 1 if somewhat worse 0.0124 0.0196   -0.0006 0.0029  
 1 if considerably worse 0.0382 0.0327   0.0096 0.0062  
 1 if much worse -0.0058 0.0358   0.0315 0.0207  
Stress included
 t2012 -0.0009 0.0121   0.0004 0.0022  
 PSS 0.0071 0.0029 ** 0.0013 0.0006 **
Unemployment included
 t2012 0.0009 0.0121   0.0006 0.0023  
 1 if unemployed 0.0738 0.0387 * 0.0037 0.0056  
 Income included       
 t2012 0.0021 0.0120   0.0001 0.0023  
 real income -0.0008 0.0022   -0.0010 0.0005 *
Working hours included
 t2012 0.0041 0.0119   0.0001 0.0021  
 working hours per workday -0.0014 0.0012   -0.0004 0.0003 *
  1. 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 short-term crisis coefficient (t2009). εMissing coefficient due to perfect predictability of underweight; hence optimal weight and underweight are combined in this estimation as a benchmark. *p < 0.1, **p < 0.05, ***p < 0.01