Skip to main content
. 2012 Dec 23;12:1109. doi: 10.1186/1471-2458-12-1109

Table 2.

Covariate adjusted hazard ratio table

Covariate HR1 95% CI p-value HRCovariate-adjusted2 Change3 % Change4
High distress [age adj.]
 1.33
 1.10-1.61
 0.003
 -
 -
 -
Sociodemographic factors
  
  
  
  
  
  
Age
 1.04
 1.03-1.04
 <0.001
 1.33
 -
 -
Female sex [ref male]
 0.74
 0.62-0.90
 0.002
 1.37
 0.030
 10.36
Marital status [ref unmarried]
 1.18
 0.96-1.45
 0.120
 1.34
 0.006
 2.25
Lower education [ref medium]
 1.29
 1.02-1.64
 0.037
 1.29
 −0.028
 −9.37
Higher education [ref lower]
 0.80
 0.62-1.03
 0.084
  
  
  
Race/ethnicity [ref white]
 2.63
 1.27-4.02
 <0.001
 1.33
 −0.003
 −0.95
Annual household income <15,000 [ref 15,000 - <35,000]
 1.10
 0.89-1.35
 0.398
 1.31
 −0.013
 −4.52
Annual household income ≥ 35,000 [ref 15,000 - <35,000]
 0.60
 0.39-0.91
 0.016
  
  
  
Health
  
  
  
  
  
  
More energetic [ref same energy]
 0.74
 0.59-0.93
 0.009
 1.14
 −0.155
 −54.25
Less energy [ref same energy]
 2.09
 1.65-2.66
 <0.001
  
  
  
Health status [ref ‘no’]
 2.36
 1.88-2.95
 <0.001
 1.19
 −0.114
 −39.90
One/some health problems [ref ‘none’]
 1.70
 1.38-2.10
 <0.001
 1.26
 −0.055
 −19.20
Lifestyle
  
  
  
  
  
  
Inactive/sedentary [ref moderately active]
 0.95
 0.74-1.23
 0.698
 1.31
 −0.013
 −4.54
Active [ref moderately active]
 0.72
 0.57-0.89
 0.003
  
  
  
Smoking [ref no]
 1.17
 0.95-1.44
 0.137
 1.32
 −0.006
 −2.21
Additional analysis5
  
  
  
  
  
  
BMI [kg/m2]
 1.12
 1.11-1.14
 <0.001
 1.29
 0.013
 5.38
BMI category ≤ 18.5 [ref 18.5-24.9]
 2.19
 0.95-5.06
 0.065
 1.29
 0.009
 3.61
BMI category 25.0-29.9 [ref 18.5-24.9]
 2.09
 1.55-2.81
 <0.001
  
  
  
BMI category ≥ 30 [ref 18.5-24.9] 6.13 4.61-8.14 <0.001      

1 The HR of each individual covariate is reported, adjusted for high distress and age. Health status and health problems are additionally adjusted for time dependent interactions (not shown).

2 The ‘HRcovariate adjusted’ is the HR of high distress, adjusted for age and each covariate independently.

3 Change in the strength (logHR) of association of high distress before and after each covariate adjustment = Change in logHR = (Bcrude model – Bnew model), and B = logHR.

4 The percentage change = 100*(Change in logHR)/Bcrude model. Absolute changes >1% were considered relevant, and included in subsequent multivariate regression models.

5 Information on BMI was not available until wave 13, and separate analysis were done for BMI, based on 452 diabetes events and 4780 cases. The age adjusted HR of high distress was 1.28 (95% CI = 1.03-1.57, p = .023) in that sample size.