Table 2

Contributions of the variables to 1-year SF-36 Mental Health scores




Correlation (r) with


Quantitative variables
Mean
STD
1 year SF36-MH
R2
p-value


Mean bodily pain at 3 and 6 month
2.03
1.11
-0.450

0.203
<0.0001
Baseline SF36-MH
61.62
18.15
0.436

0.190
<0.0001
Baseline bodily pain
3.9
0.99
0.118

0.014
0.1642
Age
42.84
10.07
0.042

0.002
0.623





1 year SF36-MH


Categorical variables
Category
n
Mean
STD




Sex
Male
86
74.88
20.34
0.001
0.724

Female
55
73.60
21.92


Workers' compensation
Yes
42
68.00
25.05
0.040
0.018

No
99
77.09
18.44


Crossover
0 month
120
76.83
18.54
0.079
0.001

3 or 6 month
21
60.38
28.03



R2 from the model including all above variables
0.435


Note: R2 was calculated using the model including only one variable.

In covariate selection, linear regression with backward selection method (p = 0.1) was applied, where 1 year SF-36 MH was the dependent variable; age, sex, workers' compensation, baseline bodily pain, prior episodes, and the SF-36 General Health, Vitality, Social Function, and Emotional Role subscales were independent variables; and age, sex and workers' compensation were fixed in the model.

141 of 162 surgical patients with both 1 year mood data and all 7 covariates and DNA were used in the 1 year mood analysis.

"Crossover" refers to the time point when a questionnaire showed that a patient who initially chose nonsurgical treatment underwent lumbar discectomy.

Max et al. Molecular Pain 2006 2:14   doi:10.1186/1744-8069-2-14