Table 6.
Multiple regression analysis of predictive variables
Variable | Adjusted odds ratio and 95% confidence interval | Adjusted p-value |
---|---|---|
Age# | ||
≥ 43 | 1.00 | |
< 43 | 0.48 (0.23 to 1.02) | 0.404 |
Sex | ||
Female | 1.00 | |
Male | 1.62 (0.83 to 3.20) | 0.161 |
Ethnicity | ||
White | 1.00 | |
Non-white | 0.41 (0.18 to 0.96) | 0.039 |
Marital Status | ||
Married/living with partner | 1.00 | |
Single/divorced/separated/widowed | 1.93 (0.98 to 3.79) | 0.056 |
Physical Activity | ||
Low | 1.00 | |
Medium | 1.76 (0.80 to 3.85) | |
High | 1.43 (0.55 to 3.84) | 0.258 |
Age at leaving full time education# | ||
18 years or under | 1.00 | |
Over 18 | 1.54 (0.77 to 3.09) | 0.120 |
Zung depression# | ||
≥ 23 | 1.00 | |
< 23 | 1.97 (0.80 to 4.83) | 0.975 |
Modified Somatic Perception score# | ||
≥ 8 | 1.00 | |
< 8 | 0.75 (0.33 to 1.70) | 0.876 |
In paid employment | ||
No | 1.00 | |
Yes | 0.98 (0.45 to 2.17) | 0.967 |
Duration of current episode | ||
3 months or less | 1.00 | |
Over 3 months | 0.55 (0.24 to 1.28) | 0.438 |
Episodes over past 12 months | ||
Continuous/on-off | 1.00 | |
Episodic | 2.64 (1.25 to 5.60) | 0.005 |
Length of usual episode | ||
12 weeks or less | 1.00 | |
Over 12 weeks | 1.03 (0.43 to 2.50) | 0.587 |
Days of pain in last month | ||
28 days or more | 1.00 | |
Less than 28 days | 1.02 (0.45 to 2.33) | 0.730 |
Von Korff Pain Scale Grade: |
||
0 pain free | - | |
I low disability/low intensity | 1.00 | |
II low disability/high intensity | 0.33 (0.131 to 0.86) | |
III high disability/-moderately limiting | 0.97 (0.36 to 2.60) | |
IV high disability/severely limiting | 0.19 (0.04 to 0.92) | 0.354 |
SF-36 Physical Functioning# | ||
≥ 50 | 1.00 | |
< 50 | 0.54 (0.20 to 1.41) | 0.301 |
Baseline Roland Morris# | ||
≥ 11 | 1.00 | |
< 11 | 1.95 (0.78 to 4.85) | 0.249 |
Odds Ratios with 95% confidence interval for recovery including variables with unadjusted p-values of less than 0.1 in the univariate analysis, adjusted for all variables in table.
Variables denoted with "#" have been dichotomised from a continuous scale in relation to the relevant median.