Table 3.

Prospective impact of psychosocial characteristics and previous back pain at baseline on sickness absence spells.

Characteristic	Age and sex adjusted (Model I)†			Adjusted for the final model (Model III)‡
	n	RR	95% CI	n	RR	95% CI
Occupation (SES)
Non-manual	184	1 [ref]	.	.	.	.
Manual	112	1.18	0.89-1.57	.	.	.
Self-employed	40	0.28	0.12-0.62	.	.	.
Smoking, current	331	0.79	0.55-1.15	.	.	.
BMI	339	1.01	0.97-1.05	.	.	.
Alcohol consumption
Lowa	263	1 [ref]
Mediumb	36	0.41	0.21-0.83	.	.	.
Highc	25	0.65	0.34-1.26	.	.	.
Disease (self report diagnosis)	340	1.38	1.04-1.83	.	.	.
Past sickness absence	340	1.36	1.21-1.51	.	.	.
I Work environment
Job strain	323	2.28	1.01-5.15	298	1.50	0.67-3.40
Psychological demands	326	1.02	0.96-1.08	299	1.03	0.97-1.09
Decision latitude	337	0.95	0.90-1.00	309	1.00	0.95-1.07
Social support at work	318	0.95	0.88-1.01	289	0.96	0.89-1.02
Overcommitment	329	1.00	0.97-1.03	300	1.02	0.98-1.05
II Emotional support	337	1.42	1.14-1.77	307	1.41	1.12-1.76
III Psychological resources
1. Sense of coherence	337	0.99	0.97-1.00	308	1.00	0.98-1.01
2. Mastery	330	0.96	0.92-1.01	300	0.99	0.94-1.04
3. Self-esteem	328	0.96	0.93-0.99	299	0.98	0.95-1.01
4. Perceived control	311	0.99	0.97-1.01	285	1.00	0.98-1.03
Previous back pain	329	0.94	0.69-1.28	301	0.83	0.61-1.13

The rate ratio in italics denotes significance (p < 0.05). (Result from Model II, see text). In model III rate ratios for confounders are not reported as separate regressions were run for each of the 10 psychosocial variables and previous back pain.

n = number of observations used; RR = rate ratio; CI = confidence interval.

^aLow (< 80 g/week for women, < 110 g/week for men)

^bMedium (between 80-139 g/week for women, 110-169 g/week for men)

^cHigh (≥ 140 g/week for women, ≥ 170 g/week for men)

^†Adjusted for age, sex; ^‡ age, sex, BMI, smoking, alcohol consumption, SES, disease (self report of diagnosis), and past sickness absence.