TABLE 3—

Summary of Individual Participant Data Odds Ratio Modification Analyses for Algorithmic Translation of Mechanical Exposure of Individual-Level Variables in Studies of Mechanical Workplace Risk Factors and Low Back Pain

	LBP in the Past 6 Months to 2 Years								Current LBP			Sick Leave Attributable to LBP
Variables	Posture 1A	Posture 1B	Posture 1C	Posture 2A	Posture 2B	Posture 2Ca	Force 1Ca	Force 2C	Posture 2A	Posture 2B	Posture 2Cab	Posture 2Aab	Posture 2B	Posture 2C	Force 1C	Force 2C
Studies, no.c	4	3	7	5	5	14	14	11	2	4	8	3	5	7	7	7
Age, y, OR
< 30	1.08		0.97	1.09									1.21
30–< 50	1.25		1.40	1.25									1.15
≥50	1.48		1.64	1.47									1.51
Sex, OR
Female	1.09			1.09							1.56				1.60
Male	1.40			1.38							1.89				2.17
Current smoker, OR						d	d
No											1.33		1.09	1.90
Yes											0.86		1.37	3.17
Ever smoker, OR
No			1.42								1.20
Yes			1.26								1.00
BMI, OR
< 30								1.49	1.19				1.12
≥ 30								1.16	1.55				1.29
Height,e OR
Low							1.80		1.08
High							1.62		1.27
Weightf
Low								1.46					1.13
High								0.94					1.38
Job satisfaction,g OR
Low	1.14			1.14			1.04									4.64
High	1.27			1.27			1.29									1.60
Psychological job demands,h OR
Low									0.88	1.08
High									1.27	1.28
Job control,i OR
Low			1.37			1.82				1.11
High			1.24			1.33				1.26

Note. BMI = body mass index (defined as weight in kilograms divided by the square of height in meters); LBP = low back pain; OR = odds ratio. Posture 1 = trunk flexion; posture 2 = nonneutral posture; force 1 = lifting; force 2 = heavy lifting; “A” exposures represent the percentage of a workday spent in a nonneutral posture or exerting a force; “B” exposures represent the percentage of job tasks requiring a nonneutral posture or exerting a force; “C” exposures represent ever experiencing a nonneutral posture or exerting a force.

^aThe outcome–exposure combinations with heterogeneity indicated in aggregate data analysis.

^bMeta-regression not done because the number of studies was < 10.

^cAnalyses included all studies of the variable of interest. The total number of studies included in each analysis may vary.

^dA variable identified as an OR modifier in the aggregate data analysis but not in the individual participant data analysis.

^eLow < 160 cm for women or < 175 cm for men; high ≥ 160 cm for women or ≥ 175 cm for men.

^fLow < 69 kg for women or < 89 kg for men; high ≥ 69 kg for women or ≥ 89 kg for men.

^gLow = not satisfied or very dissatisfied; high = satisfied or very satisfied.

^hLow < 50% of possible score; high ≥ 50% of possible score.

ⁱLow < 50% of possible score; high ≥ 50% of possible score.