Which Shift Schedules Are Linked to Back Pain?
Across 40 observational studies the night-shift subgroup carries the signal, while the rotating-shift result flips depending on which workers you compare it against.
Reviewed against primary sources on July 25, 2026 by the Soon operations research team. How we vet the evidence
The evidence in one line
A meta-analysis of 40 observational studies covering 1,839,258 participants found shift work associated with higher odds of low back pain (OR 1.31, 95% CI 1.18 to 1.47, p < 0.00001), with the night-shift subgroup at OR 1.49 (95% CI 1.24 to 1.82) and the rotating-shift subgroup showing no detectable association (OR 0.96, 95% CI 0.76 to 1.22, p = 0.49) (Chen et al., 2023). This is correlational evidence: 39 of the 40 studies were cross-sectional or case-control, so schedule and pain were largely measured at the same time and no causal claim is supported. The measured signal sat in health care (OR 1.40, 95% CI 1.20 to 1.63) and was not statistically significant outside it (OR 1.19, 95% CI 0.94 to 1.50).
The pooled number sits on the odds scale
Chen et al. (2023) pooled 40 observational studies covering 1,839,258 participants and found shift work associated with low back pain at OR 1.31 (95% CI 1.18 to 1.47, p < 0.00001). An odds ratio of 1.31 means 31% higher odds of reporting low back pain in the pooled comparison. It does not mean 31% more cases, and it does not describe the chance that a named worker will develop back pain. The random-effects pooling drew on PubMed, Embase, and Web of Science, and the 40 studies yielded 51 separate odds ratios, so several estimates come from the same study and are not fully independent.
The design mix sets the ceiling on interpretation. Thirty-one of the included studies were cross-sectional, eight were case-control, and one was prospective, so in almost every case the schedule and the pain were recorded at the same moment. The authors use associational language throughout and call for more rigorous work on the timing of shift work in relation to low back pain. Low back pain was mostly captured by questionnaire, which Chen et al. (2023) note lacks objectivity without clinical confirmation, and the included studies defined shift work differently: night shift only, a three-shift system, or occasional shift work.
Selection runs in both directions here. Workers who develop persistent back pain may move off night work, which would pull a cross-sectional estimate toward the null and make the pooled figure conservative. Workers with existing vulnerability may also be placed on, or moved off, particular rotations for reasons the underlying studies did not record. Neither pattern can be ruled out with data of this design, which is why the pooled figure is a starting point for investigation rather than a settled effect.
Nights carry the signal, rotation depends on the comparator
The subgroups split sharply. Night shift work was associated with OR 1.49 (95% CI 1.24 to 1.82, p < 0.0001), while the rotating-shift subgroup came in at OR 0.96 (95% CI 0.76 to 1.22, p = 0.49), a confidence interval that crosses 1 and therefore shows no detectable association in that pool (Chen et al., 2023). The tempting next step is to announce that rotation is safe for backs. That step is not supported.
A separate meta-analysis of 18,199 nurses points the other way. Chang and Peng (2021) compared 12,786 rotating and irregular shift nurses with 5,413 fixed day shift nurses and found back pain at OR 1.40 (95% CI 1.19 to 1.64, p < 0.001), with substantial heterogeneity (Q = 79.27, p < 0.001, I-squared = 57.11%). Back pain was the only site that reached significance in that review. Neck pain (n = 1,818), shoulder and upper limb pain (n = 2,525), and hip and lower limb pain (n = 1,894) did not, so the finding should not be widened into a general musculoskeletal claim.
The two results are not directly comparable, and saying so is more honest than picking a winner. The rotating subgroup in Chen et al. (2023) sits inside a mixed-industry pool, while Chang and Peng (2021) compare rotating plus irregular nurses against fixed day nurses inside a single profession, and they bundle irregular schedules in with rotating ones. Changing the comparison group changes what the odds ratio is measuring. The defensible reading is that the night-shift association holds up across this evidence while the rotating-shift association remains unresolved.
The measured association is concentrated in health care
Split by sector, health care workers showed OR 1.40 (95% CI 1.20 to 1.63, p < 0.0001) while non-health-care workers showed OR 1.19 (95% CI 0.94 to 1.50, p = 0.14), an interval that crosses 1 (Chen et al., 2023). Read plainly, the association was detectable in health care and was not statistically significant outside it. Carrying the headline 1.31 into warehousing, manufacturing, hospitality, or contact centers presents an untested claim as an established one.
Whether that gap reflects a real difference in physical demand or a difference in who was studied is unsettled. Chen et al. (2023) note that differences in study design, study populations, and occupations may themselves explain the variation between subgroups, and health care is heavily represented in the underlying literature. The pooled population is adult employees of companies and organizations, so neither source is a general-population sample and neither is US-specific.
On source quality, the primary paper is peer reviewed, PubMed indexed, and archived in PMC under CC BY 4.0, and the journal that published it was delisted from Web of Science in 2023. A reader can check each of those facts at the record itself, and none of them disqualifies the citation. Chang and Peng (2021) searched literature published from 2010 to 2020, so neither source captures evidence from 2021 onward.
What this means for your schedule
- Treat night work, rather than rotation in general, as the schedule feature with the most consistent back-pain association in this evidence.
- Brief managers on the odds scale and say so out loud, because 31% higher odds of reporting low back pain is not 31% more cases.
- Stop short of applying the pooled 1.31 figure to a non-health-care workforce, where the estimate was OR 1.19 with a confidence interval that crosses 1.
- Ask which comparison group any rotating-shift claim uses before acting on it, since the two meta-analyses here disagree once the comparator changes.
- Measure back-pain reports and physical task load in your own operation before and after a schedule change, because 39 of the 40 pooled studies cannot tell you what a change would do.
The business case
If a schedule review has to start somewhere, start with night work: the night-shift subgroup carries the largest association in this literature at OR 1.49 (95% CI 1.24 to 1.82), against a pooled OR of 1.31 and a rotating-shift estimate of OR 0.96 (Chen et al., 2023).
Because the evidence is observational and almost entirely cross-sectional, it supports monitoring and a controlled local pilot, not a promised change in claims or absence.
The measured signal sits in health care; extending it to other sectors would be an assumption rather than a finding.
Frequently asked questions
- Does shift work cause back pain?
- Chen et al. (2023) found shift work associated with low back pain at OR 1.31 (95% CI 1.18 to 1.47), but 39 of the 40 pooled studies recorded the schedule and the pain at the same moment, so the sorting can run in either direction. A worker whose back gives out may already have moved off nights before the survey reached them, which would understate the association, and a worker known to be vulnerable may have been kept off certain rotations for reasons no included study recorded. Neither pattern can be ruled out with this design, so the pooled figure documents an association rather than a cause.
- Are rotating shifts easier on the back than nights?
- The two meta-analyses disagree. Chen et al. (2023) found no significant association for the rotating-shift subgroup (OR 0.96, 95% CI 0.76 to 1.22, p = 0.49), while Chang and Peng (2021) found rotating and irregular shift nurses had higher odds of back pain than fixed day shift nurses (OR 1.40, 95% CI 1.19 to 1.64). The comparison groups differ, so neither result settles the question.
- Does the 1.31 figure apply to a warehouse or a contact center?
- The pooled OR 1.31 that Chen et al. (2023) reported for shift work and low back pain is an average across sectors, and it comes apart once the sectors are separated. The health care subgroup reached OR 1.40 (95% CI 1.20 to 1.63, p < 0.0001), while the non-health-care subgroup was OR 1.19 with a confidence interval of 0.94 to 1.50, which crosses 1 and is not statistically significant. Applying 1.31 to a warehouse or a contact center would import a health care result into settings where the same comparison did not reach significance.
- What does an odds ratio of 1.31 actually mean here?
- It means the odds of reporting low back pain were 31% higher among shift workers than among the comparison workers pooled by Chen et al. (2023). Odds are not risk and not case counts, and the figure describes a comparison between groups rather than the probability that any individual worker will develop back pain.
Sources
Every figure on this page is drawn from a cited primary source and checked against the original publication.
All 2 sources below are evidence syntheses: reviews that pooled many underlying studies before we cited them.
Chen, H.-M., Huang, P.-Y., Chuang, H.-Y., Wang, C.-L., Yang, C.-C., Huang, P.-J., & Ho, C.-K. (2023). Association of Low Back Pain with Shift Work: A Meta-Analysis. International Journal of Environmental Research and Public Health, 20(2), 918. https://doi.org/10.3390/ijerph20020918
Design: PRISMA systematic review and random-effects meta-analysis of 40 observational studies (31 cross-sectional, 8 case-control, 1 prospective; 1,839,258 participants; 51 odds ratios)
Chang, W.-P., & Peng, Y.-X. (2021). Differences between fixed day shift nurses and rotating and irregular shift nurses in work-related musculoskeletal disorders: A literature review and meta-analysis. Journal of Occupational Health, 63(1), e12208. https://doi.org/10.1002/1348-9585.12208
Design: Meta-analysis of observational studies of 18,199 nurses (12,786 rotating and irregular shift, 5,413 fixed day shift; literature published 2010 to 2020)
Cite these sources: BibTeX RIS
Why this page is graded moderate evidence
A consistent systematic review or meta-analysis at a lower grade, or a large observational study whose authors disclaim causality.
Who reviewed this
Every article in this library is checked against its primary sources by the Soon operations research team: each figure is traced back to the study it came from, and the wording is checked against the study design before publication. What that review covers
None of the studies cited here evaluated Soon.They examine scheduling practices, shift patterns, and working hours as studied by independent researchers, so their findings describe what those practices are associated with, not what any particular software produces.
This article summarizes published research for scheduling and operations decisions. It is not medical advice. Individual health questions belong with a qualified clinician.
Keep reading
Do Shift Work and Long Hours in Pregnancy Affect Birth Outcomes?
Sixty-two observational studies point the same direction on preterm birth and birth weight, on evidence the authors themselves graded low to very low.
Read →Does Night Shift Work Shorten Your Life?
Twenty-two years of Nurses' Health Study data show an association after about six years on rotating nights, and no life-expectancy figure anywhere in the paper.
Read →Shift Work and Health: What the Evidence Shows
A map of where the research on shift work and health is strong, where it is weak, and how to tell the difference.
Read →