Table 3.
Multivariable binary logistic regression models used to identify participants' sociodemographic and clinical characteristics associated with trajectories of pain intensity (n = 230).
| Variables | Categories | Stable severe and moderate vs other trajectories | U-shape vs other trajectories | Inverted U-shape vs other trajectories | Stable mild vs other trajectories | ||||
|---|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | ||
| Pre–COVID-19 variables | |||||||||
| Sex at birth | Female vs male | 0.7 (0.4, 1.6) | 0.464 | 2.5 (0.7, 8.8) | 0.351 | 1.2 (0.6, 2.6) | 0.645 | 0.4 (0.1, 1.8) | 0.251 |
| Low back pain been an ongoing problem over the past 6 mo | At least half the days in the past 6 mo vs less than half the days in the past 6 mo | 1.6 (0.7, 3.7) | 0.310 | 0.9 (0.4, 2.0) | 0.740 | 0.3 (0.6, 2.0) | 0.228 | ||
| Every day or nearly every day in the past 6 mo vs less than half the days in the past 6 mo | 3.3 (1.4, 7.7) | 0.007 | 0.5 (0.2, 1.2) | 0.131 | 1.0 (0.2, 4.7) | 0.989 | |||
| Low back pain spread down your leg(s) during the past 2 wk | Yes vs no/not sure | 0.6 (0.2, 1.8) | 0.391 | 0.1 (0.1, 0.7) | 0.022 | ||||
| Widespread pain (pain in most of the body) | Yes vs no | 2.7 (1.4, 5.2) | 0.004 | 0.6 (0.2, 1.8) | 0.369 | 0.1 (0.01, 0.6) | 0.013 | ||
| Opioid painkillers | Yes vs no | 0.4 (0.1, 1.0) | 0.057 | ||||||
| It is not really safe for a person with my low back problem to be physically active | Agree vs disagree | 0.5 (0.2, 1.4) | 0.200 | ||||||
| Feeling that my low back pain is terrible and never going to get any better | Agree vs disagree | 1.8 (0.8, 3.7) | 0.137 | ||||||
| Current smoker | Yes vs No | 0.8 (0.3, 2.3) | 0.692 | ||||||
| Body mass index ≥30 | ≥30 vs < 30 | 0.3 (0.1, 1.4) | 0.112 | ||||||
| Pain intensity before COVID-19 | Score from 0 = no pain to 10 = worst imaginable pain) | 1.5 (1.3, 1.9) | <0.001 | 0.8 (0.7, 0.9) | 0.014 | ||||
| Pain interference before COVID-19 | T score from 41.6 to 75.6 | 0.9 (0.8, 0.98) | 0.009 | ||||||
| Physical function before COVID-19 | T score from 22.9 to 56.9 | 1.1 (0.9, 1.2) | 0.278 | ||||||
| Onset outbreak variables | |||||||||
| Age in years | >50 vs ≤50 | 0.8 (0.4, 1.7) | 0.585 | 1.7 (0.6, 5.1) | 0.143 | 0.9 (0.4, 1.8) | 0.751 | 1.3 (0.3, 5.9) | 0.712 |
| Acute stress disorder | Score from 0 to 76 | 1.0 (1.0, 1.0) | 0.656 | 1.0 (1.0, 1.1) | 0.103 | 1.0 (0.9, 1.0) | 0.084 | 0.9 (0.9, 1.0) | 0.233 |
| Finding the COVID-19 pandemic stressful | Score from 0 (not at all) to 10 (extremely) | 1.0 (0.9, 1.2) | 0.697 | 0.9 (0.7, 1.1) | 0.290 | 1.1 (0.9, 1.3) | 0.539 | 1.3 (0.9, 1.9) | 0.110 |
| Finding lockdown measures associated with the COVID-19 pandemic stressful | Score from 0 (not at all) to 10 (extremely) | 1.1 (0.9, 1.3) | 0.327 | 0.9 (0.7, 1.2) | 0.561 | 1.0 (0.9, 1.1) | 0.892 | 1.0 (0.7, 1.3) | 0.814 |
| Time-variant variable | |||||||||
| Having COVID-19 at any point | Yes vs No | 0.5 (0.2, 1.2) | 0.109 | 3.9 (1.3, 11.3) | 0.013 | 0.8 (0.3, 1.9) | 0.571 | 1.0 (0.2, 4.6) | 0.965 |
| Model evaluation | |||||||||
| Hosmer and Lemeshow test | 0.184 | 0.915 | 0.103 | 0.067 | |||||
Significant associations were bolded. In the model of severe or stable moderate pain intensity trajectories, we forced age, sex, and finding the COVID-19 pandemic stressful and having COVID-19 at any point. In the model of U-shape pain intensity trajectories, we forced age, sex, acute stress disorder, finding the COVID-19 pandemic stressful, finding lockdown measures associated with the COVID-19 pandemic stressful, and having COVID-19 at any point. In the inverted U-shape pain intensity trajectory, and the slight pain intensity trajectory models, we forced age, sex, acute stress disorder, finding the COVID-19 pandemic stressful, finding lockdown measures associated with the COVID-19 pandemic stressful, and having COVID-19 at any point.