Table 3.
Body mass index trajectories and single and multi-site musculoskeletal pain.
| Predictor |
Outcome |
||||
|---|---|---|---|---|---|
|
Follow-up Year 1 | |||||
| MSK pain | BMI trajectoriesaOR (95%CI) |
||||
| Slightly overweight | Lower overweight-to-obese | Upper overweight-to-obese | Lower obese | Upper obese | |
| Back pain | 1.33 (0.93-1.90) | 1.58 (1.08-2.32) | 1.40 (0.89-2.22) | 1.20 (0.66-2.18) | 2.51 (0.68-9.28) |
| Hand pain | 0.84 (0.57-1.25) | 1.16 (0.77-1.74) | 1.08 (0.66-1.76) | 0.81 (0.42-1.56) | 0.99 (0.27-3.59) |
| Knee pain | 1.18 (0.79-1.77) | 1.93 (1.28-2.93) | 2.55 (1.58-4.11) | 1.49 (0.78-2.82) | 3.39 (1.00-11.49) |
| Two sites | 1.00 (0.68-1.49) | 1.54 (1.03, 2.32) | 1.75 (1.08, 2.83) | 1.04 (0.54, 2.01) | 1.64 (0.46, 5.81) |
| Three sites | 1.89 (0.93, 3.85) | 1.68 (0.78, 3.63) | 3.21 (1.48, 6.96) | 1.78 (0.62, 5.15) | 3.75 (0.68, 20.70) |
|
Follow-up Year 20 | |||||||
|---|---|---|---|---|---|---|---|
| MSK painbOR (95% CI) |
|||||||
| BMI trajectories | Back pain | Hand Pain | Hip pain | Knee pain | Two sites | Three sites | Four sites |
| Slightly overweight | 0.95 (0.55, 1.63) | 1.52 (0.96, 2.41) | 1.37 (0.82, 2.30) | 1.46 (0.92, 2.34) | 1.28 (0.78, 2.12) | 1.70 (0.94, 3.09) | 1.36 (0.55, 3.35) |
| Lower overweight-to-obese | 1.20 (0.68, 2.12) | 2.14 (1.30, 3.51) | 1.58 (0.91, 2.74) | 1.77 (1.07, 2.92) | 1.58 (0.93, 2.70) | 2.58 (1.39, 4.79) | 1.60 (0.63, 4.10) |
| Upper overweight-to-obese | 3.14 (1.57, 6.29) | 2.08 (1.14, 3.80) | 1.92 (0.99, 3.68) | 2.07 (1.13, 3.81) | 2.94 (1.46, 5.92) | 3.25 (1.56, 6.78) | 2.85 (1.00, 8.12) |
| Lower obese | 2.19 (0.76, 6.32) | 1.30 (0.56, 3.01) | 2.58 (1.03, 6.41) | 2.72 (1.12, 6.61) | 1.96 (0.72, 5.29) | 3.92 (1.37, 11.20) | 2.03 (0.45, 9.14) |
| Upper obese | 0.85 (0.06, 11.25) | - | 7.56 (0.65, 87.59) | 2.64 (0.23, 30.71) | - | 5.29 (0.33, 84.17) | 4.52 (0.31, 66.61) |
BMI, body mass index; MSK, musculoskeletal; OR, odds ratio; CI, confidence interval.
The number of observations in these models was 926 (per BMI trajectory-group: normal weight 277, then from slightly overweight to upper obese, 257, 204, 118, 58, 12, respectively). The models were created using nominal regression models with reference group normal weight (BMI trajectory groups 1 and 2) and adjusted for age, menopause status, the number of live births, smoking habits, alcohol drinking, physical activity, oral contraceptive pill use, hormone replacement therapy use, analgesic use, hysterectomy, cancer, fractures, orthopedic operations, and other major illness.
The number of observations in back pain model was 478 (per BMI trajectory-group: normal weight to upper obese, 148, 141, 113, 55, 18, 3, respectively); in hand pain 558 (per BMI trajectory-group: normal weight to upper obese, 175, 160, 124, 67, 28, 4, respectively); in hip pain 557 (per BMI trajectory-group: normal weight to upper obese, 176, 158, 124, 68, 27, 4, respectively); in knee pain 564 (per BMI trajectory-group: normal weight to upper obese, 178, 162, 123, 69, 28, 4, respectively); in two musculoskeletal pain sites 511 (per BMI trajectory-group: normal weight to upper obese, 154, 149, 118, 62, 24, 4, respectively); in three sites 482 (per BMI trajectory-group: normal weight to upper obese, 146, 140, 115, 57, 21, 3, respectively); and 470 in four sites model (per BMI trajectory-group: normal weight to upper obese, 144, 138, 113, 54, 18, 3, respectively). The models were created using binary logistic regression estimating odds of single painful site vs. no pain, two painful sites vs. single painful site or no pain, three painful sites vs. two, one or none painful site, and four vs. three, two, one or none painful site. All models were adjusted for age, the number of live births, physical activity, oral contraceptive pill use, hormone replacement therapy use, analgesic use, hysterectomy, cancer, fractures, orthopedic operations, and other major illness.
Significant results are bolded.