Mental Health of Adults Treated in Adolescence with Scoliosis-Specific Exercise Program or Observed for Idiopathic Scoliosis
Table 5
Multiple regression analysis for BDI (the zero inflation model) and GHQ-28 (the logit model for ordinal variables).
Total BDI score
Parameter estimate
Standard error
P
Odds ratio
95% CI
Gender (male)
−.18
.08
<.05*
.83
.71–.98
Scoliosis (<10° Cobb)
−.16
.08
<.05*
.84
.72–1.00
Intervention (observation)
.15
.08
.06
1.16
.90–1.38
Marital status (single)
−.21
.08
<.01*
.82
.68–.96
Employment (employed)
−.35
.13
<.01*
.80
.54–.92
Education (lower level)
−.35
.05
<.001*
.70
.62–.79
Scoliosis in immediate family (occurring)
−.11
.08
.15
.89
.75–1.06
GHQ-28 domain A*: somatic symptoms
Parameter value
Standard error
Odds ratio
95% CI
Age (older)
−.02
.03
.55
.97
.91–1.05
Gender (male)
−.56
.30
1.87
.56
.31–1.02
Scoliosis (<10° Cobb)
.06
.29
.43
1.13
.64–2.03
Intervention (observation)
.18
.30
.59
1.19
.66–2.18
Marital status (single)
.37
.30
1.21
1.45
.80–2.67
Employment (unemployed)
−.64
.45
1.33
.52
.21–1.28
Education (lower level)
−.64
.24
2.62
.52
.33–.85
Scoliosis in immediate family (occurring)
−.003
.32
.01
.99
.53–1.87
For GHQ-28 global total and domains B, C, and D no associations were found; thus, results are not shown. For BDI was not calculated, as the zero inflated model of regression analysis was used. For GHQ-28 McFadden pseudo . Significant difference.