Skip to main content

Table 2 Factors associated with diabetes mellitus type 2 in multivariate analysis

From: Diabetes mellitus type 2 in urban Ghana: characteristics and associated factors

Parameter aOR for diabetes (95% CI) aOR for diabetes without hypertension (95% CI) aOR for diabetes with hypertension (95% CI)
Residence (Kumasi outskirts) 1.93 (1.27-2.93) 1.78 (1.12-2.84) 1.96 (1.16-3.31)
Triglycerides ≥ 1.695 mmol/l 1.83 (1.13-2.97) - 2.46 (1.40-4.32)
Increased waist-to-hip ratio a 2.63 (1.76-3.93) 2.92 (1.85-4.61) 2.93 (1.79-4.81)
Diabetes family history, positive 3.79 (2.60-5.51) 3.92 (2.56-6.01) 3.72 (2.35-5.88)
Type of main work (light) 0.44 (0.25-0.78) 0.50 (0.26-0.96) 0.30 (0.15-0.59)
Working time > 40 h/week 1.76 (1.20-2.57) 1.85 (1.19-2.87) 1.84 (1.15-2.95)
Illiteracy 1.95 (1.28-2.97) 2.27 (1.43-3.62) -
Unemployment 4.23 (2.33-7.65) 2.58 (1.30-5.12) 5.71 (2.95-11.06)
Crowded living condition b 2.78 (1.71-4.51) 2.59 (1.53-4.40) 3.58 (1.97-6.48)
Carbohydrate < 55% of energy 1.63 (1.12-2.35) - 2.02 (1.27-3.22)
  1. aOR, adjusted odds ratio from multivariate logistic regression. Age and gender were a priori included. All univariately associated variables were included and stepwise backward removal of insignificantly associated factors (P > 0.05) identified independently associated parameters. R2 for diabetes, 0.53; R2 for diabetes without hypertension, 0.39; R2 for diabetes with hypertension, 0.65. All univariately associated parameters from Table 1 (partially dichotomised) were included in a fully adjusted model. The type and degree of associations held true; each aOR changed < 44%. a, male, ≥ 0.90; female, ≥ 0.85; b, > 75th percentile of the number of people per household (n > 8)