Table 2

Multiple logistic regressions for predicting postpartum glucose intolerance among women with GDM

Predictor variablesUnivariable analysisMultivariable analysisSimplified risk score
COR (95% CI)AOR (95% CI)P value
Maternal age (≥35 years)2.88 (1.111.7.46)4.04 (1.23 14.33)0.023804
Gravidity (multigravida)*3.00 (1.027 10.989)1.75 (0.39 8.66)0.47196
History of GDM1.45 (0.45 4.19)NA
Family history of DM1.85 (0.63 5.11)NA
MUAC (≥28 cm)6.48 (2.474 18.61)3.92 (1.13,15.04)0.036174
Blood pressure (mm Hg)
Systolic blood pressure0.998 (0.976 1.054)NA
Diastolic blood pressure1.015 (0.976 1.053)NA
Anaemia (haemoglobin <11 g/dL)0.34 (0.05 1.30)NA
Blood glucose level (mg/dL)
FPG at GDM diagnosis1.07 (1.03 1.13)1.08 (1.04 1.15)0.001711
1-hour PG at GDM diagnosis1.014 (0.99 1.03)NA
2-hour PG at GDM diagnosis0.99 (0.97 1.02)NA
Gestational age at diagnosis (weeks)1.03 (0.89 1.18)NA
Level of physical activity
High1
Moderate0.87 (0.23 3.47)NA
Low0.57 (0.18 2.07)
Inadequate dietary diversity*3.66 (0.97 24.03)3.07 (0.58 24,45)0.22031
Antenatal depression2.78 (1.05 7.29)5.90 (1.66 23.47)0.007705
Insulin-treated GDM1.51 (0.21 7.54)NA
  • *Gravidity and dietary diversity status were variables that were also retained in the reduced model using likelihood ratio test. Both backward and forward selection showed the same results. ORs after internal validation with bootstrapping are shown.

  • AOR, adjusted OR; COR, crude OR; DM, diabetes mellitus; FPG, fasting plasma glucose; GDM, gestational diabetes mellitus; MUAC, mid-upper arm circumference; NA, not included in the multivariate analysis .