Association between hyperglycaemia and adverse perinatal outcomes in south Asian and white British women: Analysis of data from the Born in Bradford cohort

Background: Diagnosis of gestational diabetes predicts risk of infants who are large for gestational age (LGA) and with high adiposity, which in turn aims to predict a future risk of obesity in the offspring. South Asian women have higher risk of gestational diabetes, lower risk of LGA, and on average give birth to infants with greater adiposity than do white European women. Whether the same diagnostic criteria for gestational diabetes should apply to both groups of women is unclear. We aimed to assess the association between maternal glucose and adverse perinatal outcomes to ascertain whether thresholds used to diagnose gestational diabetes should differ between south Asian and white British women. We also aimed to assess whether ethnic origin affected prevalence of gestational diabetes irrespective of criteria used. Methods: We used data (including results of a 26-28 week gestation oral glucose tolerance test) of women from the Born in Bradford study, a prospective study that recruited women attending the antenatal clinic at the Bradford Royal Infirmary, UK, between 2007 and 2011 and who intended to give birth to their infant in that hospital. We studied the association between fasting and 2 h post-load glucose and three primary outcomes (LGA [defined as birthweight >90th percentile for gestational age], high infant adiposity [sum of skinfolds >90th percentile for gestational age], and caesarean section). We calculated adjusted odds ratios (ORs) and their 95% confidence intervals (CIs) for a 1 SD increase in fasting and post-load glucose. We established fasting and post-load glucose thresholds that equated to an OR of 1·75 for LGA and high infant adiposity in each group of women to identify ethnic-specific criteria for diagnosis of gestational diabetes. Findings: Of 13 773 pregnancies, 3420 were excluded from analyses. Of 10 353 eligible pregnancies, 4088 women were white British, 5408 were south Asian, and 857 were of other ethnic origin. The adjusted ORs of LGA per 1 SD fasting glucose were 1·22 (95% CI 1·08-1·38) in white British women and 1·43 (1·23-1·67) in south Asian women (pinteraction with ethnicity = 0·39). Results for high infant adiposity were 1·35 (1·23-1·49) and 1·35 (1·18-1·54; pinteraction with ethnicity=0·98), and for caesarean section they were 1·06 (0·97-1·16) and 1·11 (1·02-1·20; pinteraction with ethnicity=0·47). Associations between post-load glucose and the three primary outcomes were weaker than for fasting glucose. A fasting glucose concentration of 5·4 mmol/L or a 2 h post-load level of 7·5 mmol/L identified white British women with 75% or higher relative risk of LGA or high infant adiposity; in south Asian women, the cutoffs were 5·2 mmol/L or 7·2 mml/L; in the whole cohort, the cutoffs were 5·3 mmol/L or 7·5 mml/L. The prevalence of gestational diabetes in our cohort ranged from 1·2% to 8·7% in white British women and 4% to 24% in south Asian women using six different criteria. Compared with the application of our whole-cohort criteria, use of our ethnic-specific criteria increased the prevalence of gestational diabetes in south Asian women from 17·4% (95% CI 16·4-18·4) to 24·2% (23·1-25·3). Interpretation: Our data support the use of lower fasting and post-load glucose thresholds to diagnose gestational diabetes in south Asian than white British women. They also suggest that diagnostic criteria for gestational diabetes recommended by UK NICE might underestimate the prevalence of gestational diabetes compared with our criteria or those recommended by the International Association of Diabetes and Pregnancy Study Groups and WHO, especially in south Asian women. Funding: The National Institute for Health Research