Women with gestational diabetes are at increased risk of non-insulin dependent diabetes and their babies are at increased risk of adverse perinatal outcomes.1 These risks can be reduced by better detection and control of diabetes.2 Identifying risk factors for gestational diabetes may improve screening programmes. As low birth weight has been related to non-insulin dependent diabetes in elderly populations,3 we decided to investigate whether women's characteristics at birth could predict their subsequent risk of gestational diabetes.
Subjects, methods, and results
We used linked generation data from the medical birth registry of Norway to study all women born in 1967-84 who had given birth between 1988 and 1998. The registry is a compulsory reporting system and files used for analysis are anonymised. Although there were 141 107 women in the cohort, we excluded 2393 who were not singletons.
We compared the birth characteristics of women with and without self reported gestational diabetes in one or more pregnancies. Data were analysed in relation to categories of birth weight; the ponderal index at birth (m/g3×100); gestational age (excluding women who were considered misclassified)4; weight for gestational age; and whether the woman had a mother whose pregnancy had been complicated by diabetes (any type), pre-eclampsia, eclampsia, placental abruption, or hypertension. We also considered diabetes in relation to the women's age and parity and their mothers' age and parity when they were born. We calculated odds ratios obtained from logistic regression analyses in which we adjusted for the women's age and parity and their mothers' diabetes.
Altogether 498 of these women aged less than 32 reported gestational diabetes. Prevalence increased with age, from 1.5 per 1000 deliveries for women aged ⩽20 to 4.2 for women aged ⩾30 (odds ratio 2.8; 95% confidence interval 1.9 to 4.3). Parity increased the risk of gestational diabetes; age adjusted odds ratios (95% confidence intervals) for women with two, three, and four or more deliveries compared with one delivery were 1.5 (1.2 to 1.9), 1.9 (1.4 to 2.5), and 3.3 (2.1 to 5.1) respectively.
Women whose mothers had had diabetes during pregnancy were at increased risk of gestational diabetes (table). The table also shows that there were significant inverse trends in diabetes in relation to birth weight and weight for gestational age (P<0.001). The increased risks of gestational diabetes were 80%, 60%, and 40% in women whose birth weights were ⩽2500 g, 2500-2999 g, and 3000-3499 g respectively compared with women in the 4000-4500 g group. We observed similar findings in relation to categories of weight for gestational age. Birth weight and weight for gestational age are strongly related; the three highest birthweight categories occur primarily in the three highest categories of weight for gestational age. We therefore limited further analyses of both variables to women whose birth weight was less than 3500 g. The inverse trend in diabetes in relation to weight for gestational age remained significant (table , P<0.01), but the variation attributed to the truncated range of birth weight was not significant. No other variables examined were associated with diabetes.
Comment
Low birth weight or low weight for gestational age or having a mother who was diabetic during pregnancy increases the risk of gestational diabetes. In women who weighed less than 3500 g at birth, weight for gestational age may provide additional predictive information on risk. No other birth characteristics were predictive of gestational diabetes. The non-significant raised risk in women weighing 4500 g or more at birth could indicate undiagnosed or unrecorded maternal diabetes. Low birth weight and low weight for gestational age may be common risk factors for gestational diabetes and non-insulin dependent diabetes. The results are compatible with the fetal origins of disease hypothesis.5 Future studies combining birth information with risk factors in adulthood may improve predictive models for identifying women at risk of gestational diabetes.
Table.
Prevalence rates and adjusted odds ratios (95% CI) for self reported gestational diabetes in relation to women's own birth characteristics
| Birth characteristics | No of women (n=138 714) | Rate (per 1000 women) | Odds ratio (95% CI)† |
|---|---|---|---|
| Mother with diabetes during pregnancy: | |||
| No (referent) | 138 518 | 3.5 | 1.0 |
| Yes | 196 | 30.6 | 9.3 (4.1 to 21.1) |
| Birth weight (g): | |||
| <2500 | 4 652 | 4.9 | 1.8 (1.1 to 3.0)** |
| 2500-2999 | 18 948 | 4.3 | 1.6 (1.1 to 2.3) |
| 3000-3499 | 51 737 | 4.0 | 1.4 (1.0 to 2.0) |
| 3500-3999 | 45 524 | 3.0 | 1.1 (0.8 to 1.5) |
| 4000-4500 (referent) | 14 852 | 2.8 | 1.0 |
| >4500 | 2 640 | 4.2 | 1.5 (0.8 to 2.9) |
| Weight for gestational age (centiles): | |||
| <10 | 16 034 | 4.6 | 1.7 (1.2 to 2.5)** |
| 10-25 | 23 480 | 4.4 | 1.6 (1.1 to 2.3) |
| >25-50 | 36 413 | 3.6 | 1.3 (0.9 to 1.8) |
| >50-75 | 31 329 | 2.9 | 1.0 (0.7 to 1.5) |
| >75-90 (referent) | 15 576 | 2.8 | 1.0 |
| >90 | 10 125 | 3.4 | 1.2 (0.7 to 1.8) |
| Ponderal index at birth (g/cm3) ×100: | |||
| <2.5 | 25 752 | 3.5 | 1.0 (0.8 to 1.4) |
| 2.5-2.635 | 28 568 | 4.5 | 1.3 (1.0 to 1.7) |
| 2.635-2.752 (referent) | 27 441 | 3.4 | 1.0 |
| 2.752 to 2.89 | 28 441 | 3.4 | 1.0 (0.8 to 1.4) |
| ⩾2.89 | 27 809 | 3.2 | 0.9 (0.7 to 1.3) |
| Gestational age (weeks): | |||
| 28-34 | 1 414 | 5.7 | 1.6 (0.8 to 3.3) |
| 35-36 | 3 261 | 3.7 | 0.9 (0.5 to 1.7) |
| 37-39 | 39 010 | 3.4 | 1.0 (0.8 to 1.2) |
| 40 (referent) | 37 815 | 3.4 | 1.0 |
| 41-42 | 43 818 | 3.8 | 1.1 (0.9 to 1.4) |
| 43-44 | 5 652 | 3.7 | 1.1 (0.7 to 1.7) |
| Weight for gestational age in women <3500 g at birth (centiles): | |||
| <10 | 16 034 | 4.6 | 2.9 (0.7 to 12.0)* |
| 10-25 | 23 480 | 4.4 | 2.8 (0.7 to 11.5) |
| >25-50 | 25 803 | 3.6 | 2.2 (0.5 to 9.1) |
| >50-75 | 5 688 | 3.5 | 2.1 (0.5 to 8.9) |
| >75-90 (referent) | 975 | 2.1 | 1.0 |
| >90 | 272 | 3.7 | 1.3 (0.1 to 18.9) |
Because of missing variables, not all characteristics were included for all women.
Adjusted for women's age and parity and their mother's diabetic status.
P<0.01; **P<0.001.
Acknowledgments
Barbro Mork Emblem was instrumental in linking the generational birth information and in setting up the generational analytical database.
Footnotes
Funding: Medical Birth Registry of Norway and Locus for Registry-based Epidemiology, University of Bergen.
Competing interests: None declared.
References
- 1.Peters RK, Kjos SL, Xiang A, Buchanan TA. Long-term diabetogenic effect of single pregnancy in women with previous gestational diabetes mellitus. Lancet. 1996;347:227–230. doi: 10.1016/s0140-6736(96)90405-5. [DOI] [PubMed] [Google Scholar]
- 2.De Veciana M, Major CA, Morgan MA, Asrat T, Toohey JS, Lien JM, et al. Postprandial versus preprandial blood glucose monitoring in women with gestational diabetes mellitus requiring insulin therapy. N Engl J Med. 1995;333:1237–1241. doi: 10.1056/NEJM199511093331901. [DOI] [PubMed] [Google Scholar]
- 3.Lithell HO, McKeigue PM, Berglund L, Mohsen R, Lithell UB, Leon DA. Relation of size at birth to non-insulin-dependent diabetes and insulin concentrations in men aged 50-60 years. BMJ. 1996;312:406–410. doi: 10.1136/bmj.312.7028.406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Melve KK, Skjærven R, Gjessing HK, øyen N. Recurrence of gestational age in sibships: implications for perinatal mortality. Am J Epidemiol. 1999;150:756–762. doi: 10.1093/oxfordjournals.aje.a010078. [DOI] [PubMed] [Google Scholar]
- 5.Barker DJP. Fetal origins of coronary heart disease. BMJ. 1995;311:171–174. doi: 10.1136/bmj.311.6998.171. [DOI] [PMC free article] [PubMed] [Google Scholar]