News

Autism risks increase after in utero antidiabetic medication exposure


 

AT EASD 2014

References

VIENNA – Children whose mothers had gestational diabetes and required antidiabetic medications may face a 22% increased risk of developing an autism spectrum disorder.

About a third of that increase was significantly associated with very-early diabetes diagnosis and treatment, suggesting that early first trimester hyperglycemia may affect fetal brain development, Anny H. Xiang, Ph.D., said at the annual meeting of the European Association for the Study of Diabetes.

Dr. Anny Xiang 					Michele G. Sullivan/Frontline Medical News

Dr. Anny Xiang

“My speculation is that, since fetal brain development is critical in the first and early second trimesters, exposure to high glucose during that time might damage its development,” said Dr Xiang, a senior research scientist at the Kaiser Permanente Department of Research and Evaluation in Pasadena, Calif. “Later exposure, perhaps, is not as problematic,” for the developing brain.

Her longitudinal cohort study included 315,827 singleton infants who were born at 28-44 weeks’ gestation between 1995 and 2009. This included 290,792 who were not exposed to gestational diabetes (GDM), and 25,035 who were. Of these, 5,891 had been exposed to antidiabetic medications.

All of the mothers were enrolled in Kaiser Permanente Southern California at the time of the birth, and all the children were still in the health plan by ages 1-2 years. Women who had preexisting type 1 or 2 diabetes were excluded form the study.

In accordance with the literature, mothers with GDM had lower educational levels and lower household incomes. They were older and more of them were multiparous. They were more likely to have had preeclampsia or eclampsia with a prior pregnancy. Asians and Pacific Islanders were more likely to have GDM than were other ethnicities.

Women who required antidiabetic treatment were diagnosed a mean of 4 weeks earlier than those who did not (23 vs. 27 weeks). Gestational ages at delivery did not significantly differ between the groups (39 weeks), but the infants of mothers with GDM were significantly heavier than were those of mothers without it. Infants exposed to antidiabetic medications weighed a mean of 3,453 g, compared with 3,380 g in nonmedication exposed infants, and 3,309 g in infants not exposed to GDM.

In the Kaiser system, standard well-baby care includes neurodevelopmental assessments at 1 year, 18 months, 2 years, and annually thereafter. Any concerns prompt a referral to a pediatric specialty clinic. For this study, diagnosis of autism required two diagnostic codes, one of which had to be from such a clinic.

Overall, the incidence rate of an autism spectrum disorder was 1.8/1,000 person-years. It was highest among those who had been exposed to antidiabetic medications (2.75/1,000 person-years) – significantly higher than that of both the GDM nonexposed and the non-GDM groups (1.96 and 1.77 per 1,000 person-years, respectively).

A multivariate analysis adjusted for a number of maternal and fetal factors, including maternal age, parity, race/ethnicity, comorbidity, education, income, prior preeclampsia/eclampsia, the infant’s gender, and gestational age at GDM diagnosis.

In the unadjusted analysis, infants who were exposed to the medications were significantly more likely to develop an autism spectrum disorder than were GDM infants who were not exposed (hazard ratio, 1.45). Significance remained high when each of the other confounding factors were individually considered. The final adjusted analysis, which included all of these, found an increased risk of 22%; however, this just missed statistical significance with a P value of .06.

A second analysis examined only the medication-exposed infants. The unadjusted analysis found a 36% increased risk, compared with nonexposed GDM infants. Again, significance remained strong when the individual confounders were examined separately. But the 26% risk in the fully adjusted model just missed significance – again with P value of.06.

The final analysis was adjusted for all of the confounders, plus gestational age when the mother was diagnosed. The hazard ratio of 1.17 suggested that early diagnosis and treatment explained about one-third of the overall increased risk, Dr. Xiang said.

The study continues with an investigation of the possible effect of maternal obesity on the current findings, she added.

Kaiser Permanente sponsored the study. Dr. Xiang is an employee of the company.

Msullivan@frontlinemedcom.com

On Twitter @alz_gal

Recommended Reading

Hormonal birth control linked to gestational diabetes
MDedge ObGyn
Noninvasive assisted reproduction linked to increased risk of congenital malformations
MDedge ObGyn
Fertility treatment was associated with limited cancer risks
MDedge ObGyn
VIDEO: Vitrification makes oocyte banking a reality
MDedge ObGyn
ESHRE writes first premature ovarian insufficiency guidelines
MDedge ObGyn
Quick freezing oocytes and embryos transforms assisted reproduction
MDedge ObGyn
Hyperovulation before IVF linked to higher obstetric complication risk
MDedge ObGyn
Gestational diabetes and the Barker Hypothesis
MDedge ObGyn
Higher cholesterol levels linked to reduced fertility
MDedge ObGyn
Estrogen-containing OCs cloud measurement of ovarian reserve
MDedge ObGyn