Hemoglobin A_1c Levels During Pregnancy and Risk of Autism Spectrum Disorders in Offspring

Abstract

This cohort study uses electronic medical record (EMR) data to examine the association between maternal hemoglobin A_1c levels during pregnancy and risk of autism spectrum disorder in offspring.

Maternal preexisting type 1, type 2, or gestational diabetes diagnosed relatively early in pregnancy are associated with increased risk for autism spectrum disorders (ASDs) in offspring.^1,2 This study extends previous observations by examining the association between maternal hemoglobin A_1c (HbA_1c) levels during pregnancy and risk of ASD in offspring.

Methods

This retrospective cohort study included singleton children born at 28 to 44 weeks’ gestation in Kaiser Permanente Southern California (KPSC) hospitals between January 1, 2012, and December 31, 2013. KPSC implemented HbA_1c screening in the early prenatal period for all pregnancies starting in 2012. Children who enrolled as KPSC health plan members at age 1 year were tracked through electronic medical records until the first of the following: (1) clinical diagnosis of ASD with at least 1 diagnostic code, (2) last date of continuous KPSC membership, (3) death, or (4) study end date of December 31, 2017. The KPSC institutional review board approved this study and provided waiver of participant consent.

The last HbA_1c level in the first 2 trimesters of pregnancy was obtained from the electronic laboratory database. HbA_1c was analyzed as a continuous variable and a categorical variable classified as less than 5.7%, 5.7% to 5.9%, 6.0% to 6.5%, and greater than 6.5%. Methods to identify ASD were previously described.^1,2 Cox regression was used to estimate hazard ratios of ASD associated with HbA_1c exposure. Kaplan-Meier plots were used to display the crude cumulative incidences of ASD by the HbA_1c categories. Potential confounders listed in the Table footnote “a” were evaluated. Interaction with gestational age at HbA_1c testing was also assessed. SAS Enterprise Guide version 7.1 (SAS Institute Inc) was used for data analysis. A 2-sided P < .05 was considered significant.

Table. Risk of ASD in Offspring Associated With Maternal HbA_1c Levels in Early Pregnancy^a.

HbA1c	ASD Cases/Total Children, No. (%)	Unadjusted Hazard Ratios (95% CI)	P Value	Adjusted Hazard Ratios (95% CI)b	P Value
As a continuous variable (per 1% increase)		1.23 (1.08-1.41)	.002	1.12 (0.96-1.31)	.15
As a categorical variablec
<5.7%	574/30 419 (1.89)	1 [Reference]		1 [Reference]
5.7%-5.9%	101/4172 (2.42)	1.24 (1.01-1.54)	.04	1.08 (0.87-1.35)	.46
6.0%-6.5%	17/871 (1.95)	0.98 (0.61-1.59)	.94	0.79 (0.48-1.28)	.33
>6.5%	15/357 (4.20)	2.17 (1.30-3.62)	.003	1.79 (1.06-3.00)	.03

Abbreviations: ASD, autism spectrum disorder (includes autistic disorders, Asperger syndrome, or pervasive developmental disorder not otherwise specified); GD, gestational diabetes; HbA_1c, hemoglobin A_1c; T1D, preexisting type 1 diabetes; T2D, preexisting type 2 diabetes.

^a The median gestational age at HbA_1c testing was 9.0 weeks (interquartile range, 6.7-11.4 weeks); 83% of the testing was in the first trimester. Potential confounders assessed included maternal age at delivery, prepregnancy body mass index, parity, education, self-reported race/ethnicity; gestational age at HbA_1c testing; presence of comorbidity (diagnosis of heart, lung, kidney, or liver disease or cancer) prior to pregnancy; presence of anemia or hemoglobinopathies, smoking, preeclampsia/eclampsia, depression, use of selective serotonin reuptake inhibitors during pregnancy; and child’s sex. None of them changed the model estimates by more than 10% except for maternal prepregnancy body mass index and race/ethnicity.

^b Estimated by Cox regression models adjusted for maternal race/ethnicity and prepregnancy body mass index (calculated as weight in kilograms divided by height in meters squared).

^c Distribution of diabetes status: overall: 0.4% T1D, 5.7% T2D, 9.3% GD, and 84.6% no diabetes; for HbA_1c greater than 6.5%: 19.9% T1D, 70.6% T2D, 8.1% GD, and 1.4% no diabetes; for HbA_1c between 6.0% and 6.5%: 2.6% T1D, 37.3% T2D, 31.3% GD, and 28.7% no diabetes; for HbA_1c between 5.7% and 5.9%: 0.3% T1D, 14.2% T2D, 17.4% GD, and 68.1% no diabetes; and for HbA_1c less than 5.7%: 0.1% T1D, 2.8% T2D, 7.6% GD, and 89.5% no diabetes.

Results

A total of 35 819 mother-infant pairs (51% boys) met the study criteria with complete HbA_1c and covariate information. The maternal mean (SD) age was 30.9 (5.6) years; the mean (SD) prepregnancy body mass index (calculated as weight in kilograms divided by height in meters squared) was 27.2 (6.4); and 51% were Hispanic, 24%, white, 7%, black, 14%, Asian/Pacific Islander, and 3%, other. The median gestational age at HbA_1c testing was 9.0 weeks (interquartile range, 6.7-11.4 weeks); 83% of the testing was in the first trimester. The median HbA_1c level was 5.4% (interquartile range, 5.2%-5.6%); 84.9% of pregnancies had HbA_1c levels less than 5.7%, 11.7% between 5.7% and 5.9%, 2.4% between 6.0% and 6.5%, and 1% greater than 6.5%. A total of 99% of women with HbA_1c levels greater than 6.5% had diabetes during pregnancy (Table footnote “c”).

Children were followed up for a median of 4.5 years (interquartile range, 4.0-5.0 years) after birth, during which time 707 (2.0%) had a clinical diagnosis of ASD. In the multivariable analysis, none of the potential confounders changed the model estimates by more than 10% except for maternal prepregnancy body mass index and race/ethnicity. After adjusting for these variables, the hazard ratio of ASD associated with each 1% increase of HbA_1c level was 1.12 (95% CI, 0.96-1.31; P = .15) for the continuous measure and 1.08 (95% CI, 0.87-1.35; P = .46) for HbA_1c level between 5.7% and 5.9%, 0.79 (95% CI, 0.48-1.28; P = .33) for HbA_1c level between 6.0% and 6.5%, and 1.79 (95% CI, 1.06-3.00; P = .03) for HbA_1c level greater than 6.5% relative to HbA_1c level less than 5.7% (Table; Figure). The risk associated with HbA_1c levels did not vary by gestational age at HbA_1c testing (P > .19 for interaction).

Figure. Crude Cumulative Incidences of Autism Spectrum Disorder in Offspring by Early Pregnancy Hemoglobin A_1c (HbA_1c) Exposure Status.

The median gestational age at HbA_1c testing was 9.0 weeks (interquartile range, 6.7-11.4 weeks); 83% of the testing was in the first trimester. Autism spectrum disorders include autistic disorders, Asperger syndrome, or pervasive developmental disorder not otherwise specified.

Discussion

In this study, there was no association between maternal HbA_1c levels during early pregnancy and ASD in offspring when HbA_1c levels were analyzed as a continuous variable or as a categorical measure if less than 6.5%. An association with HbA_1c levels greater than 6.5% was found but was based on only 15 affected children. These findings are consistent with previous observations and the preconception counseling recommendation to optimize glycemic control with HbA_1c levels greater than 6.5%.³

These results suggest that maternal glycemic control in early pregnancy may be important for ASD risk in offspring. Study limitations include the lack of information on maternal glycemic control throughout pregnancy, other prenatal and early life risk factors, paternal risk factors, and genetic factors as well as the relatively small sample size for pregnancies with HbA_1c levels greater than 6.5%.

Section Editor: Jody W. Zylke, MD, Deputy Editor.