Abstract
Context:
Lower neurocognitive development scores at age 2 yr have been reported in association with euthyroid hypothyroxinemia during early pregnancy.
Objective:
The objective of this study was to further explore this association with euthyroid hypothyroxinemia during early pregnancy.
Design:
This was an observational, nested case-control study.
Setting:
The study was conducted at physician offices and prenatal clinics throughout Maine.
Study Subjects:
Between May 2004 and March 2006, TSH was measured in 5734 women in conjunction with second-trimester Down syndrome screening. After completion of pregnancy, free T4 was measured in stored second-trimester sera from euthyroid women (TSH 0.1–3.5 mIU/ml; n = 5560). Women with free T4 at the third centile or less (n = 99) were matched with women whose free T4 was at the 10th to the 90th centile (n = 99).
Interventions:
There were no interventions.
Main Outcome Measure:
Bayley Scales of Infant Development (BSID III) were administered to the 198 offspring at age 2 yr. Scores for cognitive, language, and motor development were compared between matched pairs of offspring from the two groups before and after correcting for relevant variables.
Results:
Unadjusted BSID-III scores (cognitive, language, and motor) were lower by about 3% at age 2 yr among offspring of 98 hypothyroxinemic women (cases), reaching borderline significance for cognitive and motor scores. After adjustment for gestational age, the child's age at testing, maternal weight, and education, all differences diminished and became nonsignificant. Scores less than 85 were more frequent among case children but did not reach statistical significance (P = 0.14).
Conclusions:
Isolated hypothyroxinemia during the second trimester is not associated with significantly lower BSID-III scores at age 2 yr, compared with scores for offspring of matched euthyroxinemic women.
Fetal central nervous system development relies on T4 as a substrate for conversion to T3, the active form of thyroid hormone; T3 does not cross the fetal blood-brain barrier (1–3). Before 12 wk gestation, the mother is the sole source for T4, after which fetal production of T4 begins and slowly increases. In 1976, Man and Serunian (4) presented preliminary evidence that low butanol-extractable iodine levels in mothers' serum at 24 wk gestation were associated with impaired psychomotor development in the offspring at age 7 yr. Butanol extractable iodine, an indirect measure of thyroid hormones, was used before the availability of more refined assays (i.e. TSH, T4, and free T4). More recent studies, using these newer assays, again indicated that deficient maternal thyroid function can have an adverse impact on the offspring's neurocognitive development when a mother has either diagnosed (5) or undiagnosed (6) mild to moderate thyroid deficiency (as defined by elevated TSH) or isolated hypothyroxinemia associated with normal TSH (6–10). The present study further explores the extent to which isolated maternal hypothyroxinemia with normal TSH in midpregnancy in an iodine-sufficient region might be associated with the offspring's neurocognitive development by evaluating this relationship in the context of variables with known influence on child development.
Materials and Methods
Study subjects
The study protocol was approved by the Foundation for Blood Research (FBR) Institutional Review Board. Pregnant women presenting for second-trimester screening between May 2004 and March 2006 gave informed consent for the serum TSH measurement (at the time of serum screening for Down syndrome) and free T4 (in stored sera, after pregnancy completion). This initial recruitment phase formed a cohort of pregnancies that would be sufficiently large to identify and enroll 100 women with isolated hypothyroxinemia (euthyroid women with free T4 at the third centile or less) and 100 paired, matched euthyroxinemic women (euthyroid women with free T4 between the 10th and 90th centiles). Serum TSH values of the euthyroid women ranged between 0.26 and 3.34 mIU/liter. Recruitment pool size was modeled by taking into account the known distributions of TSH and free T4 in the Maine pregnant population, the proportion of infants expected to meet inclusion criteria, and the anticipated enrollment rate among women identified for further testing.
Birth record information was available through a data-sharing agreement between FBR and the Bureau of Vital Statistics, Maine Department of Health and Human Services. These data were linked to FBR's clinical patient records in two stages, approximately 1 yr apart, because offspring reached the age for psychometric testing over a 2-yr period. Women were excluded from the study if birth records or patient records indicated multiple gestation, delivery at less than 36 wk gestation, delivery of a baby with congenital anomalies, or a birth weight less than 2500 g.
At least three euthyroxinemic women were matched to each woman with isolated hypothyroxinemia. The seven matching criteria (in descending order of importance) were the child's sex (exact match required); maternal parity (exact match required); mother's years of education; child's age; mother's age; geographic location (northern, central, or southern Maine); and, among parous women, age of next oldest sibling (<2 yr, ≥ 2 yr). The euthyroxinemic women were ranked in order of match stringency for each potential subject with isolated hypothyroxinemia, using assigned weights for each variable that reflected distance from an exact match. Based on stringency, euthyroxinemic women were matched either for all seven criteria (n = 57), for six (n = 28), for five (n = 7), for four (n = 6), or for one (n = 1).
Women with isolated hypothyroxinemia were contacted by mail and follow-up phone call 1 month before their child's second birthday. Psychometric testing for the child was scheduled based on verbal consent; women gave written informed consent at the testing appointment for both the child's testing and access to their child's newborn thyroid screening data. Once a hypothyroxinemic woman was scheduled for testing, the contact and enrollment process began for a matched euthyroxinemic woman. Supplementary information about each child's health, including medical problems and hospitalizations, was systematically collected at the time of the visit for Bayley Scales of Infant Development, third edition (BSID III) testing. One matched pair was excluded from the present analyses due to a diagnosis of autism in a control child. No other cases of autism were reported in either cases or controls.
Laboratory measurements
Second-trimester maternal serum TSH levels were measured using an Immulite 2000 chemiluminescence analyzer (Siemens Healthcare Diagnostics, Deerfield, IL) with kits from the manufacturer. Maternal serum samples from euthyroid women were then stored at −20 C for the later measurement of free T4 on the same platform. For the BSID III testing to be scheduled after the second birthday, it was necessary to enroll the children before they reached 2 yr of age. This, in turn, required that free T4 be measured in two batches: the first when the first half of the recruitment pool had completed their pregnancies (samples collected between May 2004 and April 2005) and the second when the remaining women had completed their pregnancies (samples collected between April 2005 and March 2006). The third, 10th, and 90th centiles for free T4 were 0.92, 1.00, and 1.34 ng/dl, respectively, for 2928 samples from euthyroid women in the first stage and 0.97, 1.04, and 1.37 ng/dl, respectively, for 2632 samples from euthyroid women in the second.
Total T4 and antithyroid peroxidase antibodies (TPO; positive result defined as ≥ 35 IU/ml) were measured in stored second-trimester sera from the 198 women participating in the psychometric testing stage of the study, using the same platform. Neonatal total T4 results for this group of pregnancies were obtained by blood spot measurements performed at the University of Massachusetts as part of Maine's newborn screening program (reference range, total T4 > 5 μg/dl). The data were provided to FBR by the Maine Genetics Program.
Although it was not possible to assess iodine status in our study population, National Health and Nutrition Examination Survey data indicate that iodine status is generally acceptable in the United States (11). In an earlier analysis, we examined urine iodine data from nonpregnant adults in National Health and Nutrition Examination Survey III and determined that T4 concentrations were not diminished in the presence of urine iodine concentrations between the first and ninth centiles (12).
Neurocognitive testing
The BSID III (11) was administered to obtain standard scores for cognitive, language (receptive and expressive subtests), and motor (gross and fine subtests) scales. As described by Bayley (13), the BSID III uses standardized administration and scoring procedures to provide the infant and toddler with situations and tasks that capture his or her interest and that provide an observable set of behavioral responses. The cognitive scale assesses play skills, information processing, information processing speed, problem solving, and number concepts. The language scale assesses both receptive (ability to hear, understand, and respond) and expressive (ability to communicate) language skills and is important for identifying critical delays. The motor subtests assess quality of movement, sensory and perceptual motor integration, and basic locomotion milestones. The BSID III composite test scores are scaled on a range of 40–160, to have a mean of 100, and a sd of 15. Although not considered an intelligence quotient test, the Bayley Scales reliably identify infants with developmental delays, as indicated by scores less than 85 on the cognitive, language, or motor composite scores. Testing was performed by two individuals experienced in child developmental testing. Both the person scheduling the appointments and the testers themselves were blinded as to the thyroid status of the study subjects. Assignments for children to be tested were based on tester availability. All tests were double scored to ensure accuracy in calculating the age of the child, number of items passed, and conversion of raw scores to standard scores as published in the test manual. Two tests were administered to collect data about the child's environment: the Four Factor Index of Social Status (Hollingshead Scale) (14), which measures family socioeconomic status (calculated from the educational and occupational status of both parents), and the Home Screening Questionnaire (HSQ) (15), which assesses the child's environment without requiring a home visit.
Data analysis
Data were analyzed using statistical software from the SAS Institute, Inc. (Cary, NC) and Analyze-It for Excel (Microsoft Inc., Seattle, WA). Hypothyroxinemic women who enrolled were compared with those who did not enroll using the χ2 test, and the t test. Demographic and pregnancy-related data were compared between study subgroups, using the paired t test for continuous data and either the McNemar's test or Wilcoxon test, as appropriate, for categorical data. A linear model was fitted to predict the difference between each of the three BSID III scores for case/control sets, after controlling for gestational age, child's age at testing, maternal weight, and years of education. Significance was two sided at the P = 0.05 level. The 95% confidence intervals were computed for the observed and adjusted differences in BSID III scores. To determine whether controls with free T4 concentrations in the upper percentiles might score higher on BSID III testing, the 98 pairs were divided into two groups, according to free T4 percentiles of controls (10th to 49th and 50th to 90th), allowing subanalyses to be performed.
Results
Identification of study groups
Table 1 documents the process for identifying eligible hypothyroxinemic and euthyroxinemic women among the 5734 women who participated in the recruitment phase of the study. After excluding 174 women who were not euthyroid (3%) and 1410 women with pregnancy outcomes and free T4 levels that did not meet eligibility criteria (24%), 4150 euthyroid women remained. Among these women, 144 were identified with hypothyroxinemia, 99 of whom (68%) enrolled into the psychometric testing phase of the study. Reasons for nonenrollment included declining to participate or withdrawing from the study (n = 21), moving out of state (n = 13), inability to be contacted (n = 9), and the child not living with the mother (n = 2). Those not enrolled were significantly younger (P = 0.002), had fewer years of education (P = 0.003), were less likely to be white (88.9 vs. 98%, P = 0.03), and were more likely to be primiparous (51.1 vs. 33.3%, P = 0.03).
Table 1.
Identification of study subjects from the initial recruitment pool
| Number excluded | Number remaining | |
|---|---|---|
| Potential pool of study subjects | 5734 | |
| Exclusion criteria | ||
| Not euthyroida | 174 | 5560 |
| Free T4 out of rangeb | 890 | 4670 |
| Multiple gestation | 65 | 4605 |
| Birth less than 36 wk | 161 | 4444 |
| Birth weight less than 2500 gc | 103 | 4341 |
| Congenital abnormality | 34 | 4307 |
| Fetal/neonatal death | 0 | 4307 |
| Not delivered in Maine | 2 | 4305 |
| No matched birth record | 155 | 4150 |
| Hypothyroxinemia among 4150 eligible subjects | 144d | |
| Euthyroxinemia among 4150 eligible subjects | 4066 |
Euthyroid was defined as second-trimester maternal serum TSH 0.1–3.5 mIU/liter.
Second-trimester maternal free T4 level was outside the ranges defined for hypothyroxinemia (less than the third centile) or euthyroxinemia (10th to 90th centile).
One hundred seventy-five neonates with birth weight less than 2500 g are counted in the multiple gestation and birth less than 36 wk categories.
Ninety-nine of these women were enrolled (see Results).
Description of study groups
Table 2 compares selected variables between case and control women and their offspring, with the six matching variables examined first. Matching was highly successful, as judged by comparing the distribution of differences calculated between each case/control set. Case women were slightly but not significantly older (30.3 vs. 29.7 yr, P = 0.6). Given the small absolute difference, this is unlikely to have clinical significance.
Table 2.
Comparison of demographic and laboratory data among enrolled women, stratified by thyroid status
| n | Case mean | Control mean | Comparison (case vs. control) | P value (test) | |
|---|---|---|---|---|---|
| Matching variables | |||||
| Fetal sex male (%) | 99 | 46 | 46 | All pairs matched for fetal sex | 1.00 (McNemar's) |
| Primiparous (%) | 99 | 33 | 33 | All pairs match primiparous/multiparous | 1.00 (McNemar's) |
| Education (yr) | 99 | 14.1 | 14.1 | 81 pairs matched years | 1.00 (Wilcoxon) |
| Maternal age (yr) | 99 | 30.3 | 29.7 | Mean difference 0.60 (−0.02 to 1.21) | 0.06 (paired t) |
| Location in Mainea | 99 | 1.82 | 1.77 | 54 matched location | 0.62 (Wilcoxon) |
| Next sibling's age (yr) | 66 | 4.24 | 4.25 | Mean difference − 0.01 (− 0.95 to 0.94) | 0.98 (paired t) |
| Other demographic variables | |||||
| Child's age (yr) | 99 | 2.27 | 2.44 | Mean difference − 0.17 (− 0.22 to − 0.1) | <0.001 (paired t) |
| Maternal weight (kg) | 94/97 | 87.8 | 77.74 | Mean difference 11.0 (5.6 go 16.5) | <0.001 (paired t) |
| Gestational age (wk) | 99 | 17.3 | 16.8 | Mean difference 0.43 (0.13 to 0.73) | 0.006 (paired t) |
| BMI (kg/m2) | 89/95 | 32.9 | 28.5 | Mean difference 4.61 (2.59 to 6.64) | <0.001 (paired t) |
| Hollingshead score | 99 | 3.39 | 3.38 | 81 pairs within 1 (range 1 to 5) | 0.95 (Wilcoxon) |
| Home score | 99 | 35.8 | 36.1 | 45 pairs within 2 (range 25 to 41) | 0.36 (Wilcoxon) |
| Chemistry test results | |||||
| Free T4 (ng/dl) | 99 | 0.90 | 1.17 | Mean difference − 0.27 (− 0.28 to − 0.25) | <0.001 (paired t) |
| TSH (mIU/liter) | 99 | 1.39 | 1.29 | Mean log difference − 1.09 (− 1.26 to − 0.92) | <0.001 (paired t) |
| Total T4 (μg/dl) | 92 | 12.0 | 14.0 | Mean difference − 2.03 (− 2.88 to − 1.19) | <0.001 (paired t) |
| TPO antibody (percentile ≥35 IU/ml) | 92 | 14.1 | 10.1 | Mean difference 4.0 (−5.4 to 13.4) | 0.52 (McNemar's) |
The state was divided into southern, middle, and northern areas and coded 1, 2, and 3, respectively.
Among six demographic variables not used in matching, four differed significantly between cases and controls, as follows: children's ages at testing (0.17 yr earlier for cases), maternal weight and body mass index (BMI) (11.0 kg heavier and BMI 4.61 kg/m2 higher for case women), and gestational age of pregnancy at enrollment (0.43 wk later for case women). All enrolled women were recruited between 15 and 20 wk gestation, with 80 controls and 82 cases recruited between 16 and 19 wk. Neither of the tests measuring socioeconomic status (Hollingshead and HSQ scores) differed between the two groups.
Chemistry test results are compared in the two groups in the bottom of Table 2. Due to study selection criteria, free T4 and total T4 measurements were significantly lower among hypothyroxinemic women, as expected. TSH was slightly but significantly higher in the hypothyroxinemic subgroup. Percent of TPO antibody measurements 35 IU/liter or greater did not differ significantly between cases and controls. Among euthyroxinemic women in the lowest free T4 quintile, median BMI was 29.5 kg/m2, as opposed to 24.6 kg/m2 in the highest quintile (data not shown in Table 2).
Not shown are case-control comparisons for race and smoking. Consistent with Maine's racial distribution, 98% of cases and 99% of controls were white (P = 0.1). Smoking rates also did not differ (9 and 8%, respectively, P = 1.0). No enrolled child was identified with congenital hypothyroidism by newborn screening. Total T4, measured in newborn blood spots, was 17.3 μg/dl [95% confidence interval (CI) 16.3–18.2] and 18.1 μg/dl (95% CI: 17.2–18.9) in the hypothyroxinemic and euthyroxinemic groups, respectively; all subjects had total T4 greater than 5 μg/dl.
We found significant positive linear relationships between BSID III scores and three measures of socioeconomic status (Hollingshead score, HSQ score, mother's education) in both thyroid status subgroups (data not shown). These relationships are well known and reinforce the integrity of the data set. Negative linear associations were present between BMI and BSID III scores among control children (cognitive: r = −0.215, P = 0.036; language: r = −0.279, P = 0.006; motor: r = 0.184, P = 0.074). These relationships were not present among children of hypothyroxinemic mothers.
Maternal hypothyroxinemia and infant development
Table 3 shows BSID III composite scores (cognitive, language, and motor) in 98 cases and controls. In this analysis, one case-control pair was removed, as the control child was diagnosed with autism. Unadjusted scores were lower among case children in all three domains. The mean matched differences reached borderline significance for cognitive and motor but not for language scores. Similar results were seen for both male (n = 45 pairs) and female (n = 53 pairs) offspring (data not shown). When the 98 pairs were divided into two groups, according to free T4 percentiles of controls (10th to 49th and 50th to 90th), there was no consistent relationship between differences in unadjusted scores and control free T4 percentile range. The analysis was performed again, accounting for paired differences in gestational age, child's age at testing, maternal weight, and years of education. Maternal weight was missing for seven women (five cases and two controls). To allow these subjects to be included in the analysis, they were assigned the average weight in their respective group. After adjustment, the mean matched differences for the 98 pairs were reduced in all three domains; all became nonsignificant. Among the free T4 subgroups, differences between cases and controls lessened and remained nonsignificant.
Table 3.
BSID III composite test scores among offspring of women with isolated hypothyroxinemia or euthyroxinemia
| Domain | Mean BSID III composite score (95% CI) |
Mean matched difference (95% CI, P value) |
||
|---|---|---|---|---|
| Hypothyroxinemia | Euthyroxinemia | Unadjusted | Adjusteda | |
| FT4, 10th to 90th (n = 98)b | ||||
| Cognitive | 97.8 (95.6–99.9) | 100.6 (98.4–102.7) | − 2.8 (− 5.6 to − 0.0, 0.05) | −1.4 (−4.9 to 2.0, 0.43) |
| Language | 102.8 (99.3–106.3) | 105.4 (102.1–108.8) | −2.6 (−7.1 to 1.9, 0.25) | −0.6 (−6.2 to 5.0, 0.84) |
| Motor | 97.6 (95.3–99.9) | 101.0 (98.4–103.6) | −3.4 (−6.8 to 0.01, 0.05) | −1.9 (−6.1 to 2.4, 0.39) |
| FT4, 10th to 49th (n = 59)b | ||||
| Cognitive | 98.2 (95.6–100.8) | 100.4 (97.8–103.0) | −2.2 (−5.5 to 1.1, 0.18) | −1.0 (−5.0 to 3.1, 0.65) |
| Language | 103.5 (99.3–107.7) | 104.3 (99.9–108.6) | −0.8 (−6.3 to 4.8, 0.78) | 1.0 (−6.1 to 8.0, 0.79) |
| Motor | 97.2 (94.5–99.8) | 101.4 (98.0–104.8) | − 4.2 (− 8.4 to − 0.1, 0.047) | −2.1 (−7.3 to 3.1, 0.42) |
| FT4, 50th to 90th (n = 39)b | ||||
| Cognitive | 97.1 (93.1–101.0) | 100.8 (97.0–104.6) | −3.7 (−8.9 to 1.5, 0.16) | 0.2 (−6.3 to 6.8, 0.94) |
| Language | 101.7 (95.4–108.1) | 107.2 (101.8–112.6) | −5.5 (−13.3 to 2.3, 0.16) | −1.0 (−11.4 to 9.4, 0.85) |
| Motor | 98.4 (94.1–102.6) | 100.4 (96.4–104.5) | −2.1 (−8.0 to 3.8, 0.48) | −1.1 (−9.2 to 7.1, 0.80) |
After accounting for gestational age, child's age at testing, maternal weight, and years of education.
Matched pairs, defined by percentile of controls. One of 99 pairs removed due to autism in control (FT4 50th to 90th percentile).
One or more BSID III cognitive, language, and/or motor composite scores less than 85 were obtained in 21 of 98 offspring of hypothyroxinemic women and 13 of 98 offspring of euthyroxinemic women (P = 0.17). The same relationships were seen separately for cognitive (six vs. two), language (14 vs. 11), and motor scores (12 vs. five). Scores less than 70 were obtained in five offspring of hypothyroxinemic women and three offspring of euthyroxinemic women (P = 0.73).
Discussion
The present data demonstrate that isolated hypothyroxinemia during the second trimester is not associated with meaningful changes in composite BSID III scores at age 2 yr, when compared with scores for the offspring of matched euthyroxinemic women. Although euthyroid maternal hypothyroxinemia is hypothesized to be reflected in the fetal circulation at this time in gestation, it is not possible to obtain comparative measurements directly from the fetus. Child neurodevelopment is influenced by multiple factors, which need to be considered when evaluating the effect of an environmental exposure (16). Matching, although successful in minimizing differences for six potential confounders, does not eliminate potential confounding for others. Adjusting for these factors further reduced the small observed differences between the two subgroups.
In the current study, maternal weight (and BMI) both were higher among hypothyroxinemic women. Maternal weight was used to adjust the differences in scores because it had only seven missing results, compared with 14 missing results for BMI. Among euthyroxinemic women, the 4.9 kg/m2 difference between women in the lowest and highest free T4 quintiles is more pronounced than the free T4/BMI association reported in nonpregnant Australian adults, in whom BMI averaged 1.7 kg/m2 higher in the lowest, as opposed to highest, quintile of free T4 values (17). As reported by others, we found that markers of socioeconomic status, including maternal age and education, had a positive association with offspring developmental scores, whereas maternal weight (or BMI) had weaker negative associations (18–21).
The study reported by Pop et al. in 2003 (7) is most comparable with ours, in that it used a case-control design to examine neurocognitive development among offspring of euthyroid hypothyroxinemic women. Although neither study measured iodine concentrations in maternal urine, populations in both regions are generally considered iodine sufficient (7). Like Pop et al. (6), Henrichs et al. (9) reported a relationship between low first-trimester maternal levels of free T4 and cognitive delay among the offspring of hypothyroxinemic euthyroid Dutch women. However, the study's cross-sectional design and type of child evaluation (a self-administered structured parental report) complicate direct comparison with our data.
Using an earlier version of the Bayley Scale of Infant Development (second edition), Pop et al. (6) found that overall mental and motor scores among offspring of hypothyroxinemic mothers at age 1 and 2 yr were significantly lower than controls (by eight to 10 points). Although our results are in the same direction, our effect size is approximately 3- to 4-fold smaller. Differences in population make-up are unlikely to explain the different findings. The study population in the report by Pop et al. was described as Dutch Caucasian, and ours is 98% white. Design differences between the two studies, however, might account for at least a portion of the discrepancy. Pop et al. (7) obtained free T4 measurements at 12 wk gestation and followed up with measurements at 24 and 32 wk. Initial measurements made in the first trimester might be more strongly associated with offspring performance. It should be noted, however, that subsequent shifts in free T4 percentiles at 24 and 32 wk did not dilute the association with mental/motor development, as determined from the 12-wk measurements. In that study, matching was limited to parity and gravidity, and adjustments were not made for socioeconomic status (high incomes were reported for 49% of families in the euthyroxinemic group vs. 39% in the hypothyroxinemic group). Socioeconomic status is an important factor for child development (16, 19, 22), and this could result in a bias that favors higher performance in controls. Differences in the rates of TPO antibody positivity between the two studies might also influence results. Among euthyroxinemic women in our study, 10.1% were TPO antibody positive, as opposed to 14.1% of hypothyroxinemic women. Corresponding figures for The Netherlands study are 4.2 and 16.7% (8), suggesting further health differences between case and control women. In addition, The Netherlands study classified free T4 measurements below the 10th centile as hypothyroxinemia, whereas controls were selected from women with free T4 measurements between the 50th and 90th centiles. To explore whether this might explain differing findings between the two studies, we compared BSID III scores of our controls whose mothers' free T4 measurements were between the 50th and 90th centiles with those of their matched cases (Table 3). Our results were unchanged.
The concept of screening pregnant women for subclinical thyroid deficiency is controversial (23, 24), and discussion has focused primarily on identifying women with elevated TSH measurements with known adverse clinical consequences for the mother and fetus (25–29). Further data are needed, including evaluation of older children, for both subclinical hypothyroidism and isolated hypothyroxinemia before the screening debate can be resolved. Although addressing hypothyroxinemia, the results of this study cannot be extrapolated to children of mothers with subclinical hypothyroidism. Two randomized controlled trials are currently underway that address this subject: the Controlled Antenatal Thyroid Screening Study (ISTCTN46178175) and Thyroid Therapy for Mild Thyroid Deficiency in Pregnancy (NCT00388297). Our study shows that any impairment in neurocognitive development among the offspring of women with isolated hypothyroxinemia in the second trimester of pregnancy would be, at most, small. It also reinforces, however, the importance of accounting for other, more dominant, factors influencing child development in study design and analysis.
Acknowledgments
We thank June L. Gagnon, M.A., and Giovanna Hurley, M.Ed., for conducting the children's neurocognitive assessments, and Patricia Jewett for arranging follow-up evaluations.
This work was supported by Grant HD 44111 from the National Institute of Child Health and Human Development.
Disclosure Summary: The authors have nothing to disclose.
Footnotes
- BMI
- Body mass index
- BSID III
- Bayley Scales of Infant Development, third edition
- CI
- confidence interval
- FBR
- Foundation for Blood Research
- HSQ
- Home Screening Questionnaire
- T4
- thyroxine
- TPO
- thyroid peroxidase antibodies.
References
- 1. Morreale de Escobar G. 2001. The role of thyroid hormone in fetal neurodevelopment. J Pediatr Endocrinol Metab 14(Suppl 6):1453–1462 [PubMed] [Google Scholar]
- 2. Morreale de Escobar G, Obregón MJ, Escobar del Rey F. 2000. Is neuropsychological development related to maternal hypothyroidism or to maternal hypothyroxinemia? J Clin Endocrinol Metab 85:3975–3987 [DOI] [PubMed] [Google Scholar]
- 3. Morreale de Escobar G, Obregon MJ, Escobar del Rey F. 2004. Role of thyroid hormone during early brain development. Eur J Endocrinol 151(Suppl 3):U25–U37 [DOI] [PubMed] [Google Scholar]
- 4. Man EB, Serunian SA. 1976. Thyroid function in human pregnancy. IX. Development or retardation of 7-year-old progeny of hypothyroxinemic women. Am J Obstet Gynecol 125:949–957 [PubMed] [Google Scholar]
- 5. Smit BJ, Kok JH, Vulsma T, Briët JM, Boer K, Wiersinga WM. 2000. Neurologic development of the newborn and young child in relation to maternal thyroid function. Acta Paediatr 89:291–295 [PubMed] [Google Scholar]
- 6. Pop VJ, Kuijpens JL, van Baar AL, Verkerk G, van Son MM, de Vijlder JJ, Vulsma T, Wiersinga WM, Drexhage HA, Vader HL. 1999. Low maternal free thyroxine concentrations during early pregnancy are associated with impaired psychomotor development in infancy. Clin Endocrinol (Oxf) 50:149–155 [DOI] [PubMed] [Google Scholar]
- 7. Pop VJ, Brouwers EP, Vader HL, Vulsma T, van Baar AL, de Vijlder JJ. 2003. Maternal hypothyroxinaemia during early pregnancy and subsequent child development: a 3-year follow-up study. Clin Endocrinol (Oxf) 59:282–288 [DOI] [PubMed] [Google Scholar]
- 8. Kooistra L, Crawford S, van Baar AL, Brouwers EP, Pop VJ. 2006. Neonatal effects of maternal hypothyroxinemia during early pregnancy. Pediatrics 117:161–167 [DOI] [PubMed] [Google Scholar]
- 9. Henrichs J, Bongers-Schokking JJ, Schenk JJ, Ghassabian A, Schmidt HG, Visser TJ, Hooijkaas H, de Muinck Keizer-Schrama SM, Hofman A, Jaddoe VV, Visser W, Steegers EA, Verhulst FC, de Rijke YB, Tiemeier H. 2010. Maternal thyroid function during early pregnancy and cognitive functioning in early childhood: the generation R study. J Clin Endocrinol Metab 95:4227–4234 [DOI] [PubMed] [Google Scholar]
- 10. Vermiglio F, Lo Presti VP, Moleti M, Sidoti M, Tortorella G, Scaffidi G, Castagna MG, Mattina F, Violi MA, Crisà A, Artemisia A, Trimarchi F. 2004. Attention deficit and hyperactivity disorders in the offspring of mothers exposed to mild-moderate iodine deficiency: a possible novel iodine deficiency disorder in developed countries. J Clin Endocrinol Metab 89:6054–6060 [DOI] [PubMed] [Google Scholar]
- 11. Perrine CG, Herrick K, Serdula MK, Sullivan KM. 2010. Some subgroups of reproductive age women in the United States may be at risk for iodine deficiency. J Nutr 140:1489–1494 [DOI] [PubMed] [Google Scholar]
- 12. Haddow JE, McClain MR, Palomaki GE, Hollowell JG. 2007. Urine iodine measurements, creatinine adjustment, and thyroid deficiency in an adult United States population. J Clin Endocrinol Metab 92:1019–1022 [DOI] [PubMed] [Google Scholar]
- 13. Bayley N. 2006. Bayley Scales of Infant and Toddler Development. 3rd ed San Antonio, TX: Harcourt Assessment [Google Scholar]
- 14. Hollingshead A. 1975. Four Factor Index of Social Status. New Haven, CT: Yale University [Google Scholar]
- 15. Frankenburg WK, Coons CE. 1986. Home Screening Questionnaire: its validity in assessing home environment. J Pediatr 108:624–626 [DOI] [PubMed] [Google Scholar]
- 16. Mink PJ, Goodman M, Barraj LM, Imrey H, Kelsh MA, Yager J. 2004. Evaluation of uncontrolled confounding in studies of environmental exposures and neurobehavioral testing in children. Epidemiology 15:385–393 [DOI] [PubMed] [Google Scholar]
- 17. Makepeace AE, Bremner AP, O'Leary P, Leedman PJ, Feddema P, Michelangeli V, Walsh JP. 2008. Significant inverse relationship between serum free T4 concentration and body mass index in euthyroid subjects: differences between smokers and non-smokers. Clin Endocrinol (Oxf) 69:648–652 [DOI] [PubMed] [Google Scholar]
- 18. Record RG, McKeown T, Edwards JH. 1969. The relation of measured intelligence to birth order and maternal age. Ann Hum Genet 33:61–69 [DOI] [PubMed] [Google Scholar]
- 19. Palti H, Adler B. 1994. Do infant and early childhood tests predict intelligence quotients better than maternal characteristics? Public Health Rev 22:27–37 [PubMed] [Google Scholar]
- 20. Rodgers JL, Cleveland HH, van den Oord E, Rowe DC. 2000. Resolving the debate over birth order, family size, and intelligence. Am Psychol 55:599–612 [DOI] [PubMed] [Google Scholar]
- 21. Neggers YH, Goldenberg RL, Ramey SL, Cliver SP. 2003. Maternal prepregnancy body mass index and psychomotor development in children. Acta Obstet Gynecol Scand 82:235–240 [DOI] [PubMed] [Google Scholar]
- 22. Lawlor DA, Najman JM, Batty GD, O'Callaghan MJ, Williams GM, Bor W. 2006. Early life predictors of childhood intelligence: findings from the Mater-University study of pregnancy and its outcomes. Paediatr Perinat Epidemiol 20:148–162 [DOI] [PubMed] [Google Scholar]
- 23. Gharib H, Tuttle RM, Baskin HJ, Fish LH, Singer PA, McDermott MT. 2005. Subclinical thyroid dysfunction: a joint statement on management from the American Association of Clinical Endocrinologists, the American Thyroid Association, and The Endocrine Society. J Clin Endocrinol Metab 90:581–585 [DOI] [PubMed] [Google Scholar]
- 24. Surks MI. 2005. Commentary: subclinical thyroid dysfunction: a joint statement on management from the American Association of Clinical Endocrinologists, the American Thyroid Association, and The Endocrine Society. J Clin Endocrinol Metab 90:586–587 [DOI] [PubMed] [Google Scholar]
- 25. Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, Gagnon J, O'Heir CE, Mitchell ML, Hermos RJ, Waisbren SE, Faix JD, Klein RZ. 1999. Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. N Engl J Med 341:549–555 [DOI] [PubMed] [Google Scholar]
- 26. Glinoer D, Soto MF, Bourdoux P, Lejeune B, Delange F, Lemone M, Kinthaert J, Robijn C, Grun JP, de Nayer P. 1991. Pregnancy in patients with mild thyroid abnormalities: maternal and neonatal repercussions. J Clin Endocrinol Metab 73:421–427 [DOI] [PubMed] [Google Scholar]
- 27. Leung AS, Millar LK, Koonings PP, Montoro M, Mestman JH. 1993. Perinatal outcome in hypothyroid pregnancies. Obstet Gynecol 81:349–353 [PubMed] [Google Scholar]
- 28. Allan WC, Haddow JE, Palomaki GE, Williams JR, Mitchell ML, Hermos RJ, Faix JD, Klein RZ. 2000. Maternal thyroid deficiency and pregnancy complications: implications for population screening. J Med Screen 7:127–130 [DOI] [PubMed] [Google Scholar]
- 29. Casey BM, Dashe JS, Wells CE, McIntire DD, Byrd W, Leveno KJ, Cunningham FG. 2005. Subclinical hypothyroidism and pregnancy outcomes. Obstet Gynecol 105:239–245 [DOI] [PubMed] [Google Scholar]