Levothyroxine overtreatment during pregnancy increases odds for preterm delivery
Click Here to Manage Email Alerts
A cohort of pregnant women in Canada who were overtreated for hypothyroidism with levothyroxine were more likely to have a preterm delivery than women with normal thyroid-stimulating hormone, according to a study published in Thyroid.
“Overtreatment with levothyroxine in pregnancy was associated with preterm delivery among pregnant women using thyroid replacement therapy prior to conception,” Patricia Lemieux, MD, department of medicine at the University of Calgary Cumming School of Medicine in Alberta, Canada, told Healio. “All pregnant women on thyroid replacement prior to pregnancy should have their TSH checked in pregnancy. This is not happening, but it should be.”
Lemieux and colleagues conducted a retrospective cohort study of women aged 15 to 49 years who delivered a baby in Alberta, Canada, from October 2014 to September 2017. Women who had a prescription filled for any thyroid replacement medication in the 2 years before conception were included.
Researchers collected delivery data from the Alberta Perinatal Health Program database. Data were linked to the Discharge Abstract Database for hospitalization data and province-wide databases and the Pharmaceutical Information Network database for information on prescription medications and outpatient physician visits. Frequency of TSH testing was broken down by trimester.
A normal TSH level was defined as 0.1 mIU/L to 4 mIU/L based on American Thyroid Association guidelines. Women were defined as overtreated with levothyroxine if they had one TSH measurement of less than 0.1 mIU/L during pregnancy, and the undertreated group included women with at least one TSH measurement of 10 mIU/L or higher.
There were 10,680 deliveries with women on thyroid hormone prior to conception during the study period. Of the cohort, 82.2% had at least one TSH measurement during pregnancy, and 62.8% had two or more tests performed.
Of 9,869 women included in analysis of levothyroxine dosage, 43.7% had at least one adjustment during pregnancy, with the most common timing for first adjustment during weeks 5 and 6 of gestation. Levothyroxine dosage increased as pregnancy progressed. Compared with preconception levothyroxine dosage, the median increases were 17.9% at first trimester, 35.7% at second trimester and 43.6% at third trimester.
Of the 8,774 women who had TSH testing during pregnancy, 4% were in the levothyroxine overtreatment group and 9.1% were in the undertreatment group. After adjusting for confounders, the overtreatment group was more likely to have preterm delivery compared with those with a normal TSH level (adjusted OR = 2.14; 95% CI, 1.51-2.78). There were no associations between neonatal ICU admission and overtreatment or undertreatment with levothyroxine in adjusted analysis. However, when researchers restricted analysis to women with two or more prescriptions filled before pregnancy, there was an association between neonatal ICU admission and overtreatment (aOR = 1.49; 95% CI, 1.03-2.16).
Women were more likely to fall in the overtreatment group if their prepregnancy TSH was less than 1.5 mIU/L or the levothyroxine dosage was at least 100 µg per day in the year before pregnancy (P < .001). Women diagnosed with thyroid cancer or treated for Graves’ disease were more likely to have TSH suppression, whereas those diagnosed with primary hypothyroidism were less likely to fall in the overtreatment cohort.
“Clinicians should consider exercising caution to avoid overtreatment with levothyroxine in pregnancy in women on thyroid replacement prior to conception,” Lemieux said. “The usual management of preexisting hypothyroidism is to increase levothyroxine dosage by 25% to 30% in all pregnant women, as recommended by the American Thyroid Association. However, a subgroup of pregnant women with preconception TSH less than 1.5 mIU/L or levothyroxine dose greater than 100 µg per day may benefit from a more conservative approach to levothyroxine increments during pregnancy.”
For more information:
Patricia Lemieux, MD, can be reached at Patricia.Lemieux@albertahealthservices.ca.
Perspective
Back to Top
Terry F. Davies, MBBS, MD, FRCP, FACE
Thyroid replacement in pregnancy is a straightforward process dependent on the willingness of the patient to have frequent TSH evaluations throughout her three trimesters. Nevertheless, the common habit of adjusting doses before a change in TSH (as recommended by some professional associations) can indeed lead to significant overdosing. In this exhaustive, retrospective, Canadian study of over 10,000 pregnancies by Dr. Lemieux and colleagues, they find that almost 18% of women taking thyroid replacement before pregnancy were not even tested! Something is seriously amiss with the health care these patients received. It is possible that such a prescription-based study included women who did not need replacement therapy, but this is unlikely to explain such a large deficit in testing.
The question then arises as to whether this failure placed these mothers and babies at any risk. The risks of clinical hypothyroidism in pregnancy are well documented, ranging from increased pregnancy complications, including pregnancy loss, to babies who develop more ADHD or a lower IQ than expected. However, the data on consequences of mild increases in TSH remain controversial because of the difficulty in studying and treating early enough in pregnancy when brain development is at its peak and before most women present for assessment.
Information on the consequences of thyroid overactivity is less impressive and in fact mild Graves’ hyperthyroidism is often left untreated which allows the immune system to deal with it as pregnancy progresses. So another important finding from the study by Lemieux and colleagues is their demonstration of an increase in premature delivery among those women overtreated with thyroxine in pregnancy, raising the issue of a sin of commission (OR = 2.14). That is the problem with treating very mild hypothyroidism. Studies have shown that a significant proportion of non-pregnant patients end up with a suppressed TSH and is a major argument against treating mild thyroid failure — somewhat depending on the compliance of your patient population. Nevertheless, 10% or more of healthy women show a low TSH in early pregnancy secondary to the highly variable influence of human chorionic gonadotropin on the thyroid — so called gestational hyperthyroidism — and there are no data, yet, to suggest this phenomenon has adverse effects. This certainly needs to be reevaluated in light of this study.
Although thyroid antibodies have been consistently associated with an increased miscarriage rate, the lack of data on pregnancy loss in this population study and the absence of thyroid antibody status in the women will require a different approach to their evaluation since these parameters were not available in the data bases accessed.
Collapse
Read more about
Click Here to Manage Email Alerts