University of Southern California Dissertations and Theses

Study of maternal free thyroxine and thyroid-stimulating hormone's relationship with infant birthweight

(USC Thesis Other)

Study of maternal free thyroxine and thyroid-stimulating hormone's relationship with infant birthweight

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content STUDY OF MATERNAL FREE THYROXINE AND THYROID-STIMULATING HORMONE’S RELATIONSHIP WITH INFANT BIRTHWEIGHT By Jiazheng Chen A Thesis Presented to the FACULTY OF THE USC KECK SCHOOL OF MEDICINE UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF SCIENCE (BIOSTATISTICS) August 2022 Copyright 2022 Jiazheng Chen i Table of Contents List of Tables...................................................................................................................................ii Abstract...........................................................................................................................................iv Chapter 1. Introduction.....................................................................................................................1 Chapter 2. Methods 2.1. Participants……………………………………………………………………..................…..2 2.2. Measurements...........................................................................................................................3 2.3. Statistical Analysis....................................................................................................................4 Chapter 3. Results 3.1. Variables Examination..............................................................................................................6 3.2. Univariate Analysis.................................................................................................................13 3.3. Multivariable Analysis............................................................................................................14 3.3.1. Main Effects.........................................................................................................................14 3.3.2. ExploratoryAnalysis............................................................................................................16 Chapter 4. Discussion 4.1. Selection of Variables for Analysis........................................................................................17 4.2. Comparison with Other Studies..............................................................................................18 4.3. Strength, Limitations and Questions.......................................................................................19 4.4. Conclusions.............................................................................................................................20 References......................................................................................................................................22 ii List of Tables Table 1. Description of Independent Variables...............................................................................7 Table 2. Univariate Analysis Results Table...................................................................................14 Table 3. Multivariable Analysis Results Table..............................................................................15 Table 4. Interactions between TSH and Maternal Race................................................................17 iii List of Figures Figure 1. Comparison of distribution of TSH in the 1st and the 3rd trimester................................8 Figure 2. Comparison of distributions of FT4 in the 1st and the 3rd trimester...............................9 Figure 3. Comparison of distributions of birthweight in the 1 st and the 3 rd trimester; identified outliers are circled red....................................................................................................................10 Figure 4. Scatter Plots and Regression Lines Comparing the Relationship between FT4 and TSH with and without Outliers in the 1st Trimester; excluded observations are circled red.................11 Figure 5. Scatter Plots and Regression Lines Comparing the Relationship between FT4 and TSH with and without Outliers in the 3rd Trimester; excluded observations are circled red................12 iv Abstract Context Free thyroxine (FT4) and thyroid-stimulating hormone (TSH) are two essential thyroid hormones that affect the development of the human body. The fetus obtains thyroid hormones from mothers, and the fetus's demand for maternal thyroid hormones is different in the early and late pregnancy stages. Objective This thesis investigates the association of FT4 and TSH with birthweight and to quantitatively compare the associations of thyroid hormones in first and third trimesters with birthweight. Participants This study recruited mothers from either LAC+USC Medical Center, Eisner Health (main site), USC Obstetrics & Gynecology, South Central Clinic and the community in Los Angeles area. The data were collected from 2015 to September 2021 by obtaining medical records and questionnaires. Main Goals The primary goal of this thesis is to investigate the association between maternal FT4 and TSH and birthweight in the first and the third trimester, and to ascertain the difference in associations between FT4 and TSH of the 1 st and 3 rd trimester and birthweight. Results v In the univariable analysis, birthweight was associated with the third trimester FT4 (p = 0.0024). In the multivariable analysis, birthweight was associated with the third trimester TSH (p = 0.0140) and FT4 (p = 0.0012). Significant interactions were found between TSH in both trimesters and birthweight. Conclusions As maternal FT4 in the third trimester is associated with baby’s birthweight, and FT4 and TSH affect each other biologically, the association implies that thyroid hormones are associated with baby’s development. The associations between thyroid hormones and birthweight are different in the first and the third trimester. 1 1. Introduction Thyroid hormones, including free thyroxine (FT4) and thyroid stimulating hormone (TSH), are essential hormones for the function of the human body. They affect both development and the metabolism. FT4 is a type of thyroid hormone that is physically detached from proteins and enters human tissues when tissues need the hormone to function, such as to facilitate digestion and brain development. Thyroid glands control the secreting activity of FT4 by using a feedback loop involving other glands, including hypothalamus and pituitary. As such, thyroxine affects the whole body and all systems to some extent. Some new animal experiments and studies suggest that thyroxine may protect against myocardial cell apoptosis, Doxorubicin-induced cardiac injury and cardiac dysfunction in mice. Results also indicate that excessive maternal FT4 is able to modify the conditions of rats’ proliferative activity and angiogenic profile during the period between birth and weaning. Thyroid-stimulating hormones, also known as TSH or thyrotropin, is a hormone that physiologically interacts with FT4. TSH stimulates the production of FT4, and when FT4 level reaches a certain level, the negative feedback loop starts to work, slowing the production of TSH. Therefore, it is expected that TSH and FT4 are negatively correlated. Moreover, TSH, like FT4, is one of the essential hormones that impact cardiovascular conditions. As TSH is closely related to metabolism, Hyperthyroidism, the disease caused by high levels of the thyroid hormone, is able to accelerate the metabolism and cause faster energy consumption, showing as symptoms such as high heart rate, shaky hands and weight loss. Hypothyroidism (low levels of the thyroid hormone) has the opposite effects on the human body, with symptoms such as low heart rate, fatigue and weight gain. 2 Early-stage fetuses depend heavily on maternal thyroid hormones, as their organs are not developed enough to produce their own thyroid hormones. Studies show that abnormal concentration of thyroid hormones is associated with miscarriage and fetal or neonatal death, but few studies quantitatively associate mothers’ thyroid hormone levels with babies’ birth outcome, including fetuses’ growth and development if birth is successful. Among the few studies that have been conducted, the quantitative relationships are not clearly identified. There are studies indicating that the developing fetus depends heavily on maternal thyroid hormones, and that thyroid hormones are vital to differentiation and maturation of fetal organs. These studies show that abnormal thyroid hormones result in immaturity of hypothalamic-pituitary- adrenal axis. Towards the end of pregnancy, the demand of thyroid hormones from the fetus decreases, and the different effects of thyroid hormones in different fetus development stages is a relatively under-researched area. In this study, infants’ birthweight is considered the main indicator of fetus growth and development. The objective is to investigate the association of FT4 and TSH with birthweight. 2. Methods 2.1. Participants This study uses the data collected from Maternal and Developmental Risks from Environmental and Social Stressors (MADRES) Pregnancy Cohort12. The process of data collection (including mothers’ and newborns’) and recruitment of mothers started from late 2015 and is ongoing. This study uses data through September 2021. Mothers enrolled in the study cohort prior to 30 weeks of gestation. They were given informed consent and gave HIPAA authorization. All mothers were recruited from LAC+USC Medical Center, Eisner Health (main 3 site), USC Obstetrics & Gynecology, South Central Clinic and the community. “The mothers recruited from the community” means those who participated via self-referral. To be included in the study, mothers must be (1) at or below the 30-week gestation at the time of recruitment, (2) over 18 years old, and (3) able to speak English or Spanish fluently. Mothers who meet the following criteria were excluded: (1) HIV-positive; (2) disabled (including physical, mental or cognitive that keep mothers from normally participating the study or providing informed eligible consent); (3) currently in custody; or (4) have multiple gestation. 2.2. Measurements Study visits occurred at three times during pregnancy, roughly corresponding to the first, second and third trimesters (defined as pregnancy time <20 weeks, 18-27 weeks and 30-34 weeks, respectively). All variables about the mother’s demographic and lifestyle information, including education level and pre-pregnant BMI, were obtained via questionnaires during the first, second and third trimesters. Medical history was collected during the third trimester. Mothers’ blood samples were collected during the first and third trimester when they visited the testing sites for FT4 and TSH testing. FT4 was measured in plasma using the Roche Cobas e411 (a fully automated immunoassay analyzer platform performing moderate-complexity thyroid testing using a 2-point calibration for analytes). The Elecsys Free T4 test was used to test plasma FT4 concentrations and has been calibrated against equilibrium dialysis. The accuracy was checked by calibration verification and by the CAP external proficiency testing relative to users of this method. Precision was assessed on an ongoing basis from human serum pool (HSP) and Bio-Rad Lyphocheck Immonoassay Plus Tri-level Controls in the low, medium, and high measurements of the reportable range. TSH was measured in plasma using the Roche Cobas 4 e411. Elecsys TSH assay employs monoclonal antibodies specifically directed against human TSH. The Elecsys TSH method is standardized against the International Reference Preparation MRC 68/558. The accuracy is checked by calibration verification and by the CAP external proficiency testing relative to users of this method. Bio-Rad QCs are also included with each run for peer-group determinations of accuracy. Birthweights and gender of babies were obtained from either medical records, medical centers, or questionnaires from mothers at birth or during the 7 to 14 days after birth. Gestational age was solely obtained from medical records at birth. 2.3. Statistical analysis Statistical analysis, including univariable and multivariable analysis, was performed using SAS 9.4. The inclusion criteria of selecting covariates for multivariable analysis were either of the following: (1) factors that reveal baby’s or mother’s biological characteristics; (2) factors that did not contain more than 20% missing values in the final dataset; and (3) factors that may affect baby’s birthweight. Therefore, family thyroid disease history was excluded from the study because of high rate of missing values in the first trimester (22.07% missing). The statistical analysis used four datasets: (1) the birth outcome dataset was used as it contains the baby's gender, a commonly used factor that needs to be controlled, and birthweight in grams. (2) the thyroid dataset was used for its TSH level (in mIU/mL), FT4 level (in ng/dL) and trimester of collection. (3) The pregnancy outcome dataset was used for its mother’s body mass index (BMI) during pregnancy and gestational age (in weeks). (4) The maternal demographics dataset was used for extracting mother’s education level. The following variables were selected as a priori: FT4, TSH, gender of baby, the trimester of data collection, mother’s BMI during 5 pregnancy, gestational age and the mother’s education level. Therefore, the following variables were selected into the multivariable model: maternal pre-pregnant BMI, highest education level, baby’s gender and gestational age at birth. The pairwise interaction terms among all covariates were examined. The categorical variables were treated by creating dummy variables. Mother’s education level was categorized into low-education level (below and equal to some college or technical school) and high education level (completed 4-year college or higher). The low-education level was set as the reference group when performing multiple linear regression. "Female" was the reference group for the baby’s gender compared to "male." To investigate potentially different associations of TSH and FT4 in different trimesters, two variables containing only first trimester’s or third trimester’s thyroid hormone data were created. All categorical variables’ distributions were examined. The variable of mothers’ smoking status was also excluded from the model because the smoking population was too small (n = 6 in the 1st trimester; n = 5 in the 3rd trimester). As linear regression is used in the analysis, distribution of each dependent and independent variable was separately examined to see if it satisfies linear regression assumptions and if it is appropriate to be included in the regression model. Associations of birthweight with each trimester’s FT4 and TSH were separately examined. Simple linear regression was used when conducting univariate analysis. In order to take the covariates into consideration and to check if there were confounding variables, multiple linear regression was used when conducting multivariable analysis. Potential outliers or influential points of birthweight, TSH and FT4 in the 1st trimester (n = 9) and the 3rd trimester (n = 2) were identified based on scatterplots. The identified outliers 6 were further examined to ensure if they caused significant difference in the analyses and if they were due to data entry errors. There was no significant difference between the analyses with and without outliers, and there was no confirmed data entry error. As such, no outliers were excluded from the study. Confounders were defined as the variables which were able to influence the parameter estimations of FT4 or TSH to have more than 20% change, and there was no covariate caused more than 20% change of the parameter of hormones in the multivariable models. Interactions were explored by examining the coefficients of each variable times another variable. If a pair of covariates which were known to have collinearities or have variance inflation factor (VIF) higher than 10, then one of the covariates were chosen to be excluded. For all 2- sided P values, 0.05 was chosen as the cut point for significance level. 3. Results 3.1. Variable examination In the analysis of the first trimester data, there were 213 records of FT4 (mean = 1.08 ng/dL) and TSH (mean = 1.33 mIU/mL). 171 records of FT4 (mean = 0.91 ng/dL) and TSH (mean = 1.85 mIU/mL) were included for the analysis of the third trimester data (Table 1). Potential outliers or influential points of TSH and FT4 were found in both the first and the third trimester (Figure 1 and Figure 2). TSH and FT4 were negatively correlated in both trimesters with or without outliers; the difference in the linear relationships between FT4 and TSH in both trimesters were further examined (Figure 3 and Figure 4). To investigate the effect of outliers, outliers caused by TSH and FT4 were identified using scatterplots (Figure 1) and the outlier caused by birthweight was identified by using boxplot. Nine observations were excluded from the 1st trimester dataset, and one of the outlier 7 was excluded because of the corresponding birthweight was out of range. Two observations were excluded from the 3rd trimester dataset and both were excluded because the birthweight was out of range. The exclusion criteria for the outliers were: (1) 1st trimester TSH < 4 mIU/mL; (2) 1st trimester FT4 < 1.8 ng/dL; (3) 3rd trimester TSH > 0.1 mIU/mL; (4) birthweight > 2000 grams. Separated univariate and multivariable analyses were performed using the datasets with and without outliers and influential points. It was shown that the coefficients of thyroid hormones in all the analyses without outliers were not significantly different from the analyses with outliers. All previously identified potential outliers or influential points were biologically plausible to be included in the study, and there were no data entry error identified. Therefore, all data points were included in the analysis. Table 1. Description of Independent Variables 1st Trimester 3rd Trimester (n = 213) (n = 171) TSH, mean (SD) (mIU/mL ) 1.33 (0.98) 1.85 (0.93) FT4, mean (SD) (ng/dL) 1.08 (0.19) 0.91 (0.13) Maternal Pre-pregnant BMI, mean (SD) (kg/m 2 ) 28.97 (6.38) 29.00 (6.30) Baby's Gestational Age (GA) at Birth, mean (SD) (Weeks) 38.89 (1.73) 39.09 (1.39) Male (%) 107 (50.23%) 90 (52.63%) Female (%) 106 (49.77%) 81 (47.37%) High (%) 35 (16.43%) 28 (16.37%) Lower (%) 178 (83.57%) 143 (83.63%) Hispanic (%) 168 (78.50%) 134 (79.76%) Non-Hispanic (%) 45 (21.50%) 37 (20.24%) Characteristics Baby's Gender Maternal Education Level Maternal Race 8 Figure 1. Comparison of distribution of TSH in the 1st and the 3rd trimester 9 Figure 2. Comparison of distributions of FT4 in the 1st and the 3rd trimester 10 Figure 3. Comparison of distributions of birthweight in the 1 st and the 3 rd trimester; identified outliers are circled red 11 Figure 4. Scatter Plots and Regression Lines Comparing the Relationship between FT4 and TSH with and without Outliers in the 1st Trimester; excluded observations are circled red 12 Figure 5. Scatter Plots and Regression Lines Comparing the Relationship between FT4 and TSH with and without Outliers in the 3rd Trimester; excluded observations are circled red In the multivariable analysis, SAS was used to perform linear regression on the variables (Table 1). There were no variables with VIF higher than 10, so no covariates were excluded 13 because of high collinearity. The pre-pregnant BMI and baby's gestational age at birth were fairly normally distributed. 3.2. Univariable analysis In the analysis of the first trimester data, 213 records were used for the analysis. The result shows that 1 mIU/mL increases in maternal TSH was associated with 49.05 gram decreases in birthweight (95% CI = -123.93, 25.84; P value = 0.1981); 1 ng/dL increases in maternal FT4was associated with 82.28 gram decreases in birthweight (95% CI = -471.52, 306.96; P value = 0.6773). TSH and FT4 parameters were not significantly different from 0, indicating there was no statistically significant association of birthweight with TSH or FT4. In the analysis of the third trimester (n = 171), the linear relationship shows that 1 mIU/mL increases in maternal TSH was associated with 26.45 gram decreases in birthweight (95% CI = -101.52, 48.61; P value = 0.4876); 1 ng/dL increases in the third trimester maternal FT4 was associated with 846.33 gram decreases in birthweight (95% CI = -1389.36, -303.30; P value = 0.0024). This significant result indicates that only the third trimester FT4 was significantly associated with birthweight (Table 2). Based on the linear regression parameter estimations, the association between TSH and birthweight in the first and the third trimester were not significantly different, while the association of FT4 were significantly different between the two trimesters. 14 Table 2. Univariate Analysis Results Table Trimester Variable Parameter (β) Standard Error 95% Confidence Interval P value 1st (n = 213) TSH -49.05 37.99 -123.93 25.84 0.1981 FT4 -82.28 197.46 -471.52 306.96 0.6773 3rd (n = 171) TSH -26.45 38.03 -101.52 48.61 0.4876 FT4 -846.33 275.08 -1389.36 -303.30 0.0024 3.3. Multivariable analysis 3.3.1. Primary Analysis There are 213 participants included in the first trimester dataset and 171 participants included in the third trimester dataset. For the analysis of TSH, in the first trimester analysis, 1 mIU/mL higher in maternal TSH was associated with 7.69 gram higher birthweight of the baby (95% CI = -58.79, 74.18; P value = 0.8198). The third trimester analysis showed that, 1 mIU/mL higher in maternal TSH was associated with 158.35 gram lower birthweight of the baby (95% CI = -284.19, -32.52; P value = 0.0140). For the analysis of FT4, in the first trimester analysis, 1 ng/dL higher in maternal FT4 was associated with 186.62 lower in birthweight (95% CI = -509.70, 136.47; P value = 0.2561); in the third trimester analysis, 1 ng/dL higher in maternal FT4 was associated with 833.37 lower in birthweight (95% CI = -1332.08, -334.65; P value = 0.0012). The results showed that the association between FT4/TSH in the third trimester and birthweight were significant. The parameters of FT4 in the first and the third trimester are significantly different. 15 Table 3. Multivariable Analysis Results Table Trimester Variable Parameter (β) Standard Error 95% Confidence Interval P value 1st (n = 213) TSH (mIU/mL) 7.69 33.72 -58.79 74.18 0.8198 Maternal Race (Hispanic) 0 . . Maternal Race (Non- Hispanic) 56.57 131.95 -203.59 316.74 0.6686 Baby's Gender (Female) 0 . . . . Baby's Gender (Male) 34.56 60.67 -85.05 154.17 0.5695 Maternal Pre-pregnant BMI (kg/m²) 13.83 4.87 4.22 23.44 0.005 Baby's Gestational Age at Birth (Weeks) 181.91 17.82 146.78 217.03 <.0001 Maternal Education (Low) 0 . . . Maternal Education (High) 46.68 89.79 -130.34 223.7 0.6037 1st (n = 213) FT4 (ng/dL) -186.62 163.88 -509.7 136.47 0.2561 Maternal Race (Hispanic) 0 . . . . Maternal Race (Non- Hispanic) 192.7 80.05 34.88 350.52 0.017 Baby's Gender (Female) 0 . . Baby's Gender (Male) 17.55 60.81 -102.34 137.43 0.7732 Maternal Pre-pregnant BMI (kg/m²) 12.93 4.95 3.17 22.68 0.0097 Baby's Gestational Age at Birth (Weeks) 181.64 17.89 146.37 216.9 <.0001 Maternal Education (Low) 0 . . . Maternal Education (High) -0.66 87.94 -174.05 172.72 0.994 3rd (n = 171) TSH (mIU/mL) -158.35 63.73 -284.19 -32.52 0.014 Maternal Race (Hispanic) 0 . . . . Maternal Race (Non- Hispanic) 57.17 155.42 -364.07 249.72 0.7135 Baby's Gender (Female) 0 . . Baby's Gender (Male) 20.08 64.12 -106.54 146.69 0.7546 Maternal Pre-pregnant BMI (kg/m²) 12.36 5.17 2.15 22.58 0.018 Baby's Gestational Age at Birth (Weeks) 145.4 23.55 98.89 191.91 <.0001 Maternal Education (Low) 0 . . . Maternal Education (High) 43.45 89.05 -132.38 219.29 0.6262 3rd (n = 171) FT4 (ng/dL) -833.37 252.57 -1332.08 -334.65 0.0012 Maternal Race (Hispanic) 0 . . . . 16 Maternal Race (Non- Hispanic) 178.18 81.26 17.74 338.62 0.0297 Baby's Gender (Female) 0 . . . . Baby's Gender (Male) 16.36 62.67 -107.38 140.1 0.7944 Maternal Pre-pregnant BMI (kg/m²) 11.52 5.09 1.47 21.57 0.0249 Baby's Gestational Age at Birth (Weeks) 144.07 23.07 98.53 189.62 <.0001 Maternal Education (Low) 0 . . . . Maternal Education (High) 38.77 87.19 -133.39 210.94 0.6571 3.3.2. Exploratory Analysis During the first trimester, the analysis of TSH showed that the maternal pre-pregnant BMI (estimated β = 15.25, P value = 0.0023) and baby’s gestational age at birth (estimated β = 173.00, P value < 0.0001) were significantly associated with birthweight. In the analysis of FT4, maternal pre-pregnant BMI (estimated β = 14.53, P value = 0.0038) and baby’s gestational age at birth (estimated β = 175.68, P value < 0.0001) showed significant association with birthweight (Table 4). In the analysis of third trimester data, maternal pre-pregnant BMI (estimated β = 14.26, P value = 0.0075) and baby’s gestational age at birth (estimated β = 135.30, P value < 0.0001) were significantly associated with birthweight in the analysis of association between TSH and birthweight. The analysis of FT4 showed that maternal pre-pregnant BMI (estimated β = 13.00, P value = 0.0118) and baby’s gestational age at birth (estimated β = 138.55, P value < 0.0001) were associated with birthweight (Table 4). Interactions were explored pairwise; it showed that only TSH, in both the first and the third trimester, had significant interactions with maternal race (Table 5). Therefore, the estimated birthweight was differed by maternal race. In the first trimester analysis, the Hispanic mother’s baby was estimated to be 7.69 grams heavier if the baby had 1 mIU/mL higher in maternal TSH, 17 but was estimated to be 151.52 lighter if the mother was non-Hispanic. Therefore, the birthweight of the Hispanic mother’s baby was estimated to be 159.21 grams heavier than that of non-Hispanic mother’s baby. In the third trimester analysis, the Hispanic mother’s baby was estimated to be 158.35 grams lighter if the baby had 1 mIU/mL higher in maternal TSH, but was estimated to be 57.80 grams heavier if the mother was non-Hispanic, so the birthweight of the Hispanic mother’s baby was estimated to be 216.15 grams lower than that of non-Hispanic mother’s baby. Table 4. Interactions between TSH and Maternal Race Trimester Race Parameter of the Interaction term Parameter of Race Effect of TSH Difference in Birthweight (Hispanic - Non-Hispanic) 1st (n = 213) Hispanic -215.78 0 7.69 159.21 Non- Hispanic 56.57 -151.52 3rd (n = 171) Hispanic 158.98 0 -158.35 -216.15 Non- Hispanic 57.17 57.80 4. Discussion 4.1. Selection of Variables for Analysis There were three variables that could serve as proxies for fetus development during maternal pregnancy: baby’s birthweight, length and head circumference. We did not choose length as an indicator of development because babies do not usually lie straight during the measurement so the data is possibly inaccurate. The head circumference was excluded as well because not all of the babies’ heads were measured in the same spot, leading to inconsistencies so that the accuracy was weak. Baby’s birthweight is a common indicator of fetus development and is more accurate as it has less random errors. 18 There were three biological factors included in the study as covariates, except birthweight, TSH and FT4 concentrations: baby’s gender, maternal pre-pregnant BMI and baby’s gestational age at birth. Babies with different genders possibly have different developmental paces, so baby’s gender was considered as well. Maternal pre-pregnant BMI is possibly affected by both genetic and environmental factors: studies show genes account for 25% of the predisposition to be obese while being underweight can be caused by duplicated chromosome locus or illness, including eating disorders. Baby’s gestational age was taken into consideration because proper gestational age is essential for fetus development. Shorter or longer than normal gestational age can affect both baby’s development and birthweight. Demographic factors were also selected. Maternal education level was selected as it is known to have association with income and body weight. Mothers with higher education levels tend to have more income and thus are more likely to have access to nutritious foods and an active lifestyle. Although we had mother’s income recorded in the dataset, we excluded it as a variable in the analysis because (1) income highly depends on occupations and industries, (2) income and education level may cause higher collinearity as they are known to be statistically associated, and (3) income in our dataset had more than 20% missing values so that income fell into the exclusion criteria. 4.2. Comparison with Other Studies This study, in both the first and the third trimester, the parameter estimations of TSH and FT4 were negative. It is supported by other studies which show negative association between thyroid hormones and birthweight. The results, however, are not completely consistent with other studies. First, other studies concludes that TSH and FT4 are significantly negatively 19 associated with birthweight, while this study did not suggest significant associations; second, there are studies concluding that the thyroid hormones show FT4 and TSH in the first trimester have significantly lower parameter estimation of association with birthweight than in the third trimester; On the contrary, the multivariable analysis in this study shows that thyroid hormones in the third trimester have greater impacts to birthweight. The possible cause of controversial conclusions are: (1) this study involves relatively smaller sample size than other studies because missing values lowered the sample size we could use, and (2) the covariates included in this study and other studies are different; for example, there is one study only adjusted baby’s sex, and there might be more confounders unnoticed. Overall, this study and other studies support the idea that maternal thyroid hormones matters in fetus’s development. The difference between early and late thyroid hormones’ association with birthweight might worth taking more statistical and biological studies. 4.3. Strength, Limitations and Questions This study showed that the first or the third trimester maternal TSH and FT4 were not significantly associated with baby’s birthweight in all analyses, except in the univariate analysis of the third trimester of FT4. Maternal pre-pregnant BMI was significantly associated with birthweight in the analysis of first trimester TSH, first trimester FT4 and third trimester FT4; the association of maternal pre-pregnant BMI was “marginally” associated with birthweight when analyzing the first trimester FT4 (P value = 0.0717). These results supported the hypothesized reasons when we chose covariates, and may motivate the future statistical, sociological or biological studies. 20 There are also limitations or unanswered questions of this study as well: (1) the variable indicating mothers’ smoking status was distributed unevenly: the number of mothers who smoked during pregnancy was only 17 (2.5%) while who did not smoke was 670 (97.5%). Therefore, smoking status was not an appropriate variable to be included in the analysis because it did not provide sufficient observations of smoking population, and was not taken into consideration. (2) There are many factors that we could not record in the dataset or the analysis; for example, thyroid hormone level is affected by environmental factors such as food and pollutants, and these factors are able to affect fetuses’ development. If an environmental factor was truly a confounder but not included in the dataset, we cannot adjust it. (3) TSH and FT4 concentrations in human body keep fluctuating, so the concentrations of TSH and FT4 are dynamic instead of constant and always changes. Besides, the testing happened at different times for each mother as we could not take the test at exactly the same time point for each mother. Since we were not able to monitor TSH and FT4 during the entire pregnancy, TSH and FT4 were set to be tested twice during the first and the third trimester, and deviation from the true mean value of TSH and FT4 exists. (4) In both univariate and multivariable analyses, the third trimester FT4 values showed an overwhelmingly negative effect to birthweight; this result contradicts the biological study which indicates that fetuses’ dependency of maternal thyroid hormones decreases towards the end of pregnancy and the reason is unknown. The possible explanations are: (1) the distribution of FT4 determines such overwhelming negative effect: the means of FT4 in the first and the third trimester are 1.082 ng/dL and 0.912 ng/dL, respectively, so 1 ng/dL change of FT4 or natural log of FT4 can cause a significant impact; (2) babies grow faster in the third trimester than in the first trimester, so we have greater variations in the 21 outcome to detect effects; (3) maternal thyroid hormones are possibly have interactions with fetal thyroid hormone levels, and this study did not capture this possible relationship. 4.4. Conclusions Based on this study, the univariate analysis shows that birthweight is significantly associated with third trimester FT4; birthweight is not significantly associated with first trimester TSH, first trimester FT4 or third trimester TSH. After adjusting covariates (baby’s gender, maternal pre-pregnant BMI, maternal highest education level and baby’s gestational age at birth), birthweight is associated with the third trimester FT4 only. In both trimesters, the significant interaction was found between TSH and race. 22 References American Thyroid Association [Internet]. Falls Church (VA): American Thyroid Association; c2017. Thyroid Function Tests [cited 2017 May 22]; [about 2 screens]. Available from: https://www.thyroid.org/thyroid-function-tests “Thyroxine | You and Your Hormones from the Society.” You and Your Hormones, https://www.yourhormones.info/hormones/thyroxine/. Wang, Yuan et al. “Thyroxine Alleviates Energy Failure, Prevents Myocardial Cell Apoptosis, and Protects against Doxorubicin-Induced Cardiac Injury and Cardiac Dysfunction via the LKB1/AMPK/,, Axis in Mice.” Disease markers vol. 2019 7420196. 16 Dec. 2019, doi:10.1155/2019/7420196 Ribeiro, Lorena Gabriela Rocha et al. “Excess Maternal Thyroxine Alters the Proliferative Activity and Angiogenic Profile of Growth Cartilage of Rats at Birth and Weaning.” Cartilage vol. 9,1 (2018): 89-103. doi:10.1177/1947603516684587 Sun, Xianglan et al. “Association of thyroid-stimulating hormone and cardiovascular risk factors.” Internal medicine (Tokyo, Japan) vol. 54,20 (2015): 2537-44. doi:10.2169/internalmedicine.54.4514 “Hyperthyroidism (Overactive Thyroid).” Mayo Clinic, Mayo Foundation for Medical Education and Research, 14 Nov. 2020, https://www.mayoclinic.org/diseases- 23 conditions/hyperthyroidism/symptoms-causes/syc- 20373659#:~:text=Hyperthyroidism%20(overactive%20thyroid)%20occurs%20when,a%20rapid %20or%20irregular%20heartbeat. “Hypothyroidism (Underactive Thyroid).” Mayo Clinic, Mayo Foundation for Medical Education and Research, 19 Nov. 2020, https://www.mayoclinic.org/diseases- conditions/hypothyroidism/symptoms-causes/syc-20350284. Benhadi, N et al. “Higher maternal TSH levels in pregnancy are associated with increased risk for miscarriage, fetal or neonatal death.” European journal of endocrinology vol. 160,6 (2009): 985-91. doi:10.1530/EJE-08-0953 Medici, Marco et al. “Thyroid function in pregnancy: what is normal?.” Clinical chemistry vol. 61,5 (2015): 704-13. doi:10.1373/clinchem.2014.236646 Alkemade, Anneke. “Thyroid hormone and the developing hypothalamus.” Frontiers in neuroanatomy vol. 9 15. 20 Feb. 2015, doi:10.3389/fnana.2015.00015 Chung, Hye Rim. “Adrenal and thyroid function in the fetus and preterm infant.” Korean journal of pediatrics vol. 57,10 (2014): 425-33. doi:10.3345/kjp.2014.57.10.425 Bastain, Theresa M et al. “Study Design, Protocol and Profile of the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) Pregnancy Cohort: a 24 Prospective Cohort Study in Predominantly Low-Income Hispanic Women in Urban Los Angeles.” BMC pregnancy and childbirth vol. 19,1 189. 30 May. 2019, doi:10.1186/s12884-019- 2330-7 Gestational age at birth was calculated and standardized using a hierarchy of methods. Citation is Committee Opinion No 700: Methods for Estimating the Due Date. Obstet Gynecol, 2017. 129(5): p. e150-e154. “Why People Become Overweight.” Harvard Health, 24 June 2019, https://www.health.harvard.edu/staying-healthy/why-people-become- overweight#:~:text=Genetic%20influences&text=Research%20suggests%20that%20for%20som e,of%20treating%20your%20weight%20problems. Imperial College London. "'Gene overdose' causes extreme thinness." ScienceDaily. ScienceDaily, 31 August 2011. <www.sciencedaily.com/releases/2011/08/110831160038.htm>. Tantleff-Dunn, S et al. “How did you get so thin? The effect of attribution on perceptions of underweight females.” Eating and weight disorders : EWD vol. 14,1 (2009): 38-44. doi:10.1007/BF03327793 Garn, S M et al. “Level of education, level of income, and level of fatness in adults.” The American journal of clinical nutrition vol. 30,5 (1977): 721-5. doi:10.1093/ajcn/30.5.721 25 Babić Leko, Mirjana et al. “Environmental Factors Affecting Thyroid-Stimulating Hormone and Thyroid Hormone Levels.” International journal of molecular sciences vol. 22,12 6521. 17 Jun. 2021, doi:10.3390/ijms22126521 Medici, Marco et al. “Maternal thyroid hormone parameters during early pregnancy and birth weight: the Generation R Study.” The Journal of clinical endocrinology and metabolism vol. 98,1 (2013): 59-66. doi:10.1210/jc.2012-2420 Zhou, Bin et al. “Effect of Gestational Weight Gain on Associations Between Maternal Thyroid Hormones and Birth Outcomes.” Frontiers in endocrinology vol. 11 610. 3 Sep. 2020, doi:10.3389/fendo.2020.00610 Zhang, Chen et al. “Association Between Maternal Thyroid Hormones and Birth Weight at Early and Late Pregnancy.” The Journal of clinical endocrinology and metabolism vol. 104,12 (2019): 5853-5863. doi:10.1210/jc.2019-00390

Abstract (if available)

Abstract Free thyroxine (FT4) and thyroid-stimulating hormone (TSH) are two essential thyroid hormones that affect the development of the human body. The fetus obtains thyroid hormones from mothers, and the fetus's demand for maternal thyroid hormones is different in the early and late pregnancy stages. This thesis investigates the association of FT4 and TSH with birthweight and to quantitatively compare the associations of thyroid hormones in first and third trimesters with birthweight. This study recruited mothers from either LAC+USC Medical Center, Eisner Health (main site), USC Obstetrics & Gynecology, South Central Clinic and the community in Los Angeles area. The data were collected from 2015 to September 2021 by obtaining medical records and questionnaires. The primary goal of this thesis is to investigate the association between maternal FT4 and TSH and birthweight in the first and the third trimester, and to ascertain the difference in associations between FT4 and TSH of the 1st and 3rd trimester and birthweight. In the univariable analysis, birthweight was associated with the third trimester FT4 (p = 0.0024). In the multivariable analysis, birthweight was associated with the third trimester TSH (p = 0.0140) and FT4 (p = 0.0012). Significant interactions were found between TSH in both trimesters and birthweight. As maternal FT4 in the third trimester is associated with baby’s birthweight, and FT4 and TSH affect each other biologically, the association implies that thyroid hormones are associated with baby’s development. The associations between thyroid hormones and birthweight are different in the first and the third trimester.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Association of maternal and environmental factors with infant feeding behaviors in a birth cohort study

PDF

Associations of cumulative pollution burden and environmental health vulnerabilities with gestational weight gain in a cohort of predominantly low-income Hispanic women

PDF

The interplay between tobacco exposure and polygenic risk score for growth on birthweight and childhood acute lymphoblastic leukemia

PDF

Personal exposure to particulate matter PM2.5 sources during pregnancy and birthweight

PDF

Relationship between progression of atherosclerosis and coagulation measures in a randomized-controlled trial

PDF

Generalized linear discriminant analysis for high-dimensional genomic data with external information

PDF

The risk estimates of pneumoconiosis and its relevant complications: a systematic review and meta-analysis

PDF

Leveraging functional datasets of stimulated cells to understand the relationship between environment and diseases

PDF

The impact of hormone therapy on Alzheimer’s disease biomarkers in early and late postmenopausal women

PDF

HIF-1α gene polymorphisms and risk of severe-spectrum hypertensive disorders of pregnancy: a pilot triad-based case-control study

PDF

Association between endogenous sex hormone levels and cognition: the Women’s Isoflavone Soy Health (WISH) trial

PDF

Infants in non-rhabdomyosarcoma soft tissue sarcoma

PDF

Comparing robustness to outliers and model misspecification between robust Poisson and log-binomial models

PDF

Pregnancy in the time of COVID-19: effects on perinatal mental health, birth, and infant development

PDF

Prenatal environmental exposures and fetal growth in the MADRES cohort

PDF

Psychometric study of an English version of Perceived Stress Scale in minority adolescents

PDF

Relationship of blood pressure and antihypertensive medications to cognitive change in the BVAIT, WISH, and ELITE clinical trials

PDF

Combination of quantile integral linear model with two-step method to improve the power of genome-wide interaction scans

PDF

Variation in CRY2 and MTNR1B have independent effects on insulin secretion in Mexican Americans

PDF

Detecting joint interactions between sets of variables in the context of studies with a dichotomous phenotype, with applications to asthma susceptibility involving epigenetics and epistasis

Asset Metadata

Creator Chen, Jiazheng (author)

Core Title Study of maternal free thyroxine and thyroid-stimulating hormone's relationship with infant birthweight

School Keck School of Medicine

Degree Master of Science

Degree Program Biostatistics

Degree Conferral Date 2022-08

Publication Date 07/15/2022

Defense Date 06/22/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag birthweight,free thyroxine,interaction,linear regression,maternal,multivariable analysis,OAI-PMH Harvest,thyroid-stimulating hormone,trimester,univariate analysis

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Breton, Carrie (committee chair), Li, Chun (committee member), Nuno, Michelle (committee member)

Creator Email 769253196@qq.com,jchen081@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111371439

Unique identifier UC111371439

Legacy Identifier etd-ChenJiazhe-10836

Document Type Thesis

Format application/pdf (imt)

Rights Chen, Jiazheng

Type texts

Source 20220715-usctheses-batch-953 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu