The current study uses data from a prospective birth cohort that recruited 186 socioeconomically disadvantaged pregnant women during late pregnancy and subsequently enrolled their newborn children between 8/2017 and 11/2022 (33). The inclusion criteria for the current study were: 1) mothers provided saliva samples during the late second or third trimester, and 2) tooth eruption information was available for children at any of the 6-, 12-, 18-, or 24-month study visits. The final analytic sample included 142 mother-child dyads. The birth cohort study was approved by the university's Research Subject Review Board (#1248). Written informed consent was obtained from all participants, and consent for minors was obtained from their legal guardians.

Comprehensive oral examinations were performed on children at multiple visits (1 week, 2, 4, 6, 12, 18, & 24 months) up to 24 months of age by one of three calibrated dentists using standard equipment. A tooth was considered erupted if its cusp was visible above the gum line. The number and type of erupted teeth were recorded at each visit for each child.

Prenatal stress was not directly measured in our participants; instead, a proxy measure was used, based on a documented diagnosis of depression or anxiety during pregnancy in the medical record, according to previous studies (34, 35). The variable was coded as the presence or absence of depression or anxiety. Saliva samples were collected once from each woman during late pregnancy. A detailed description of salivary hormone assessment has been provided in our previous publication (36). Briefly, prior to sample collection, participants were instructed to refrain from eating, drinking, or brushing their teeth for 2 h. Each woman provided approximately 3–5 ml of unstimulated saliva during the late 2nd or 3rd trimester by spitting into a sterile centrifuge tube. Salivary hormone levels of cortisol, estradiol, progesterone, testosterone, triiodothyronine (T3), and thyroxine (T4) were measured using the MILLIPLEX MAP Multi-Species Hormone Magnetic Bead Panel (Cat# MSHMAG-21K, Merck Millipore, Darmstadt, Germany). Results were obtained on a Luminex 200 instrument (Luminex, USA) and reported based on standard curve values.

Covariates included in the model were selected based on previous literature (3, 15, 37). Maternal and child demographic and socioeconomic information (maternal age, educational level, marital status, employment status, parity, smoking status, child sex, child race/ethnicity, birthweight, and breastfeeding status) was obtained through self-reported questionnaires and medical records. Maternal medical conditions (diabetes and hypertension) were self-reported and confirmed using electronic health records. Gestational age at sample collection was calculated from the due date and the sample collection date. Breastfeeding status was determined primarily from the 6-month visit. If data at the 6-month visit were missing, information from earlier visits (2- and 4-month) was used; if unavailable, breastfeeding data from the 12-month visit were used.

All hormone levels were natural log-transformed to approximate a normal distribution. To reduce model complexity and avoid overfitting, covariates were selected for the final models through univariate analyses between each covariate and the number of erupted teeth at each visit, using a p-value threshold of 0.2. The final models included maternal age, educational level, employment status, nulliparity, smoking status, diabetes, hypertension, and child sex, race/ethnicity, birthweight, and breastfeeding status. Pearson and Spearman correlations were used to assess relationships among stress-related hormones and, separately, among the number of erupted teeth across visits. Group-based trajectory modeling was applied to identify distinct patterns of tooth eruption over time.

Because the outcome variable, the number of erupted teeth, was a count variable, generalized linear models were used to examine associations between prenatal stress, stress-related hormones, and tooth eruption at each visit, adjusting for covariates. A Poisson distribution with a log link function was used for the 12-, 18-, and 24-month visits, while a negative binomial distribution was applied at the 6-month visit to account for overdispersion. Overall associations across all visits were assessed using negative binomial mixed-effects models. Sensitivity analyses excluded infants born with low birthweight (<2,500 grams; n = 1), and cases with maternal hormone levels identified as outliers, defined as values beyond the median ± 4 × interquartile range. All statistical analyses were conducted using STATA 18.0 (College Station, TX, USA).

The characteristics of the mother-child dyads (n = 142) are summarized in Table 1. Most mothers were employed (>53%), had a high school education or lower (60%), had more than one pregnancy (>76%), and did not breastfeed around 6 months postpartum (59%). All children were born full-term, with an average birthweight of 3,308 grams. More than half of the children were African Americans (>52%).

By 6 months of age (Figure 1), 17 children (15.2%) had erupted teeth, ranging from 1 to 6 (mean = 0.33, SD = 0.87). The lower central incisors (12%) were the most common teeth erupted at this age (Figure 2A). By 12 months, 3 children (2.5%) had no erupted teeth, while the median number of erupted teeth was 6 (range: 0–12; mean = 5.76, SD = 2.69). More than 95% of children had erupted lower central incisors, and approximately 80% had erupted upper central incisors (Figure 2B). At 18 months, all children attending the study visit (n = 102) had erupted teeth, with a median of 14 (range: 3–20; mean = 13.13, SD = 3.39). By 24 months, the median number of erupted teeth was 16 (mean = 17.18, SD = 1.85), and 25% of children had a full set of 20 primary teeth (Figure 1). Fewer than 40% had erupted upper and lower second molars (Figure 2D).

Spearman's rank-order correlation was used to assess associations among tooth counts across visits. Significant correlations were observed (p < 0.01) for all pairs except between 6 and 18 months. Significant correlation coefficients were weak to modest, ranging from 0.30 to 0.57 (Figure 3A). However, the intraclass correlation coefficient (ICC) across visits was close to “0” (ICC = 7.24 × 10⁻¹⁰), suggesting limited stability in individual eruption patterns (Figure 3B). Group-based trajectory modeling identified two distinct tooth eruption patterns during the first two years of life, based on the lowest Bayesian information criterion (BIC, Figure 3C): a continuous growth pattern (97.3% of children) and a rapid growth pattern between 12 and 18 months (2.7% of children).

During pregnancy, 36.6% of women had a documented diagnosis of depression or anxiety. Levels of stress-related hormones measured in maternal salivary samples collected during the late 2nd or 3rd trimester are presented in Table 2. All hormones were significantly correlated (p < 0.001), with correlation coefficients ranging from 0.58 to 0.89 (Figure 4). No significant associations were observed between depression or anxiety diagnoses and hormone levels.

Due to the small-to-moderate correlations and low ICC for the number of erupted teeth across visits, associations between prenatal depression/anxiety, salivary hormones, and tooth eruption were first assessed separately at each child visit. Prenatal maternal depression/anxiety was not significantly associated with the number of erupted teeth at any time point (p > 0.05). In contrast, several prenatal stress-related hormones, including cortisol, estradiol, progesterone, testosterone, and T3, were positively associated with tooth counts (Table 3).

Specifically, higher prenatal cortisol levels were associated with a greater number of erupted teeth at 6 months (β = 0.79, 95% CI: 0.12–1.45), corresponding to an average difference of ∼4 teeth between the lowest and highest cortisol levels. At 12 months, higher estradiol (β = 0.08, 95% CI: 0.002–0.15) and testosterone (β = 0.04, 95% CI: 0.004–0.08) were positively associated with tooth count, corresponding to average differences of 0.4 and 0.3 teeth, respectively. At 24 months, progesterone (β = 0.02, 95% CI: 0.003–0.03) and testosterone (β = 0.01, 95% CI: 0.001–0.02) were associated with tooth count, corresponding to average differences of 0.1 and 0.07 teeth. T3 levels were also positively associated with tooth counts at 18 months (β = 0.03, 95% CI: 0.002–0.05) and 24 months (β = 0.01, 95% CI: 0.003–0.02), corresponding to average differences of 0.2 and 0.09 teeth, respectively, between the lowest and highest hormone levels.

When overall associations across all visits were assessed, the results were non-significant, likely reflecting the low ICC and weak correlations across visits. Most findings remained robust in sensitivity analyses; however, associations between cortisol and tooth count at 24 months (β = 0.03, 95% CI: 0.0002–0.05) and between estradiol and tooth count at 24 months (β = 0.02, 95% CI: 0.0005–0.04) became significant. In contrast, associations with estradiol (p = 0.09) and testosterone (p = 0.05) at 12 months were attenuated.