Sample summary statistics are reported in . The mean age of the study sample was 31.3 years (SD = 4.3 years) and this was the first pregnancy for half of the study participants (50.7%). The majority of the sample self-identified as White (86.8%) and just over half (54.4%) had a household income of more than $100,000 USD/year. Thirty-five percent of participants had a high-risk pregnancy, while 18% of participants had a midwife as a primary maternity care provider. The mean FOBS score was 61.2 (SD = 21.8) with 62% of participants having FOBS > 54, indicating tokophobia. Three percent of participants gave birth to low birth weight infants, while 5.6% gave birth preterm.
Prediction 1—Factors related to maternal health and healthcare experience will be associated with tokophobia
Odds ratios and 95% confidence intervals are presented in . Consistent with our prediction, tokophobia was significantly more common among individuals self-identifying as African American/Black (reference white: odds ratio (OR) = 1.90, 95% CI = 1.25, 2.97), those in the lowest household income category (<$50,000 per year, reference group $100,000+: OR = 1.39, 95% CI = 1.02, 1.91) and those in the lowest education category (No college degree, reference group advanced degree, OR = 1.37, 95% CI = 1.07, 1.76). In addition, having a high-risk pregnancy (OR = 1.65, 95% CI = 1.35, 2.03), a pre-existing health condition (OR = 1.69, 95% CI = 1.17, 2.48), prenatal depression (OR = 4.95, 95% CI = 3.84, 6.45) and being in the third trimester of pregnancy during the survey (relative to the first, OR = 1.97, 95% CI = 1.35, 2.89) were all associated with increased odds of tokophobia. There was a trend toward increased odds of tokophobia among participants with an obstetrician for primary care provider (reference: midwife, OR = 1.22, 95% CI = 0.95, 1.55) and for participants preparing for their first birth (reference: previous birth, OR = 1.15, 95% CI = 0.95, 1.39), but these associations were not statistically significant.
Self-reported Black ethnicity, low education, low income, high-risk pregnancy, prenatal depression, pre-existing health conditions and being in the third trimester were significantly associated with tokophobia. Reference categories for categorical variables were selected based on the group hypothesized to be least likely to have tokophobia. The figure presents the unadjusted odds ratio and 95% CI. * = P < 0.05; ** = P < 0.01; *** = P < 0.001.
Prediction 3—Tokophobia measured in pregnancy will be prospectively associated with gestational age at birth and birth weight
Consistent with our prediction, tokophobia was significantly associated with preterm birth before (OR= 1.85, 95% CI = 1.04, 3.50) and after adjusting for covariates (adjusted odds ratio (aOR) = 1.91, 95% CI = 1.02, 3.76). Individuals who had a high-risk pregnancy (aOR = 1.91, 95% CI = 1.07, 3.42) or who were in the third trimester of their pregnancy at the time of the survey (relative to first; aOR = 0.365, 95% CI = 0.14, 1.05) had significantly higher odds of delivering preterm. No other covariates were significantly associated with the odds of preterm birth.
Inconsistent with our prediction, low birth weight was not significantly associated with tokophobia in unadjusted (OR = 1.55, 95% CI = 0.72, 3.60) or adjusted (OR = 1.50, 95% CI = 0.65, 3.68) models. Cesarean section (OR = 2.25, 95% CI = 1.02, 4.93) and high-risk pregnancy (aOR = 2.83, 95% CI = 1.30, 6.48) were associated with higher odds of low birth weight birth. No other covariates were significantly associated with the odds of low birth weight.
Conclusions and implications
Within the fear of childbirth literature, empirical research has mainly focused on fear of childbirth predictors rather than on associated outcomes. This study is the first to demonstrate a positive relationship between childbirth fear and preterm birth, finding that individuals with tokophobia have nearly twice the odds of giving birth preterm. Importantly, this association remains significant after controlling for both elective and emergency cesarean section, which previous research indicates are associated with tokophobia . For context, this effect size was nearly identical to the risk of giving birth preterm for individuals diagnosed with a high-risk pregnancy.
The prevalence of tokophobia in our sample is much higher than was reported in a review of tokophobia studies in Europe, Canada, Australia and the USA, which found a childbirth fear prevalence ranging from 6% to 14% across studies . However, a study in Turkey, which also used the FOBS employed in the present study, reported a median score of 65, as was observed here . Our number is also slightly higher than the reported mean (56) of a similar survey among US child bearers during the COVID-19 pandemic; however, their sample only included participants anticipating a vaginal birth, and the median was not reported .
The high prevalence of tokophobia reported in this study is even more notable since our sample was relatively privileged, with self-identified white, highly educated participants over-represented relative to the general US birthing population. Tokophobia may, therefore, be even higher in more nationally representative samples. Additional studies are needed to understand whether the high levels of childbirth fear reported here were driven primarily by the pandemic, aspects of the US maternal health care system, high levels of pre-existing general anxiety that become elevated during pregnancy or other American cultural norms and expectations around birth.
Participants with lower income and education in this study were more likely to have tokophobia. In addition, the odds of a Black mother having tokophobia was 90% higher than the odds that a white mother did. This latter finding may reflect the unfortunate experience of obstetric racism that results in poor care delivery for many Black families in US settings . Some participants explicitly stated this concern; for example, one participant wrote:
“We chose our doctor because she was POC (though not black). We wanted a black doula (but there aren’t many in our city and the ones that are present are incredibly expensive). We are deeply aware of how race plays out in care.”
-34-year-old nullipara with low-risk pregnancy
Higher childbirth fear associated with obstetric racism may therefore be a presently underappreciated contributor to the stark inequities in maternal and infant health in the USA [25–27].
Tokophobia influences preterm birth: Potential biological pathways
Our finding that tokophobia is associated with higher odds of preterm birth is consistent with other research suggesting that maternal stress experienced during pregnancy impacts gestation length [28, 29]. Prenatal stress may affect the timing of parturition through an accelerated trajectory of placentally derived corticotrophin-releasing hormone (CRH) release [30, 31] (). Specifically, stress exposure activates the evolutionarily conserved hypothalamic–pituitary–adrenal (HPA) axis, leading to the production of hypothalamic CRH. This peptide triggers the release of adrenocorticotropic hormone (ACTH) from the pituitary, which in turn stimulates cortisol release from the adrenal cortex.
Prenatal stress triggers a hormonal cascade that has been hypothesized to impact gestation length. The HPA-axis generally operates with a negative feedback system, with cortisol production leading to a downregulation of CRH from the hypothalamus (steps 1–4). Among anthropoid primates, cortisol uniquely stimulates pCRH production from the placenta, which is believed to influence timing of labor by stimulating uterine contractions.
The HPA-axis is usually regulated through a negative feedback loop, with cortisol decreasing sensitivity to CRH in the hypothalamus and reducing subsequent ACTH and cortisol release. However, during pregnancy, cortisol stimulates the production of placental CRH, leading to a positive feedback loop of pCRH, ACTH, and cortisol. These peptides/hormones increase exponentially across gestation and play multiple vital roles, including facilitating fetal maturation and influencing the timing of spontaneous labor [32–34]. Elevated CRH levels and steeper increases in pCRH are associated with preterm birth . Tokophobia may, therefore, impact gestation length through influences on the trajectory of pCRH increase across gestation.
COVID-19 has exacerbated childbirth fear
The COVID-19 pandemic exacerbated existing weaknesses in the US maternity care system , potentially contributing to high tokophobia in the present sample. Rapidly changing hospital regulations following the announcement of the pandemic in the USA limited the number of support persons allowed in the delivery room while also leading parents to fear or even experience separation from their newborns after testing positive for COVID-19 . Furthermore, our study participants expressed concern that hospital-based care may increase COVID-19 infection risk. A critical aspect of the COVID-19 pandemic is, therefore, that it overlapped with and amplified many pre-existing factors leading to childbirth fear, including fear of unwanted medical intervention, lack of support, and risk of harm to self or child. As an example, one participant wrote:
“I’m terrified all the way around as it is my first pregnancy. You name it, I’ve worried about it. The COVID-19 stuff just adds more fear.”
-30-year-old nullipara with low-risk pregnancy
One of the specific pathways through which the COVID-19 pandemic has exacerbated childbirth fear is the threat many participants experienced of having to give birth alone. One participant wrote that she was:
“…afraid I’d end up with a c-section due to my lack of support from a doula, midwives, or other friends.”
-39-year-old multipara with high-risk pregnancy
Assisted childbirth is considered a human universal and is regarded as an important component of our evolutionary history [3, 36]. There is strong empirical evidence that emotional support is essential for positive birth outcomes; a Cochrane review found that continuous birth support was positively associated with odds of spontaneous vaginal birth (i.e. without forceps or cesarean) . Thus, pandemic-associated changes to the availability of emotional support during labor clearly impacted participant childbirth fear.
Evolutionary perspectives on childbirth fear and gestation length
Childbirth is an essential component of human reproduction. And yet, many individuals experience severe childbirth fear, clinically defined as tokophobia. We can imagine at least two plausible evolutionary hypotheses that could help explain why pathologically high rates of childbirth fear occur. One explanation is that tokophobia reflects the development of an extreme version of an adaptive response. Specifically, a modest amount of childbirth fear may be adaptive if it promotes a desire for assistance in labor and delivery, thereby reducing maternal and infant mortality . Since childbirth fear exists on a spectrum, pathologically high levels of childbirth fear may develop in a subset of individuals at the high end of the childbirth fear distribution. However, our findings show that tokophobia increases the odds of preterm birth, which could negatively affect offspring survival. Maternal depression and anxiety resulting from tokophobia could also negatively impact longer-term maternal and child well-being. Our results, therefore, do not support this hypothesis, but instead suggest that tokophobia would likely be selected against given probable negative health and fitness effects, as documented here.
A second, and more likely, hypothesis for the presence of tokophobia in cultural contexts like the USA is that it reflects a mismatch between the birth settings we likely experienced for much of our evolutionary history (i.e. familiar environments with known care attendants) and those where many individuals currently give birth (i.e. unfamiliar birth facilities with unknown care attendants) [3, 38]. The mismatch may also occur at the level of knowledge and expectations about birth since the shift in birth location and the medicalization of childbirth has influenced where and how knowledge about birth is learned (i.e. from medical professionals versus maternal kin and social networks) [38, 39]. This mismatch is likely most significant for socially disadvantaged individuals, who are more likely to experience bias and mistreatment in maternity care settings . A prediction of this hypothesis is that maternity care settings characterized by safety and trust will be associated with reduced tokophobia. Consistent with this hypothesis, the community-birth setting has been associated with lower childbirth fear [7, 41]. In addition, a post-hoc analysis in our sample demonstrated a significant inverse association between satisfaction with provider and childbirth fear (r = −0.20, P < 0.0001; see  for description of provider satisfaction measure). Importantly, if tokophobia is indeed the result of an evolutionary mismatch scenario, it is unlikely to be adaptive.
Notably, full-term pregnancy in humans (280 days, or 40 weeks) is defined because 39–40 weeks is the most common gestation length and the gestation length associated with the lowest incidence of neonatal mortality . This coupling suggests that there has likely been a strong selection of the genetic and physiological mechanisms that determine gestation length .
That said, shorter gestation, including preterm birth, may not be a maladaptive response in all instances; for example, Pike argued that shorter gestation length may be adaptive if it increases maternal or offspring survival in the face of compromised maternal or environmental conditions (e.g. maternal infection, undernutrition or psychosocial stress) . Yet, while potentially beneficial in some instances for ensuring maternal longevity or offspring survival, early parturition can also result in tradeoffs that compromise an offspring’s long-term health. This hypothesis suggests that there is plasticity in human gestation length, with tradeoffs between current and future conditions potentially affecting the optimal timing of parturition. If true, this evolved sensitivity likely emerged in response to fundamentally different types of ecological and psychosocial stressors than those experienced in contemporary society. This means that the response of earlier parturition following psychosocial stress from obstetric racism, for example, may reflect an evolutionary mismatch.
Our sample consists of a disproportionally high number of participants who self-identify as white, are highly educated, and are from high-income households relative to the general US birthing population. As indicated by our findings, tokophobia is higher among Black mothers, likely due to racism and other inequities in maternity care experienced by this population [25, 40]. Additionally, these fears may have been exacerbated during the pandemic due to the disproportionate risk of infection and severe COVID-19 symptoms experienced by communities of color . More research is needed on nationally representative samples to understand the true prevalence of tokophobia within the general population, and how this prevalence varies according to participant demographics and COVID-19-related factors.
Recent research has highlighted the strengths of the FOBS in both clinical settings and application across cultural contexts, mainly since there are issues of translation of several instrument items of the Wijma scale into English . A qualitative study in the USA also questioned the utility of the Wijma scale measure for this population due to the lack of inclusion of questions about factors strongly associated with childbirth fear in this context, including provider mistreatment . That said, the Wijma scale is considered by many to be the gold standard for fear of childbirth research. It would be helpful to conduct a study in the USA comparing the general utility of the Wijma to the FOBS, as has been done in Australia .
Consistent with other online survey samples conducted during the COVID-19 pandemic, we had substantial attrition between the first and second surveys . Specifically, 31% of participants who consented to be re-contacted after the prenatal survey did not participate in the postnatal survey. Individuals with tokophobia were significantly more likely to be lost to follow-up (P = 0.04); however, the difference was slight (62% in overall sample with tokophobia vs. 60% in follow-up sample). Nonetheless, the loss to follow-up among individuals with higher childbirth fear has likely biased our results towards the null since we would predict that higher childbirth fear scores would be more strongly associated with adverse birth outcomes.