- Top of page
- Methods and Materials
- Strengths and Limitations of the Study
- Future Directions
- Implications for Practice
Objective: To investigate race and ethnicity as risk factors for breech presentation.
Design: Case-control study using a population-based birth certificate registry that included linkage to Medicaid/Women, Infants, & Children (WIC) eligibility data for a socioeconomic proxy variable.
Participants: 912,107 mothers of singletons born in Florida 1999 to 2003.
Methods: Maternal race and ethnicity were evaluated as risk factors for breech presentation using logistic regression. The dependent variable was birth presentation. Covariates were variable measures that have been repeatedly identified as risk factors for breech presentation in the literature and are known to be highly accurate birth certificate variable measures.
Results: White women were 69% more likely to have a breech baby (adjusted odds ratio [OR]=1.69, 95% confidence interval [CI] 1.63, 1.76) than Black women. Higher socioeconomic status was a risk factor in the bivariate analyses, but not in the adjusted analysis. Prematurity, nulliparity, female infant, and advancing maternal age were risk factors in the final model. The final model accounted for <5% of the total variance (Max Rescaled R2=4.18%), and thus was poorly fit (Hosmer and Lemeshow goodness of fit <0.0001).
Conclusions: White women were at increased risk of having a breech baby. However, important variables appear to be missing from the model.
Breech presentation is the most common birth malpresentation (3%–4%) (Rayl, Gibson, & Kickok, 1996) and is associated with increased perinatal morbidity and mortality (Albrechtsen, Rasmussen, Dalaker, & Irgens, 1998a). Surprisingly, although breech presentation is associated with a higher incidence of preterm and low-birth-weight babies, it occurs disproportionately in White women (Martin et al., 2005) and in middle- and upper-class women (Pop et al., 2004; Roberts, Algert, Peat, & Henderson-Smart, 1999), populations that are generally protected from other adverse pregnancy outcomes. This phenomenon is the breech paradox. Why is breech presentation an exception to the tendency for adverse pregnancy outcomes to occur disproportionately in disadvantaged populations?
Fewer than 50 studies were identified that evaluated risk factors for breech presentation. Retrospective birth registry studies have consistently identified younger gestational age, lower birth weight, older maternal age, a female fetus, and nulliparity as risk factors for breech presentation babies when compared with cephalic presentation babies (Amoa, Sapuri, & Klufio, 2001; Jonas & Roder, 1993; Rayl et al., 1996; Roberts et al., 1999; Vendittelli et al., 2008). Birth registry studies have also found that breech presentation is more common in parents who were born breech (Nordtveit, Melve, Albrechtsen, & Skjaerven, 2008), in a mother who has already had a breech baby (Albrechtsen, Rasmussen, Dalaker, & Irgens, 1998c), or in a mother who had a previous Cesarean section (Vendittelli et al.). Other variables of potential interest such as amniotic fluid volume (Luterkort, Persson, & Weldner, 1984; Rayl et al.), placental implantation site (Rayl et al.; Roberts et al.; Westgren, 1985), maternal smoking and drinking (Pop et al., 2004; Rayl et al.; Roberts et al.), and maternal disease (Pop et al.; Rayl et al.; Roberts et al.) are inconclusive risk factors and at times are inappropriately evaluated in databases, such as U.S. birth certificates, where these variables are known to be inaccurately documented (DiGiuseppe, Aron, Ranbom, Harper, & Rosenthal, 2002; Roohan et al., 2003; Yasmeen, Romano, Schembri, Keyzer, & Gilbert, 2006).
Inadequate prenatal care was evaluated as a risk factor for breech presentation in only two identified studies. A Papua New Guinea labor ward register study (N=1,004) (Amoa et al., 2001) and a Washington State birth certificate study (N=11,771) (Rayl et al., 1996) indicated mothers of breech babies received less prenatal care than did mothers of cephalic presentation babies. The Washington State findings must be accepted cautiously, however, because birth certificate data generally does not accurately document prenatal care (Northam & Knapp, 2005). Other important considerations for pregnancy outcomes such as genetics, the environment, stress, stress-related hormones, diet, and rest have not been evaluated as risk factors for breech presentation.
Neonatal orthopedic and neurologic evaluations and maternal anthropometric studies have been conducted less frequently and use smaller sample sizes than do birth registry studies; they also arrive at less conclusive results. The largest and best designed case-control study done to date of neonatal orthopedic and neurologic characteristics shows there is virtually no difference between breech presentation babies and cephalic presentation babies (Bartlett, Okun, Byrne, Watt, & Piper, 2000; Bartlett, Piper, Okun, Byrne, & Watt, 1997). Maternal pelvic size and shape were evaluated in only one identified study of risk factors for breech presentation. The pelvic inlet and pelvic outlet were not significantly different in Swedish mothers of breech babies and mothers of cephalic presentation babies (Luterkort et al., 1984). Recent magnetic resonance imaging (MRI) studies indicate there is no significant difference between White women and African American women in important pelvimetry measurements such as the interspinous diameter and the anteroposterior conjugate; the pelvic inlet and the pelvic outlet are statistically significantly wider in White women than in African American women (Handa et al., 2008). The authors did not indicate if the difference was clinically significant.
The cause of turning to cephalic presentation remains unknown (Sekulic, 2000). The purpose of this study was to determine if maternal race and ethnicity add further explanatory value to a model of accepted risk factors for breech presentation. The study hypothesis is that risk for breech presentation varies by maternal race and ethnicity after controlling for socioeconomic confounders in the logistic regression model.
Race and ethnicity are important variable measures to consider as possible risk factors for breech presentation for four reasons: (a) breech presentation unexpectedly occurs more commonly in White women than in Black or Hispanic women and thus offers a unique opportunity to explore adverse outcomes in a typically protected population; (b) differential health outcomes are now understood to be influenced, in part, by socio-political-economic processes for which race and ethnicity act as proxy variable measures; (c) the breech presentation baby is typically delivered by Cesarean section, a higher cost birth option, thus all potential risk factors should be investigated in an attempt to prevent or reverse breech presentation and decrease the rate of Cesarean sections (Agency for Health Care Administration: State Center for Health Statistics, 2006; Martin et al., 2009); and (d) race and ethnicity have not been published as risk factors in midwifery and obstetric texts, although this is common practice for other conditions (Cunningham et al., 2005).
Racial categorization is based mostly on skin color; genetics has only a small (15%) influence on apparent racial differences (Sarich & Miele, 2004). As such, group membership to any particular race assigns certain opportunities or disadvantages that far outweigh any shared genetic traits (Molnar, 2002). Ethnicity, in contrast, is a self-identified association with a group by phenotypic, cultural, linguistic, or behavioral characteristics (Molnar). A growing body of evidence suggests that social constraints differentially experienced by race and ethnic groups might contribute to adverse pregnancy outcomes such as low-birth-weight and preterm births (David & Collins, 2007; Dominguez, 2008; Williams & Mohammed, 2008). The rate of preterm and low-birth-weight births varies by state, but Blacks consistently have poorer pregnancy outcomes than do Whites or Hispanics in the United States (Martin et al., 2009). Although Black and low socioeconomic status women typically bear the brunt of adverse pregnancy outcomes, women who work in professional or executive occupations might have reproductive outcomes similar to those experienced by working-class women; white-collar workers appear to be protected (Boivin, Sanders, & Schmidt, 2006). The reasons for these shared pregnancy outcomes are not well understood, and pregnancy outcomes for professionals and executives have not been extensively studied.
We conducted a large population-based case-control study to evaluate the role of race, ethnicity, and socioeconomic factors in influencing breech presentation. Here we present our findings and suggest future directions for research.
- Top of page
- Methods and Materials
- Strengths and Limitations of the Study
- Future Directions
- Implications for Practice
This investigation employed the second largest database to date to evaluate risk factors for breech presentation and was the first study to include race and ethnicity as risk factors for breech presentation while controlling for other confounding and contextual variables. We found that White women were 69% more likely than Black women to have a breech baby. In the bivariate analyses Hispanic ethnicity was not a risk factor for breech presentation. However, after controlling for the other variables in the logistic regression model, Hispanic women were 46% more likely to have a breech baby than Black women. This finding is supported in existing literature that indicates Hispanic birth outcomes are more similar to those of White than to Black women (Alexander et al., 2003). The findings from the current study must be cautiously interpreted because the overall fit of the model was poor and explained less than 5% of total variance. Race, ethnicity, and the other variable measures in the model do not appear to be substantial risk factors for breech presentation. However, interpretation of the results might suggest future directions for research.
Risk factors for breech presentation appear to vary by race and ethnicity for all variable measures. Advanced education produced the greatest risk of having a breech baby for Hispanic women followed by Black women in the bivariate analyses. In the United States educational attainment is differentially distributed between races and ethnicities where Whites acquire more education than Blacks, and Blacks complete more years of education than Hispanics (U.S. Census Bureau, 2005). Although advanced education was not a risk factor in the final model, it is interesting to note the risk associated with education for each race and ethnic group was the inverse of the pattern of educational achievement viewed in the population. This association suggests that a sociocultural component might be influencing breech presentation. Whether this finding is related to socioeconomic status or role incongruency is unclear because the education variable and the Medicaid/WIC eligibility variable measures might not be adequately capturing these domains.
Role incongruency represents a lack of cultural consonance where individuals behave or think differently than their ascribed sociocultural role (Dressler, 1991). This incongruency has been associated with an increased risk of impaired physical (Dressler, Oths, Ribeiro, Balieiro, & Dos Santos, 2008) and mental health (Dressler, Balieiro, Ribeiro, & Dos Santos, 2007). Educational attainment and use of Medicaid/WIC eligibility as a socioeconomic variable are likely not sensitive enough to capture these finer sociocultural domains. However, this points to the need for more research in this area.
Women who achieve high levels of education often delay childbearing (Lampic, Svanberg, Karlstrom, & Tyden, 2006). Although advanced education and ineligibility for Medicaid/WIC were not risk factors in the final model, advancing maternal age remained a risk factor. The unadjusted and adjusted ORs for maternal age, however, were in the same direction and of similar magnitude. This lack of change in magnitude and direction suggests maternal age acts independently of education, and all other variables in the model including race and ethnicity, as a risk factor. Black women of advanced maternal age were at more risk of having a breech baby (age 35–39, unadjusted OR=2.12, 95% CI 1.92, 2.35) than were Hispanic women (unadjusted OR=1.45, 95% CI 1.35, 1.55) or White women (unadjusted OR=1.31, 95% CI 1.25, 1.36). However, Black women are less likely to have a baby at that time in their life (20%) than Hispanic women (26%) or White women (29%). This difference in reproductive patterns might, in part, explain why Blacks are less likely to have a breech baby. The weathering hypothesis suggests the increased risk of adverse pregnancy outcomes experienced by Black women beyond their teens and early twenties is related to consistent sociocultural barriers experienced by these women (Geronimus, 2001). Although the weathering hypothesis was not tested in this study, future investigations might benefit from use of this conceptual framework.
Women who have their first child after they are age 35 are often met with the dual challenges of anxiety produced by the culture of fear surrounding reproduction after age 35, and a social network that is at a different point in their reproductive life and thus not able to offer as much support (Suplee, Dawley, & Bloch, 2007). Although pregnancy after age 35 is widely believed to be more risky (Friese, Becker, & Nachtigall, 2006; Nassar & Usta, 2009), the impact of age alone as a risk factor for adverse pregnancy outcomes has not been rigorously tested by controlling for confounding factors such as lifestyle (Carolan, 2003). However, the culture of fear prevails and women older than age 35 are often subjected to extensive fetal diagnostic tests. These tests, meant to determine the health of the fetus, often provoke a sense of uncertainty about the viability of the pregnancy which in turn results in anxiety, fear, distancing, and decreased attachment to the baby (Carolan & Nelson, 2007; Hjelmstedt, Widstrom, & Collins, 2006; Lawson & Turriff-Jonasson, 2006). Additional research should be conducted that evaluates advanced maternal age as a risk factor while controlling for lifestyle variables. Maternal emotions and intrauterine attachment in breech pregnancies should also be investigated.
Similar to advancing maternal age, nulliparity appeared to affect Black women differently than White or Hispanic women. In the bivariate analyses nulliparity was not a risk factor for Black women but was a moderate risk factor for White women and for Hispanic women. Unlike advancing maternal age, nulliparity was a larger risk factor in the final model (adjusted OR=1.71, 95% CI 1.67, 1.76) than in the bivariate analyses (unadjusted OR=1.49, 95% CI 1.45, 1.52). This discrepancy in the response of maternal age and nulliparity to introduction into the model suggests nulliparity and advancing maternal age are measuring different phenomena and that they are acting independently of race and ethnicity. Although nulliparity is presented as a mechanical risk factor for breech presentation in the breech literature, nulliparity is also associated with psycho-social-cultural factors such as increased fear of childbirth (Nieminen, Stephansson, & Ryding, 2009) provoked in part by lack of social support and coping mechanisms (Laursen, Hedegaard, & Johansen, 2008). This fear of childbirth might lead to an increased desire for a Cesarean section (Rouhe, Salmela-Aro, Halmesmaki, & Saisto, 2009); an increased risk of having an emergency Cesarean section has also been noted in women who fear childbirth (Laursen, Johansen, & Hedegaard, 2009). The possible association between fear and mode of childbirth suggests that emotions and coping mechanisms should be included in future studies of risk factors for breech presentation.
The risk of having a preterm breech baby was far greater in Black women (35%) than in White women (18%) in this study. Yet White women were nearly twice as likely to have a breech baby (3.2%) than were Black women (1.7%). A great proportion of the risk for Black women having a breech baby might simply be due to their propensity to have preterm babies, thus not giving the baby sufficient time to turn. Low-birth-weight and preterm births are, in part, accounted for by insufficient prenatal care (Heaman, Newburn-Cook, Green, Elliott, & Helewa, 2008), but it is unclear what role prenatal care plays in breech presentation because a sound variable measure for this parameter was not available in the database.
Culture influences birth outcomes. This has been demonstrated repeatedly by the birth outcomes of new immigrants to the West exceeding those of their race or ethnic group already in residence (Ray et al., 2007). The comparison is particularly stark between Black immigrates and African Americans, where immigrants are about one third less likely to experience preterm and low-birth-weight births than African Americans (Urquia et al., 2009). This finding suggests that sociocultural, environmental, or lifestyle factors, rather than genetics, might be primarily influencing these birth outcomes. Unfortunately, immigrant health benefits begin to diminish within months of arrival and are generally completely erased in one generation (Gravlee, 2009; Ray et al.). Although a healthy immigrant effect appears to contribute to the positive health and birth outcomes of immigrants (Callister & Birkhead, 2002; McDonald & Kennedy, 2004), social factors also play a role as demonstrated by well-educated Canadian immigrants having poorer birth outcomes than all other immigrants and native-born Canadians (Auger, Luo, & Daniel, 2008). These findings suggest role incongruency might influence birth outcomes. Similarly, the East Indian immigrant paradox, characterized by adverse pregnancy outcomes in an otherwise protected U.S. population that receives adequate prenatal care and is highly educated, points to the possibility that sociocultural factors might play a role in determining health and birth outcomes (Gould, Madan, Qin, & Chavez, 2003).
In contrast to the East Indian immigrant paradox, the Hispanic paradox illustrates the impact that positive cultural practices such as a strong social network, high valuation of children, and deeply held religious values have on birth outcomes (Gallo, Penedo, de los Monteros, & Arguelles, 2009). Although Hispanics are a disadvantaged population in terms of income, access to health care, and discrimination, they have birth and health outcomes nearly as positive as Whites in the United States (Page, 2004). The role of culture in influencing birth outcomes was also evidenced following September 11, 2001. Michigan women with Arab surnames or maiden names were protected from experiencing an increased risk of having a preterm or low-birth-weight baby 6 months after 9/11 (El-Sayed, Hadley, & Galea, 2008), whereas Californian women with Arab surnames or maiden names experienced an increased risk in these adverse pregnancy outcomes during the same time period (Lauderdale, 2006). In Michigan the Arab population concentration is nearly twice as dense as that in California suggesting a sociocultural support system might cushion the probability of experiencing adverse pregnancy outcomes.
The breech paradox represents an adverse pregnancy outcome experienced disproportionately by White women and by women in middle or upper socioeconomic strata, groups typically protected from adverse pregnancy outcomes. Although race and ethnicity themselves are probably not important risk factors for breech presentation, the breech paradox provides an opportunity to explore possible reasons why Black women and women of lower socioeconomic strata are protected from this adverse pregnancy outcome whereas White women and women who are not of lower socioeconomic strata might be at increased risk. Typically, low social status produces physiologic stress (Sabbah, Watt, Sheilham, & Tsakos, 2008). However, in industrialized countries professional or executive occupations confer general health and reproductive health benefits to men, but not to women (Frankenhaeuser et al., 1989; Hopcroft, 2006; Light et al., 1995; Steptoe et al., 2003; Weeden, Abrams, Gree, & Sabini, 2006). Little is understood about why this might occur, but further exploration might be appropriate to discover additional risk factors for breech presentation.
The apparent association between race and ethnicity and breech presentation might represent the increased likelihood for White women to acquire advanced education and work in occupations with higher status (Bureau of Labor Statistics, 2008; U.S. Census Bureau, 2005). Thus, they are more exposed to what is termed the “stress of higher status” whereas minority women are less likely to experience this type of stress. The “stress of higher status” hypothesis asserts that the benefits of higher status employment, such as autonomy and nonroutine work, are offset, in part, by increased interpersonal conflict on the job and increased work–home conflict (Schieman & Reid, 2009). It is possible that mothers of breech babies are bearing the dual burden of occupational and family accomplishments and might be experiencing work–home conflict (Frone, 2003). High expectations and multiple roles have been found to produce more stress in women than in men (Stewart, Ahmad, Cheung, Bergman, & Dell, 2000). Women who have an advanced education and elite careers are particularly at risk (Hammig & Bauer, 2009). This might be due in part to the resistance women face when attempting to ascend in the occupational ranks (Carnes, Morrissey, & Geller, 2008; Eagly & Carli, 2007). Because Black and Hispanic women in the United States are less likely to have professional or executive occupations (Bureau of Labor Statistics) they are less likely than White women to experience a glass ceiling effect and thus are in part protected from the adverse stressors of professional or executive occupations.
Although no studies were identified that evaluated the possibility that stress might influence breech presentation, some researchers have hypothesized that sympathetic dominance might promote breech presentation secondary to constriction of the lower uterine segment thus impeding the baby from turning to cephalic presentation (Peterson, 1981). Physical, emotional, and mental stress are mediated via separate neuroendocrine pathways (Papousek, Schulter, & Premsberger, 2002; Walker, Anand, & Plotsky, 2001). Although the physiologic response to physical and emotional stress is highly evolved and well organized, the response to mental stress is less well coordinated. In fact, the sympathetic response to mental stress, such as that experienced by professional and executive women, produces primarily epinephrine rather than norepinephrine, which is produced by physical and emotional stress (Panter-Brick & Pollard, 1999). The lower uterine segment is under autonomic control. Unlike norepinephrine, epinephrine stimulates uterine α-adrenergic receptors (promoting contractions) and β-adrenergic receptors (promoting relaxation). Therefore the uterus is simultaneously given contradictory instructions that lead to a tonic state (Gibbs, Karlan, Haney, & Nygaard, 2008) and could influence the baby's position.
Only two studies on risk factors for breech presentation have included the race and ethnicity of the study participants. Black race was found to be protective against breech presentation when compared to White race in one Washington State study (unadjusted OR=0.4, 95% CI 0.3, 0.5) (Rayl et al., 1996) and in a South African study (unadjusted OR=0.2,; 95% CI not indicated) (Hofmeyr, Sadan, Myer, Galal, & Simko, 1986). The findings of the current study support this surprising result. The two previous studies are limited, however, because they did not account for confounders, such as socioeconomic status, or other factors that might influence risk like preterm birth and maternal age.
Unlike other studies, this study indicates the quality of fit of the model and the variance accounted for by the model. Although White race was a risk factor for breech presentation in the logistic regression model, the model was poorly fit in that it accounted for less than 5% of the total variance, which is an important limitation. It is unclear how this compares to other studies on the topic, because their goodness of fit was not indicated; however, this does mean that interpretation of individual variables should be cautiously undertaken. A poorly fit model might reflect improper specification of variables in the model, poor data quality, or important variables missing from the model. Variables in this model were respecified numerous times, and the quality of the data was insured by using only highly valid variables. However, the goodness of fit remained low. Previously published findings of risk factors for breech presentation ought to be interpreted with caution. Seven studies were identified that used logistic regression to evaluate risk factors for breech presentation (Albrechtsen, Rasmussen, Dalaker, & Irgens, 1998b; Albrechtsen et al., 1998c; Faber-Nijholt, Huisjes, Touwen, & Fidler, 1983; Nordtveit et al., 2008; Pop et al., 2004; Rayl et al., 1996; Roberts et al., 1999; Vendittelli et al., 2008). Of these, only Roberts et al. indicated goodness of fit (p=.02), thus indicating poor fit.