Volume 198, Issue 2 , Pages 171.e1-171.e6, February 2008
Physical abuse during pregnancy and preterm delivery
Article Outline
Objective
This study was undertaken to assess the relationship between physical abuse during pregnancy and preterm delivery.
Study Design
We conducted a hospital-based survey on physical abuse during pregnancy, which included 2660 women with consecutive live births. Women were interviewed and violence was assessed using the Abuse Assessment Screen. Data on sociodemographic, behavioral, and obstetric variables were also obtained. Mothers of preterm (<37 weeks; n = 217) were contrasted with mothers of term newborn infants (n = 2428). Logistic regression analysis was performed to estimate adjusted odds ratios.
Results
Twenty-four percent of mothers of preterm newborn infants had experienced physical abuse during pregnancy compared with 8% of mothers of term newborn infants (P < .0001). Violence was associated with preterm birth even after controlling for age, marital status, education, income, parity, planned pregnancy, antenatal care, smoking, alcohol, and illicit drugs use (odds ratio = 3.14, 95% confidence interval, 2.00-4.93).
Conclusion
Women who have had physical abuse during pregnancy present a large increase in the risk of preterm delivery, independently from a large set of sociodemographic and behavioral characteristics usually recognized as determinants of preterm birth.
Key words: physical abuse, pregnancy, preterm, violence
Preterm birth remains an unsolved relevant public health issue.1 Knowledge about etiologic factors is insufficient, effective prophylactic interventions are unavailable, and in most countries, increasing rates are observed.2
See Journal Club, page 239
Only few factors were established as preterm birth determinants and most are not amenable to intervention, like ethnicity, multiple pregnancy, or a previous preterm delivery. This led to a resurgence of interest on the role of social and environmental factors like racism, violence, poverty, stress, and physical exertion in the occurrence of preterm birth and the design of preventive strategies.1, 3
Studies in Europe and North America showed that the prevalence of violence against pregnant women varied between 0.9% and 22.0%,4 and physical abuse is increasingly reported as a potentially modifiable risk factor for pregnancy adverse outcome.5, 6, 7, 8 Few studies on the association between physical abuse and preterm birth have been reported, and those that have found an effect were mainly conducted among low-income and teenage mothers.9, 10, 11, 12 Few population-based studies with a large sample size and allowing control for several confounding variables have been conducted, and some have not yielded a significant association between physical violence and preterm birth.13, 14 Differences in population characteristics and sample size, varying definitions of physical violence and birth outcome, and different study design and control for confounding may also explain the discrepant findings. In this study we aimed to assess and clarify the relationship between physical abuse during pregnancy and preterm delivery.
Materials and Methods
This survey included 2660 white women consecutively delivered of live singleton infants, at the Department of Obstetrics of Hospital de São João, in Porto, during a 10-month period in 1999-2000. This level III teaching hospital admits almost all pregnant women from the neighborhood area and is a referral hospital for several level II units.
In the first 96 hours after delivery, women were invited to participate and asked written informed consent. The central study hypothesis—the association between physical abuse and adverse pregnancy outcome—was not explicitly presented during the study description to eligible participants. Information was obtained by face-to-face interview. All interviews were performed by trained social workers in a private setting at the hospital, and women were guaranteed confidentiality. There was no refusal to participate.
Participants completed a questionnaire comprising questions on sociodemographic (age, education, employment status, cohabitation status with infant’s father, and family income), current pregnancy (parity, anthropometrics, antenatal care use, and intention to get pregnant), and behavioral characteristics during pregnancy (tobacco smoking, alcohol consumption, and illicit drugs use). Concerning antenatal care, women were classified as having no antenatal care visit, beginning antenatal care during pregnancy first trimester, or entering antenatal care later. Women’s experience of abuse was assessed by using the Abuse Assessment Screen.15 Mothers were asked whether they had been hit, slapped, kicked, or otherwise physically abused since they became pregnant. If so, they were asked to indicate the perpetrator, the number of times they had been abused, the area of injury, and the most severe incident they had suffered during pregnancy. Women were also asked if they had ever been physically abused. Data on gestational age, birthweight, and medical complications during current pregnancy were collected from maternity clinical charts.
For this analysis, the outcome variable was preterm birth defined as birth before 37 weeks’ gestation. Gestational age was determined according to the best clinical estimate considering that when the discrepancy between amenorrhea and ultrasounds ascertained gestational age was higher than 1 week, the latter was adopted. The current analysis compared 217 preterm births with 2428 term births. Fifteen births were excluded because there was no reliable information on gestational age.
We used logistic regression analysis to evaluate the association between physical abuse during pregnancy and preterm delivery, measured by using the odds ratio (OR) and its 95% confidence interval (CI) both crude and adjusted for other significant exposures. Statistical analysis was performed with Stata Statistical Software (release 7.0, Stata Corp, Cary, NC).
This study was approved by The Ethical Committee of the Medical School of Porto University.
Results
In this survey, the prevalence of physical abuse during pregnancy among women delivered of live singleton infants was 9.7% (257/2660). One in 8 women reported a lifetime experience of physical abuse. From women battered during pregnancy, 50% reported to suffer acts of violence more than 3 times during the whole pregnancy, 57.6% mentioned that the most severe act of violence they had suffered was a slap or push but without injuries or lasting pain, and 42.4% referred a more severe incident (39.3% acts such as punching, kicking, bruises, cuts, and/or continuing pain; and 3.1% beaten up, severe contusions, broken bones, head, internal, and/or permanent injury).
The prevalence of preterm birth was 8.2% (217/2645 births). Preterm delivery was significantly more frequent among physically abused than nonabused pregnant women (21.4% vs 6.8%; P < .0001). Women reporting physical abuse during pregnancy were significantly more likely to be younger than 20 years of age, in a noncohabiting relationship, less educated, with lower family income, higher parity, unplanned pregnancy, to have no antenatal care, to enter later into antenatal care, to gain less weight during pregnancy, to smoke more than 10 cigarettes per day during pregnancy, to use alcohol and illicit drugs during pregnancy, and to have no employment (TABLE 1, TABLE 2).
TABLE 1. Crude association between socio-demographic characteristics and physical abuse during pregnancy
| Physical abuse during pregnancy | |||
|---|---|---|---|
| Maternal characteristics | Yes | No | P value |
| n = 257 | n = 2403 | ||
n (%) | n(%) | ||
| Maternal age (y) | |||
| <20 | 44(17.1) | 182(7.6) | |
| 20-24 | 54(21.0) | 489(20.3) | |
| 25-29 | 59(23.0) | 814(34.0) | |
| 30-34 | 63(24.5) | 627(26.2) | |
| ≥35 | 37(14.4) | 287(12.0) | < .0001 |
| Cohabitation with infant’s father | |||
| Yes | 194(75.5) | 2268(94.4) | |
| No | 63(24.5) | 135(5.6) | < .0001 |
| School education (y) | |||
| 0-4 | 136(52.9) | 443(18.4) | |
| 5-6 | 60(23.3) | 552(23.0) | |
| 7-9 | 39(15.2) | 413(17.2) | |
| 10-12 | 20(7.8) | 475(19.8) | |
| ≥13 | 2(0.8) | 520(21.6) | < .0001 |
| Family income (Euros) | |||
| <600 | 145(56.4) | 322(13.4) | |
| 600-1999 | 102(39.7) | 1219(50.8) | |
| 2000-2999 | 9(3.5) | 507(21.1) | |
| >3000 | 1(0.4) | 353(14.7) | < .0001 |
| Paid job | |||
| Yes | 104(42.1) | 1809(77.7) | |
| No | 143(57.9) | 519(22.3) | < .0001 |
TABLE 2. Crude association between obstetrical, anthropometric and behavioral characteristics and physical abuse during pregnancy
| Physical abuse during pregnancy | |||
|---|---|---|---|
| Maternal characteristics | Yes | No | P value |
| n = 257 | n = 2403 | ||
| n(%) | n(%) | ||
| Parity | |||
| 1 | 88(34.2) | 1254(52.2) | |
| 2 | 63(24.5) | 765(31.8) | |
| 3 or more | 106(41.2) | 384(16.0) | < .0001 |
| Planned pregnancy | |||
| Yes | 48(18.7) | 1862(77.5) | |
| No | 209(81.3) | 541(22.5) | < .0001 |
| Antenatal care | |||
| No visits | 36(14.6) | 23(1.0) | |
| First visit after first trimester | 100(40.5) | 244(10.2) | |
| First visit during the first trimester | 111(44.9) | 2123(88.8) | < .0001 |
| Body mass index (kg/m2) | |||
| <20.0 | 29(11.9) | 278(11.7) | |
| 20.0-24.9 | 136(55.7) | 1335(56.0) | |
| 25.0-29.9 | 57(23.4) | 579(24.3) | |
| ≥30.0 | 22(9.0) | 193(8.1) | .956 |
| Weight gain during pregnancy (g/wk) | |||
| ≤225 | 74(33.9) | 476(20.1) | < .0001 |
| >225 | 144(66.1) | 1896(79.9) | |
| Smoking during pregnancy | |||
| Did not smoke | 186(72.4) | 2065(86.0) | |
| 1-9 cigarettes per day | 31(12.1) | 230(9.6) | |
| ≥10 cigarettes per day | 40(15.6) | 107(4.4) | < .0001 |
| Ethanol intake during pregnancy | |||
| Yes | 62(24.1) | 162(6.7) | < .0001 |
| No | 195(75.9) | 2239(93.3) | |
| Illicit drugs use during pregnancy | |||
| Yes | 17(6.6) | 8(0.3) | < .0001 |
| No | 240(93.4) | 2395(99.7) | |
The OR for the univariate association between physical abuse during pregnancy and preterm delivery was 3.72 (95% CI, 2.59-5.33). After controlling for maternal age, cohabitation status, education level, family income, parity, planned pregnancy, antenatal care, tobacco, alcohol, and illicit drugs use, the association between physical abuse during pregnancy and preterm delivery was attenuated but remained strong and statistically significant (adjOR = 3.14; 95% CI, 2.00-4.93). Higher ORs of preterm have been shown with increasing frequency of abuse experience and for more severe aggression (Table 3). Eighty-eight percent of women experiencing physical abuse during pregnancy had been victims of intimate partner violence. The risk of preterm delivery was higher when violence was inflicted either by an intimate partner or others (usually relatives; Table 3).
TABLE 3. Crude and adjusted odds ratios between violence characteristics and preterm delivery
| Preterm | Term | Crude OR (95% CI) | Adjusted OR (95% CI)a | |
|---|---|---|---|---|
| n = 217 | n = 2428 | |||
| n(%) | n(%) | |||
| Physical abuse during pregnancy | ||||
| No | 164(75.6) | 2234(92.0) | 1 | 1 |
| Yes | 53(24.4) | 194(8.0) | 3.72 (2.59-5.33) | 3.14 (2.00-4.93) |
| Frequency of physical abuse during pregnancy | ||||
| 0 | 164(75.6) | 2234(92.0) | 1 | 1 |
| 1-3 | 23(10.6) | 92(3.8) | 3.41 (2.04-5.65) | 2.97 (1.72-5.12) |
| 4-6 | 16(7.4) | 58(2.4) | 3.76 (2.03-6.89) | 3.06 (1.50-6.24) |
| ≥7 | 14(6.4) | 44(1.8) | 4.33 (2.22-8.36) | 3.75 (1.80-7.81) |
| Severity of physical abuse | ||||
| No abuse | 164(75.6) | 2234(92.0) | 1 | 1 |
| Less severe | 28(12.9) | 117(4.8) | 3.26 (2.04-5.17) | 2.84 (1.68-4.80) |
| More severe | 25(11.5) | 77(3.2) | 4.42 (2.66-7.30) | 3.69 (2.03-6.72) |
| Physical abuse perpetrator | ||||
| No physical abuse | 164(75.6) | 2234(92.0) | 1 | 1 |
| Partner | 45(20.7) | 172(7.1) | 3.56 (2.43-5.21) | 2.98 (1.86-4.76) |
| Other than a partner | 8(3.7) | 22(0.9) | 4.95 (1.99-11.92) | 4.59 (1.74-12.09) |
aLogistic regression models used to calculate adjusted odds ratios included 2613 observations. |
The association between physical abuse during pregnancy and preterm birth was similar after excluding medically induced preterm cases from analysis (adj.OR = 2.87; 95% CI, 1.69-4.87). Women who were physically abused before but not during pregnancy (n = 83) also showed an increased risk of preterm delivery when compared with never abused women, although the association was weaker (adjOR = 2.1; 95% CI, 1.00-4.31).
Abused women reported more frequently genital hemorrhage during pregnancy (16% vs 5.4%; P < .005) and the association was significant for every pregnancy trimester. Gestational hypertension was less prevalent among abused women (1.6% vs 5.4%; P < .005). No statistically significant differences were found for preeclampsia or other severe pregnancy hypertensive disorders (2.3% vs 0.9%), diabetes (3.5% vs 5.3%) or urinary tract infections (16.0% vs 18.6%). Victims of physical abuse during pregnancy reported more frequently the occurrence of unspecified pregnancy diseases (36.2% vs 16.3%) and they also delivered more frequently a small-for-gestational-age newborn infant (36.5% vs 10.3%; P < .005).
Comment
This report resulted from a cross-sectional survey conducted in a Portuguese level III hospital attending mainly urban population and delivering about 3000 newborn infants per year. Unlike other studies that have found a relationship between violence against pregnant women and preterm delivery, our study has not been conducted in a generally indigent or low-income population. From a search in Medline, no previous report on the prevalence of physical abuse during pregnancy was found for Portuguese women. We estimated that about 10% of pregnant women are physically abused during pregnancy, a rate that is within the range of prevalence estimates in other countries.4
This study disclosed a significant association between physical abuse during pregnancy and preterm birth, even after controlling for many sociodemographic and behavioral factors. The crude association between physical abuse during pregnancy and preterm was only slightly attenuated after adjustment for the potential confounders. This means that low socioeconomic status, behavioral factors such as smoking, alcohol or drug use, and inadequate antenatal care use do not largely explain the association between physical abuse during pregnancy and preterm delivery. We also have not found any interaction between physical abuse and social class (assessed through maternal education or family income), marital status, maternal age, or parity. Therefore, we only entered these factors in the multivariable models as potential confounders.
The mechanisms linking violence with preterm birth are yet unknown. Some authors found that women admitted during pregnancy after sustaining an assault, experienced higher rates of preterm delivery, whether they delivered at the assault hospitalization or were discharged after the assault.16 According to Petersen et al,17 the relationship between violence during pregnancy and adverse pregnancy outcomes must take into account two potential etiologic pathways, physical trauma, and stress. Severe direct abdominal trauma may cause adverse pregnancy outcome, but minor abdominal trauma seems less likely to be associated with perinatal results. Women abused during pregnancy are more likely than nonabused women to experience higher levels of other types of stressful life events18 and more likely to have depression and anxiety.19 There is increasing evidence that psychosocial stress may be associated with preterm delivery via both behavioral and neuroendocrine pathways. Women who had been subjected to violence are more likely to smoke, drink alcohol, and use illegal drugs, and these unhealthy behaviors may be associated with preterm delivery. On the other hand, emotional stress may activate the neuroendocrine axis, causing the release of catecholamine and other vasoconstrictors that lead to fetal hypoxia or fetal growth restriction and predispose to induced preterm or provoke the release of prostaglandins contributing to preterm labor. Nevertheless, according to McLean et al,20 besides stressors, there are effect modifiers like personal dispositions (coping behaviors and responses, perceived control of situations) and social support and networks, and these factors may interfere with the effect of violence on pregnancy outcome. Other plausible explanations that have not been adequately investigated are high physical demanding activity, sexually transmitted diseases, and nutritional deprivation.
We have found that victims of violence during pregnancy reported vaginal bleeding in every pregnancy trimester more frequently than nonabused women. These findings favor the hypothesis that choriodecidual hemorrhage and abruption placenta are mechanisms through which physical abuse during pregnancy may increase the risk of preterm delivery. It is commonly accepted that intentionally inflicted abdominal trauma may lead directly to placental dysfunction. However, vaginal bleeding may be a sign of threatened abortion or preterm labor resulting from another causal pathway. Moreover, in our study, too few women referred direct trauma to the abdominal area during pregnancy.
It has been supposed that the association between physical abuse and preterm birth could partially be explained by an increased frequency of pregnancy complications, such as preeclampsia and fetal growth restriction among abused pregnant women. In our study, we have found a significantly higher prevalence of small-for-gestational-age, but not preeclampsia among abused women. A lower prevalence of gestational hypertension among abused women may be explained by less antenatal care use and gestational hypertension underdiagnosis.
We also have found an association between physical abuse and preterm delivery when abuse exposure only occurred before pregnancy. This finding supports the presence of mechanisms other than direct maternal physical trauma. The experience of violence before pregnancy may result in posttraumatic stress, which may lead to preterm delivery.
In this study, intimate partners represented the largest single category of perpetrators of violence against pregnant women, but preterm risk was also increased among women battered by others than intimate partners, usually the father. We may hypothesize that violence inflicted by relatives other than an intimate partner may happen in a context of high dependency and psychosocial stress.
We used the Abuse Assessment Screen because its reliability and validity have already been established and it is easy to administrate. This is also the most commonly used research tool to measure violence in hospital samples. Bias in exposure status ascertainment cannot be excluded in our study. Although the use of structured questionnaires by trained staff significantly improves detection rate,21 it is still possible that domestic violence has been underidentified given the reluctance of women to report it. However, if a nondifferential bias had occurred, the true association would even be higher than the reported one. Although, in data analysis, we have made efforts to control for all confounding factors, the possibility of residual confounding still exists, but it is unlikely to be relevant given the magnitude of the observed effect even after adjustment.
The main strengths of this study are its large sample size, the high participation rate; the inclusion of women from virtually all social and economic strata and a broad range of pregnancy and delivery risk profiles. Assessment of violence exposure took place with guaranteed privacy, through face-to-face interviews that used highly trained personnel. Evaluation and control for many potential confounding factors, including smoking, alcohol intake, and illicit drug abuse; with accurate information on gestational age and pregnancy complications were essential to data quality analysis. Although the study setting was a public maternity of a general university hospital, the external validity of the results can be inferred because more than 90% of deliveries occur in public hospitals.
In Portugal, like in many other developed countries, pregnant women are not routinely screened for domestic violence. Besides uncertainties in the causal pathways, it seems to be consensual that health professionals should recognize physical abuse as a risk factor for preterm delivery and inquire pregnant women about it.
References
- Research agenda for preterm birth: recommendations from the March of Dimes. Am J Obstet Gynecol. 2005;193:626–635
- . Perspectives in the prevention of premature birth. Eur J Obstet Gynecol Reprod Biol. 2004;117(Suppl):S2–S5
- . Psychosocial stress and neuroendocrine mechanisms in preterm delivery. Am J Obstet Gynecol. 2005;192:S30–S35
- . Prevalence of violence against pregnant women. JAMA. 1996;275:1915–1920
- . Violence and reproductive health: current knowledge and future research directions. Matern Child Health J. 2000;4:79–84
- . Intimate partner violence and birth outcomes: a systematic review. Int J Fertil Womens Med. 2004;49:159–164
- . Effects of domestic violence on preterm birth and low birth weight. Acta Obstet Gynecol Scand. 2004;83:455–460
- . Abuse: a risk factor for low birth weight? (A systematic review and meta-analysis). CMAJ. 2001;164:1567–1725
- . Effects of abuse on maternal complications and birth weight in adult and adolescent women. Obstet Gynecol. 1998;92:530–534
- Physical and nonphysical partner abuse and other risk factors for low birth weight among full term and preterm babies: a multiethnic case control study. Am J Epidemiol. 1999;150:714–726
- . Severity, manifestations, and consequences of violence among pregnant adolescents. J Adolesc Health. 2001;28:55–61
- . Domestic violence: effect on pregnancy outcome. J Am Osteopath Assoc. 1999;99:254–256
- . Physical violence during pregnancy: maternal complications and birth outcomes. Obstet Gynecol. 1999;93(Pt 1):661–666
- . Intimate partner violence and adverse pregnancy outcomes: a population-based study. Am J Obstet Gynecol. 2003;188:1341–1347
- . Assessing for abuse during pregnancy: severity and frequency of injuries and associated entry into prenatal care. JAMA. 1992;267:3176–3178
- . Maternal and neonatal outcomes of assaults during pregnancy. Obstet Gynecol. 2005;105:357–363
- Violence and adverse pregnancy outcomes: a review of the literature and directions for future research. Am J Prev Med. 1997;13:366–373
- . Stressful life events and physical abuse among pregnant women in North Carolina. Matern Child Health J. 2001;5:145–152
- . Domestic violence: prevalence in pregnant women and associations with physical and psychological health. Eur J Obstet Gynecol Reprod Biol. 2004;113:6–11
- . Psychosocial measurement: implications for the study of preterm delivery in black women. Am J Prev Med. 1993;98:39–81
- . Battering in pregnancy: an assessment of two screening methods. Obstet Gynecol. 1995;85:321–325
Cite this article as: Rodrigues T, Rocha L, Barros H. Physical abuse during pregnancy and preterm delivery. Am J Obstet Gynecol 2008;198:171.e1-171.e6.
PII: S0002-9378(07)00629-1
doi:10.1016/j.ajog.2007.05.015
© 2008 Mosby, Inc. All rights reserved.
Refers to article:
- Physical abuse and preterm delivery: Rodrigues et al
Volume 198, Issue 2 , Pages 171.e1-171.e6, February 2008
