|Home | About | Journals | Submit | Contact Us | Français|
Prenatal cocaine exposure has been linked to intrauterine growth retardation and poor birth outcomes; little is known about the effects on longer-term medical outcomes, such as overweight status and hypertension in childhood. Our objective was to examine the association between prenatal cocaine exposure and body mass index and blood pressure at 9 years of age among children followed prospectively in a multi-site longitudinal study evaluating the impact of maternal lifestyle during pregnancy on childhood outcome.
This analysis includes 880 children (277 cocaine exposed and 603 with no cocaine exposure) with blood pressure, height, and weight measurements at 9 years of age. Regression analyses were conducted to explore the relationship between prenatal cocaine exposure and body mass index and blood pressure at 9 years of age after controlling for demographics, other drug exposure, birth weight, maternal weight, infant postnatal weight gain, and childhood television viewing, exercise and dietary habits at 9 years. Path analyses were used to further explore these relationships.
At 9 years of age, 15% of the children were pre-hypertensive and 19% were hypertensive; 16% were at risk for overweight status and 21% were overweight. A small percentage of women were exposed to high levels of prenatal cocaine throughout pregnancy. Among children born to these women, a higher body mass index was noted. Path analysis suggested that high cocaine exposure has an indirect effect on systolic and diastolic blood pressure that is mediated through its effect on body mass index.
High levels of in-utero cocaine exposure are a marker for elevated body mass index and blood pressure among children born full term.
Hypertension has for several decades been recognized as a risk factor for cardiovascular disease. Children with higher blood pressure are more likely to become adults with hypertension . An unfavorable environment during fetal life or insults during fetal life induce lifetime effects on the subsequent development of major disease processes . Cocaine is a potential vasoconstrictor; therefore prenatal exposure to cocaine may be associated with childhood hypertension. Two small studies [3,4] have shown conflicting associations between cocaine exposure during pregnancy and elevated blood pressure in infants, and the relationship of prenatal cocaine exposure and childhood hypertension has not been explored to date (3–4).
Another childhood morbidity that is effecting progressively younger children is obesity and overweight status, with the proportion of children above the 85th percentile of body mass index (BMI) reaching 37% in the most recent National Health and Nutrition Examination . Insults during fetal life that impact growth include substance use; smoking during pregnancy is associated with increased BMI during childhood and adolescence, while alcohol and cocaine use during pregnancy are associated with lower weight for height in childhood [6–10].
Risk factors for overweight status and hypertension include preterm birth, restricted growth during fetal life, accelerated growth during early childhood and the child’s current body mass index (BMI) [10–15]. Other risk factors that influence hypertension and obesity are Black or Hispanic race, lower socioeconomic status and high caloric diet [5, 16–22]. An increased maternal BMI status, female gender and lack of physical activity increase risk for obesity [5, 23–25].
The multi-site longitudinal study evaluating impact of maternal lifestyle on childhood outcomes (Maternal Lifestyle Study, MLS) offered an opportunity to explore the relationship between prenatal cocaine exposure and childhood obesity and hypertension at 9 years of age while controlling for risk factors associated with these conditions. We hypothesized that children born to mothers who used cocaine during pregnancy would have a higher blood pressure at 9 years of age then children born to mothers who did not use cocaine during pregnancy.
The MLS study is performed at 4 sites (Brown University, University of Miami, University of Tennessee at Memphis and Wayne State University) in the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network with additional support from the National Institute on Drug Abuse. The study was approved by the Institutional Review Board at each of the participating centers and informed consent was obtained prior to study participation.
A total of 1,388 children were enrolled in the MLS study at 1 month of age, 658 cocaine exposed and 730 infants who were not cocaine exposed and were matched by race, gender and gestational age at each clinical site. Of these children, 1,259 (91%) had data on the level of cocaine exposure by interview of mother and analysis of the infant’s meconium for cocaine metabolites using gas chromatography mass spectroscopy. Infants in the non exposed group had both a negative report of cocaine use and no metabolites in the infants’ meconium sample. The biological mother was not available for interview in the remaining 9% of cases. Among these 1,259 participants, 915 (73%) were present for the medical assessment at 9 years of age, with 23% loss to follow up of this extremely high risk population. Of these 915 participants, 880 (96%) had complete data on blood pressure, height, and weight at 9 years of age.
The demographic variables included child’s gender, mother’s race (black, white, or other), mother’s education (highest grade completed), mother’s weight at the time of delivery and clinical site.
These included birth weight, whether the child was born at term (> 36 weeks gestation) or pre-term, and whether he or she was born small for gestational age (SGA, below the 10th percentile for gestational age) or appropriate for gestational age (AGA). In addition, the change in the child’s weight from birth to four months corrected age was computed as weight gain in grams per month. A detailed history of maternal substance use (cocaine, opiate, alcohol, tobacco and marijuana) during the 3 month period before pregnancy and the three trimesters of pregnancy was obtained at the 1 month visit using the Maternal Inventory of Substance Use . In order to evaluate a possible dose response effect for prenatal cocaine exposure, a three-level cocaine exposure variable was used in the MLS: no exposure, some exposure (1–2 days per week or less), and high exposure (3 to 6 days per week or more). In addition to cocaine exposure, prenatal exposures to other drugs (alcohol, tobacco, opiates, and marijuana) were also included with each of these exposures coded as exposed vs. not exposed.
At the 9 year data collection, children were classified by caretaker interview as exercising for half an hour or more 1–2 times a week, 3–4 times or > 4 times a week. Television (TV) viewing on school days was categorized as: (1) does not watch TV on school days, (2) less than 2 hours per school day, or (3) 2 or more hours per school day. In addition, average caloric intake was calculated based on a three day food diary prior to the clinic visit at 9 years.
Body mass index (BMI) was determined separately for boys and girls from height measured on a stadiometer and weight recorded in Kg and plotted on the CDC growth curves designed for use with children 2 to 20 years of age . Risk for overweight was defined as BMI between the 85th to 95th percentile, while overweight was defined as > 95th percentile. Systolic and diastolic blood pressures (BP) were measured in a sitting position as an average of 3 measurements of right arm BP using the Dinamap portable monitor with cuff size appropriate for the upper arm. Hypertension was defined as the average of systolic blood pressure (SBP) and/or diastolic blood pressure (DBP) that is ≥95th percentile for gender, age and height, while pre-hypertension was defined as average of SBP and/or DBP between >90th and 95th percentile. Children with blood pressure levels in the hypertensive range were referred to their primary care physician for further evaluation.
Linear regression models were conducted with BMI as the outcome variable and the level of prenatal cocaine exposure as a predictor variable, controlling for the variables mentioned above. Similar regression models were conducted for systolic and diastolic blood pressure at 9 years. Because of the potential for cocaine effects to vary by whether the child was born at term vs. pre-term, a model that included an interaction term for cocaine by term/pre-term was tested initially. All of the regression models were conducted using the SAS 9.0 software program.
Following the regression models, path analyses were conducted using the M-plus software program  to explore the relationship between BMI and blood pressure. To test for the model fit, values for the comparative fit index (CFI), Tucker-Lewis Index (TLI) and the root mean square error of approximation (RMSEA) were calculated. If the model fits well, the CFI and TLI should have values of 0.90 or greater and the RMSEA should be 0.05 or less . Established approaches for dealing with missing data were incorporated into the analyses. For the regression models, multiple imputations  were used as implemented in SAS PROC MI and MIANALYZE. For the path analyses, the missing data approach in MPlus was used as this method allows for the inclusion of all observations and is based on full information maximum likelihood estimation. Models were examined both with and while removing participants with any missing data.
The mean BMI for the 880 children was 18.8 with a standard deviation (SD) of 4.5. Males had lower mean BMI (mean ± SD 18.6 ± 4.1; range of 11.9 to 36.0) than females (19.1 ± 4.9; range of 12.3 to 44.9) (P = 0.06). Children who were small for gestational age had a lower BMI (17.7 ± 3.5) than those born appropriate for gestational age (19.2 ± 4.7, P < 0.001) and those who were born pre-term had a lower BMI (18.5 ± 4.4) than those born full term (19.1 ± 4.5, P = 0.08). Cocaine exposure groups did not differ in BMI; with no exposure BMI was 18.7 ± 4.4, some exposure 18.8 ± 4.3 and high exposure 19.6 ± 5.4, respectively.
The frequencies of children in the three cocaine exposure levels is shown in Table 1. Overall 69% of children were born to women who used no cocaine in pregnancy, 21% with some use and 10% with high use. There were significant differences in maternal education, weight at delivery and alcohol and poly drug use between the three exposure groups. The frequencies of BMI categorized as normal, at risk for overweight and overweight are shown in Table 2. Overall, 16% of the children were at risk for overweight status at 9 years of age and 21% were overweight. There were significant differences among the BMI categories by birth weight, small for gestational age status, postnatal weight gain, race, and, mother’s weight at delivery.
The frequency and percentage of children in the sample classified as normal, pre-hypertensive, and hypertensive are presented in Table 3. Fifteen percent of the sample was classified as pre-hypertensive and 19 % were classified as having hypertension. The blood pressure categories varied significantly by birth weight.
The relationship between prenatal cocaine exposure and 9-year BMI, after adjusting for each of the control variables outlined above was explored. The interaction between term (vs. pre-term) birth and cocaine exposure was significant; F (2,677) = 5.36 (P = .005), suggesting that cocaine exposure has a differential impact on BMI for term and pre-term children. Therefore, separate models for these two groups were examined.
As shown in Table 4, high cocaine exposure was significantly associated with higher BMI among children born at term (P = 0.019). Higher BMI among children born at term was also associated with female gender, higher birth weight, higher postnatal weight gain for first four months of age in the child, higher maternal weight at delivery and watching 2 or more hours of TV on school days and consuming more calories at 9 years of age.
The corresponding regression model for pre-term children is shown in Table 5. Cocaine exposure was not significantly associated with BMI for either high (P = 0.69) or some (P = 0.90) exposure. Several of the other predictors of high BMI were similar to those for term children: female gender, higher birth weight, higher postnatal weight gain for child, and higher maternal weight at delivery. In addition, higher maternal education and not exercising regularly at 9 years of age were associated with higher BMI among children born at pre-term.
Regression analyses were then conducted to test whether cocaine exposure has a direct impact on blood pressure. For systolic and diastolic blood pressure, the term by cocaine interactions were not significant: Systolic: F (2,755) = 1.48 (P = .23); Diastolic: F(2,755) = 2.13 (P = .12). Hence, models that included both term and pre-term children were explored. As shown in Table 6 and Table 7, cocaine exposure was not significant for either blood pressure measure. Higher systolic blood pressure was associated with greater postnatal weight gain and higher maternal weight at delivery while lower systolic blood pressure was associated with prenatal alcohol exposure. Similarly, lower diastolic blood pressure was associated with prenatal alcohol exposure.
Based on the regression results, it appears that cocaine exposure does not have a direct impact on blood pressure after controlling for other possible predictors. Therefore, analyses exploring whether cocaine has an indirect effect on blood pressure through its impact on BMI was examined (Figure 1). The other variables (i.e., demographic, pre/neonatal, and 9-year variables) are included as control variables and have direct paths to BMI. To control for possible variation in measurement error by site, clinical site has a direct path to both BMI and blood pressure. Path analyses were conducted to test this conceptual model for systolic and diastolic blood pressure. Because the regression results indicated no significant relationship between cocaine exposure and BMI for the pre-term children, the path analyses included only the term children. The path analysis model for systolic BP is shown in Figure 2. This model fit very well (CFI = .97, TLI = .92, RMSEA = .03), supporting the hypothesis of an indirect cocaine effect on blood pressure. High cocaine exposure had a significantly positive relationship with BMI (path coefficient = 1.50, P = 0.02) while some cocaine exposure was not significant (path coefficient = 0.65, P = 0.23). There is a strong positive relationship between BMI and systolic blood pressure (path coefficient = 0.78; P < 0.001).
Similar results were noted for diastolic blood pressure (Figure 3). The model fit well (CFI = .99, TLI = .97, RMSEA = .02). High cocaine exposure was positively associated with BMI (path coefficient = 1.50, P = 0.02) and the relationship for some cocaine exposure was not significant (path coefficient = .65; p = 0.23). BMI and diastolic blood pressure were positively associated (path coefficient = 0.14; p = 0.02).
In this large, multi-site prospective longitudinal study, a high level of maternal cocaine use during pregnancy (≥ 3 days a week use) was associated with a higher BMI at 9 years among children born at term, and through this effect on higher BMI, cocaine exposure was also indirectly associated with higher blood pressure. High levels of in-utero cocaine exposure can thus be considered as a marker for elevated body mass index and blood pressure in children born full term. It should be noted that very few women in this study continued to use cocaine throughout pregnancy.
Among six to 11 year old children in the National Health and Nutrition Examination Survey, the prevalence of overweight status was 18.8 % . In our study we found that 21% of our nine year old children were overweight. This high rate may reflect the predominantly inner city, low socioeconomic status children in our study. Our findings that a higher BMI was associated with higher maternal weight at delivery, higher neonatal birth weight and accelerated weight gain, have been documented by other investigators; a trend for higher BMI with prolonged television viewing was also noted in our study [5, 14, 15, 18, 24, 25].
The association of prenatal substance use and BMI in childhood is being recognized with current longitudinal studies. Alcohol use during pregnancy is associated with a negative impact on height, weight and head circumference at eight years of age and 10 years of age [8,9]. Smoking during pregnancy has been associated with an increased BMI during early childhood, later childhood and adolescence [6,7]. Cocaine exposure is associated with lower weight for height at six years of age or that the negative effect of cocaine at birth did not persist beyond eight years (8,10). In the current study, we noted that among term infants only, high cocaine use during pregnancy was associated with increased BMI at nine years of age. We speculate that high cocaine use throughout pregnancy in this study may be associated with under nutrition which has been linked with later obesity.
Increased blood pressure levels during childhood strongly predict hypertension in young adulthood and the earlier BP values exceeded criterion, the higher the risk of hypertension later in life . In our study, at the age of nine years, 19% of children were found to be hypertensive. This rate is higher than the rate reported in the National Health and Nutrition Examination Survey (NHANES) . We speculate that this high rate reflects an atypical study group that is predominantly inner city and African American and has a low socio-economic status.
In the present study we evaluated blood pressure separately in term infants and preterm infants. We found that among both term and preterm infants, increased maternal weight at delivery was associated with an increase in systolic blood pressure. Although maternal weight at delivery has not been reported to be associated with childhood hypertension, higher blood pressure in childhood has been reported among the offspring of diabetic mothers who have an abnormal maternal metabolism as well as offspring of mothers with hypertension at delivery [32,33]. The relationship between rapid postnatal weight gain and elevated BP that we noted in this study has also been seen by other investigators .
The relationship between substance use during pregnancy and childhood hypertension is not clear. The negative association between maternal alcohol use during pregnancy and blood pressure at nine years that we found in our study is a new finding. Prenatal smoking does result in elevated childhood blood pressure . In our study when smoking was evaluated as a covariant with other substance use, we did not find a association between smoking and childhood blood pressure. We have previously examined the association with cocaine, opiate, marijuana, tobacco and alcohol use during pregnancy and hypertension at six years of age among children born at term and did not find any association . At six years of age, the children in our study had a very low frequency of risk for overweight and overweight status. In the present study the increase in overweight status at nine years may have mediated the relationship, however small, between prenatal cocaine exposure and elevated blood pressure.
The strengths of the study are the large sample (880 children) followed prospectively with information on prenatal substance use during pregnancy, maternal weight at delivery, and childhood longitudinal growth along with diet and exercise history at nine years of age. A further strength of this study is that children born at both full term and preterm birth were evaluated.
The drawbacks of this study are that 80% of the children evaluated were Black; hence the impact of race on blood pressure could not be adequately assessed. All three measurements of blood pressure were made serially at one visit; not at the beginning and end of the visit to reduce “white coat hypertension”. We did not have either maternal blood pressure measurements at delivery or at the child’s nine year visit, nor a detailed family history of heart disease, or serum markers of cardiovascular risk.
In summary, in a longitudinal study evaluating the relationship of maternal lifestyle during pregnancy on childhood medical outcomes at nine years of age, we found a high rate of overweight status and hypertension. A very small percentage of women were exposed to cocaine throughout their pregnancy. Among the children born at term to these women, a higher risk for elevated blood pressure was associated with increased BMI at nine years of age. Correlations, not causality, are implied in our study. Cocaine exposed children continue to live in a high risk environment and may be at risk for elevated BMI in later childhood.
The National Institutes of Health, the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), the National Institute on Drug Abuse (NIDA), the Administration on Children, Youth, and Families, and the Center for Substance Abuse and Treatment provided grant support for recruiting subjects into the Maternal Lifestyle Study in 1993–1995. NIDA and NICHD provided funding to conduct follow-up examinations in three phases: at 1, 4, 8, 10, 12, 18, 24, and 36 months corrected age (Phase I); at 3½, 4, 4½, 5, 5½, 6, and 7 years of age (Phase II); and at 8, 9, 10, and 11 years of age (Phase III). The funding agencies provided overall oversight of study conduct, but all data analyses and interpretation were completed independent of the funding agencies. We are indebted to our medical and nursing colleagues and the infants and their parents who agreed to take part in this study.
Data collected at participating sites of the NICHD Neonatal Research Network (NRN) were transmitted to RTI International, the data coordinating center (DCC) for the network, which stored, managed, and analyzed the data for this study. On behalf of the NRN, Drs. Abhik Das (DCC Principal Investigator) and Sylvia Tan (DCC Statistician) had full access to all the data in the study and take responsibility for the integrity of the data and accuracy of the data analysis.
The following investigators, in addition to those listed as authors, participated in this study:
Steering Committee Chair: Barry M. Lester, PhD, Brown University.
Brown University Warren Alpert Medical School Women & Infants Hospital of Rhode Island (U10 HD27904, N01 HD23159) – Barry M. Lester, PhD, Cynthia Miller-Loncar, PhD; Linda L. LaGasse, PhD; Jean Twomey, PhD.
Eunice Kennedy Shriver National Institute of Child Health and Human Development – Rosemary D. Higgins, MD.
National Institute on Drug Abuse – Vincent L. Smeriglio, PhD; Nicolette Borek, PhD.
RTI International (U10 HD36790) – W. Kenneth Poole, PhD; Abhik Das, PhD; Jane Hammond, PhD; Debra Fleischmann, BS.
University of Miami Holtz Children's Hospital (GCRC M01 RR16587, U10 HD21397) – Charles R. Bauer, MD; Ann L. Graziotti, MSN, ARNP; Rafael Guzman, MSW; Carmel Azemar, MSW.
University of Tennessee (U10 HD42638) – Henrietta S. Bada, MD; Toni Whitaker, MD; Charlotte Bursi, MSSW; Pamela Lenoue, RN.
Wayne State University Hutzel Women’s Hospital and Children’s Hospital of Michigan (U10 HD21385) – Seetha Shankaran, MD; Eunice Woldt, RN MSN; Jay Ann Nelson, BSN.
Grant Numbers: U10HD21385a, U10HD36790b, U10HD21397c, U10 HD27904d, U10HD42638e
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.