Menu
Menu
Menu
Menu
N/A
Background: There is growing evidence of the negative impact of alcohol on morbidity and mortality of individuals living with HIV but limited evidence of in utero effects of HIV and alcohol on exposure on infants. Methods: We conducted a population-based birth cohort study (N=667 mother-infant dyads) in South Africa to investigate whether maternal alcohol use and HIV affected gestational outcomes. Descriptive data analysis was conducted for all variables using frequency distributions, measures of central tendency, and estimates of variance. Hierarchical multiple regression was conducted to determine whether maternal alcohol use, maternal HIV status and other risk factors (socioeconomic status, smoking, depression) predicted infant outcomes. Results: Our results showed severity of recent alcohol use and lifetime alcohol use predicted low birth weight. Similarly lifetime alcohol use predicted shorter infant length, smaller head length, smaller head circumference, and early gestational age. However, HIV status was not a significant predictor of gestational outcomes. Conclusions: The unexpected finding that maternal HIV status did not predict any of the gestational outcomes may be due to high rates of ART usage among HIV-infected mothers. The potentially negative effects of HIV on gestational outcomes may have been attenuated by improved maternal health due to high coverage of antiretroviral treatment in South Africa. Interventions are needed to reduce alcohol consumption among pregnant mothers and to support healthy growth and psychosocial development of infants.
Data was collected from an investigation of child health, called the Drakenstein Child Health Study (DCHS). The DCHS consisted of a multidisciplinary birth cohort study following mother–child dyads from the prenatal period through to five years of age [29, 30]. The aim of DCHS is to investigate the epidemiology, etiology, and risk factors of childhood respiratory disease and the determinants of child health in a low socioeconomic area of South Africa [30]. The study investigates the role and interaction of possible risk factors in 7 areas (environmental, infectious, nutritional, genetic, psychosocial, maternal and immunological risk factors) that may impact child health. Pregnant women were recruited from two primary care clinics in the Drakenstein sub-district near Cape Town, South Africa between March 2012 and September 2014. Data collection occurred at these two clinics (maternal data) and at a central hospital (newborn outcomes). Participants were enrolled in the study at 20–28 weeks gestation upon presenting for antenatal booking. The mothers were followed throughout pregnancy and mother–child dyads are followed until five years after birth. Data collection for the DCHS remains ongoing. Throughout the study, participants receive regular medical care as per the South Africa national program and expanded program for immunization (EPI) schedule. The DCHS is located in the Drakenstein area, a peri-urban area, 60 km outside Cape Town, South Africa with a population of approximately 200,000. The community is stable, with low levels of immigration or emigration. The local economy is based around commercial agriculture and light industry. More than 90% of the population access health care in the public sector including antenatal and child health services [31]. This area has a well-established, free primary health care system, with high coverage for childhood immunizations. The public health system is comprised of 23 primary health care clinics and one centralized hospital where all births and all hospital-based pediatric care, including admissions occur [30]. To be eligible for the study, participants were: (a) over age 18, (b) pregnant, (c) 20–28 weeks gestation, and (d) receiving antenatal care at a participating study site. The two primary clinics were TC Newman clinic (serving a predominantly “Coloured” or mixed ancestry community) and Mbekweni clinic (serving a predominantly black African community). Exclusion criteria were minimal in order to maximize generalizability, and focused primarily on those individuals who did not live in the region (and could not readily complete follow-up assessments) or those who were intending to move out of the district within the following 2 years. The cohort comprised mothers presenting in an unfiltered manner to the antenatal services and were not selected for any clinical risk factors or exposures. The study was approved by the Faculty of Health Sciences, Human Research Ethics Committee, University of Cape Town (401/2009), and by the Western Cape Provincial Health Research committee. Written informed parental consent was obtained from all participants at the time of enrolment. Demographic data was collected regarding participant age, education, employment, and socioeconomic status (SES). For our purposes, a composite SES score was developed in order to categorize participants into quartiles based on their relative SES. This composite score is calculated based on current employment status and standardized scores of educational attainment, household income, and a composite asset index based on access to household resources, amenities, and market access. Participants were categorized as being of low, low-moderate, moderate-high, or high SES based on their relative composite SES score. Maternal HIV status was determined using an HIV ELISA test, which all mothers in South Africa receive during pregnancy due to the high prevalence of HIV. Those who tested positive for HIV self-reported whether she was prescribed antiretroviral therapy. Mothers who were newly diagnosed received a second confirmatory HIV ELISA test and were referred for counseling and evaluation for treatment initiation as per standard management guidelines. All babies were born at a central hospital that serves the communities of both clinics. Trained clinical staff recorded newborn birth weight (kilograms) at the time of delivery, using a digital scale with a precision level of 10 g. Gestational age (weeks), head circumference and length in centimeters were also recorded at birth. We defined low birth weight and very low birth weight according to World Health Organization’s (WHO’s) parameters for weight for age, with low birth weight defined as less than 2.5 kilograms and very low birth weight as less than 1.5 kg [32]. Percentile for infant birth weight was then calculated using the formula proposed by Mikolajcyzk et al. [33]. This formula enabled us to calculate the birth weight percentile based on gestational week of each infant. We used full-term, low-risk infants (N = 100) as the reference group for anchoring our percentile estimates. This method for estimating fetal weight percentiles has a better ability to predict adverse perinatal outcomes than other unadjusted and fully individualized formulas [33]. Risk of substance use was assessed using the alcohol, smoking and substance involvement screening test (ASSIST) at the second antenatal visit (28–32 weeks gestation). This tool was developed by the WHO to detect psychoactive substance use and related disorders in primary care settings. It has shown good reliability, feasibility and validity in international, multisite studies [34, 35], and is useful for identifying risk of substance dependence in poly-substance abusers with varying degrees of psychopathology. Seven items are included to assess alcohol and other drug use across 10 categories (i.e., tobacco products; alcoholic beverages; cannabis; cocaine; amphetamine-type stimulants; inhalants, sedatives or sleeping pills; hallucinogens; opioids; and a general category entitled “other”, in which the participant is required to specify the substance used). ASSIST scores from 0 to 10 for alcohol and 0–3 for illicit drugs indicating low-risk, scores from 11 to 26 for alcohol and 4–26 for illicit drugs indicate moderate-risk and scores above 26 indicate high risk of severe problems, with the likelihood of substance dependence. The higher the score, the greater substance-related risk. The ASSIST has good reliability scores and has been validated in numerous populations [36], as well as primary care settings [37]. For our purposes, only data for alcohol were extracted from the ASSIST, as reports of other substance use were negligible with the exception of tobacco where a biochemical measure of exposure was used as detailed below. Lifetime alcohol use was a dichotomous variable coded as 0, 1. Past 3 month alcohol use severity was calculated as a continuous variable using standard scoring techniques to calculate a sum score based on past 3-month alcohol use frequency, craving, consequences, and other symptoms of alcohol use disorders as detailed above [38]. Biochemical verification of smoking status was determined through cotinine urine analysis collected on the same day as the baseline interview. Cotinine is a nicotine metabolite that is measurable among people who smoke, use smokeless tobacco, use nicotine products, or are exposed to environmental tobacco smoke. An antenatal urine cotinine cut-off value of 500 was used to distinguish active smokers from non-smokers [39]. The Beck Depression Inventory-II (BDI-II) [40] was used to assess depressive symptomatology. This is a widely used self-report measure of the severity of depressive symptoms with strong internal consistency and concurrent validity in clinical and non-clinical samples [41, 42]. Each item is assessed on a severity scale ranging from 0, “absence of symptoms”, to 3 “severe, often with functional impairment”. A total score was obtained by summing individual item responses, with higher scores indicative of more severe depressive symptoms. Descriptive data analysis was conducted for all variables using frequency distributions, measures of central tendency, and estimates of variance. Hierarchical multiple regression was conducted to determine whether maternal HIV status, alcohol use, and other risk factors (socioeconomic status or SES, smoking, depression) predicted infant outcomes. Infant outcomes included birth weight percentile, length, head circumference, and gestational age. All outcomes were entered as continuous variables in regression analyses. The analyses were conducted in a step-wise fashion to examine the unique contributions for each block of predictor variables. The four steps/blocks of variables in the regression analyses included (in order): (1) HIV status, (2) SES score, BDI-II score, and smoking status, (3) alcohol use, and (4) interaction variables (HIV by alcohol use and HIV by smoking status). We examined separate models using the predictors past 3-month alcohol use severity and lifetime alcohol. All regression models were examined for violations of normality, linearity, homoscedasticity, redundancy, and multicollinearity. Data performed excellently on all criteria except for length of gestation. Length of gestation was non-normal (skew = −2.04, kurtosis = 7.38) and the log of this variable was used in the analysis. Statistical significance was set at the 0.05 level.
N/A
