Facts, figures, & fallacies

Failed or misleading adjustment for confounding

Effects of prenatal exposure to cocaine on the reactivity and regulation of the motor system of 825 four-month-old infants enrolled in the Maternal Lifestyle Study were examined. Videotaped assessments of 338 cocaine-exposed (CE) infants and 487 non-exposed comparison infants were coded by examiners masked to exposure status. Exposure status was determined by meconium assay and maternal self-report of prenatal cocaine use. Infants were presented with a series of 17 visual, auditory and tactile stimuli for 30-s each. Intensity and latency of limb movement responses on a subset of items were analyzed to test the following hypotheses: CE infants are more active in general; CE infants exhibit increased movement levels for a larger proportion of time in response to stimulation; the motor systems of CE infants are more reactive to stimulation (e.g., shorter latencies to respond); and CE infants are poorer regulators of the motor system.

CE infants were not more active in general and data do not indicate a more highly reactive motor system. However, CE infants exhibited increased movement levels for a larger proportion of time in response to stimulation. Additional analysis of movement exhibited during three tactile items found increased movement lability in CE infants and different patterns of responding, suggesting that the effects of prenatal cocaine exposure on the motor system may vary by context. Covariate effects for tobacco, alcohol, and marijuana are also reported.

The literature concerning the relationship between polychlorinated biphenyl (PCB) exposure and Intelligence (IQ) is not entirely consistent. Two studies showed inverse associations between PCBs and IQ in cohorts of children whose mothers consumed Great Lakes fish contaminated with PCBs and other organochlorines (Jacobson & Jacobson 1996; Stewart et al. 2008). Another study from the general US population, where women were exposed to background levels of PCBs, showed no association between PCBs and IQ (Gray et al. 2005). The current report examines two potential sources of inconsistency across studies: 1) confounding with non-PCB organochlorines [Hexachlorobenzene (HCB), dichlorodiphenyldichloroethylene (DDE), and Mirex]; and 2) the presence of negative confounding (i.e., suppressor variables). The former could confound PCBs and lead to spurious associations (Type I errors), while the latter could suppress PCB associations and obscure true associations (Type II errors). These issues were explored through the analysis of associations between placental levels of organochlorines and IQ in children at 9 and 11 years of age in the Oswego study. Neither DDE nor mirex was related to lower IQ at either age; PCBs predicted lower IQ at both ages; and hexachlorobenzene (HCB) appeared as a significant predictor of IQ at the 11-year assessment. However, analysis of the IQ data set as a whole showed that both PCBs and HCB predicted lower IQ in a generally independent fashion. There was, however, overlap in the variance in some cases, and the pattern of findings between the two was remarkably similar. These results may provide some evidence for the potential involvement of non-PCB organochlorines in the Oswego study. To explore negative confounding, we analyzed the relationships between PCB exposure and demographic variables as well as the unadjusted and adjusted relationships between PCB and IQ. Results revealed that placental PCB levels were associated with older mothers who were more educated and came from higher socioeconomic strata. Due to this fact, unadjusted relationships between PCBs and IQ appeared null or slightly positive. After control for confounders, several significant negative associations between PCBs and IQ were revealed. These data might suggest that inadequate control for confounders in PCB studies, where negative confounding is present, may bias results toward the null (Type II error) rather than spurious associations (Type I error). This pattern of confounding with PCB exposure in the Oswego study also has implications for the assessment of risk. The most highly exposed children came from families with somewhat higher socioeconomic status, and tended to score in the average to above-average range, well above IQ scores that are considered “at risk.” Further, such children were exposed prenatally to PCBs through maternal consumption of PCB-contaminated Great Lakes fish in the early 1990s, when PCB levels were higher than today.

There is growing evidence that childhood IQ is inversely associated with mortality in later life. However, the specificity of this association in terms of causes of death, whether it is continuous over the whole range of IQ scores and whether it is the same according to age and sex is not clear. In a large cohort (N = 11,603) of a complete population of children born in one city in the UK in the early 1950s, IQ measured at age 7 years (using a routinely administered picture test) was found to be inversely associated with mortality between the ages of 15 and 57 years. For every 1 SD increase in IQ at 7, the all cause mortality hazard ratio was 0.79 (95% CI 0.73, 0.85). On adjustment for a range of perinatal factors, father's social class at birth, number of sibs in the household and childhood height and weight, this was attenuated slightly to 0.81 (0.74, 0.88). Almost identical associations of IQ with mortality were seen for men and women as well as at younger (15–39) and older (40+) ages. These associations were across the entire IQ range, although some of the high mortality in the lowest category of IQ (< 70) was accounted for by causes associated with congenital disorders. Overall, external causes of death showed the strongest association, with weaker associations being seen for cancer. Further work is required to understand the mechanisms whereby childhood IQ has such a robust association with mortality in later life.

The present study examined the impact of both perinatal maternal depression and cocaine use on infant neurobehavior at 1 month of age in a large, multi-site study.

Infant neurobehavior was examined in 1053 infants at 1 month of age using the NICU Network Neurobehavioral Scale (NNNS). Mothers were interviewed using The Addiction Severity Index to determine present and past psychiatric history. Four groups were derived from the total sample: 385 prenatally cocaine-exposed infants, 76 whose mothers reported current postpartum depression (DEP/COC) and 309 without current postpartum depression (nonDEP/COC); 668 infants were not exposed to cocaine, 104 whose mothers reported current postpartum depression (DEP/nonCOC), 564 without current postpartum depression (nonDEP/nonCOC). A 2 × 2 Analysis of Covariance was used with covariates (birthweight, maternal age, SES, nicotine, alcohol, and research site) to examine infant neurobehavior in these four conditions. Secondary analyses were conducted to examine the effects of amount and timing of prenatal cocaine exposure.

DEP group by COC exposure status interactions were significant; there was only a DEP effect in the nonCOC infants. Infants in the nonCOC/DEP group had poorer self-regulation and more stress signs, excitability, and arousal than infants in the other groups.

Postpartum maternal depression has negative effects on infant neurobehavior at 1 month of age. Prenatal cocaine exposure may serve to suppress or buffer the effects of postpartum depression on infant neurobehavior. Maternal mood could explain some of the inconsistencies found in the prenatal cocaine exposure literature.

Objective To study absolute and interpeak latencies of the auditory brain response in infants exposed to cocaine and/or opiates in utero. Study design The sample included 477 exposed and 554 comparison infants matched for race, sex, and gestational age. Mothers were recruited at 4 urban university-based centers; most were black, receiving public assistance, and had received adequate prenatal care. Exposure was determined by meconium assay and self-report with alcohol, marijuana, and tobacco present in both groups. At 1 month, infants were tested by masked examiners with the auditory brain response. Results Analyses were conducted for exposed and comparison groups and for level of prenatal cocaine exposure with adjustment for covariates (alcohol, marijuana, tobacco, gestational age at birth, social class, and site). Heavy prenatal cocaine exposure (≥3 days per week, first trimester) led to an increase in the I-III, I-V, and III-V interpeak latencies and to a shorter latency to peak I. Infants with prenatal opiate exposure showed a longer latency to peak V and a longer III-V interpeak latency. Conclusions Prenatal cocaine and/or opiate exposure affects neural transmission. Detection of these effects requires a large sample with control for gestational age, other drugs, and level of cocaine use. (J Pediatr 2003;142:279-85)

From this detailed review of the literature, several conclusions can be drawn: (a) An association between caffeine consumption and a reproductive hazard is more likely to be seen in lower-quality studies than in studies that come closer to approximating the ideal. This is especially evident for “lower” birthweight and congenital anomalies. (b) The association between caffeine consumption and spontaneous abortion may well reflect the Stein–Susser epiphenomenon (women with prominent nausea tend to reduce caffeine consumption and nausea appears to be a marker of good implantation, perhaps reflecting a favorable balance of hormones produced by a healthy placenta). (c) The claim that caffeine consumption by women delays conception has not been followed by convincing support. (d) Reproductive hazards associated with cigarette smoking tend to be associated with caffeine/coffee consumption. Sometimes this appears to be a consequence of residual confounding associated with inadequate adjustment for cigarette smoking, which is over-represented among those who drink the most coffee/caffeine. Sometimes this reflects the tendency of women to underreport socially undesirable behaviors (e.g. smoking) while accurately reporting socially neutral behaviors (e.g. coffee and caffeine consumption). Thus, it seems reasonable to conclude that no convincing evidence has been presented to show that caffeine consumption increases the risk of any reproductive adversity.