Nicholas P. Jewell

Organophosphate Pesticide Exposure and Neurodevelopment in Young Mexican-American Children

Background Organophosphate (OP) pesticides are widely used in agriculture and homes. Animal studies suggest that even moderate doses are neurodevelopmental toxicants, but there are few studies in humans. Objectives We investigated the relationship of prenatal and child OP urinary metabolite levels with children’s neurodevelopment. Methods Participating children were from a longitudinal birth cohort of primarily Latino farm-worker families in California. We measured six nonspecific dialkylphosphate (DAP) metabolites in maternal and child urine as well as metabolites specific to malathion (MDA) and chlorpyrifos (TCPy) in maternal urine. We examined their association with children’s performance at 6 (n = 396), 12 (n = 395), and 24 (n = 372) months of age on the Bayley Scales of Infant Development [Mental Development (MDI) and Psychomotor Development (PDI) Indices] and mother’s report on the Child Behavior Checklist (CBCL) (n = 356). Results Generally, pregnancy DAP levels were negatively associated with MDI, but child measures were positively associated. At 24 months of age, these associations reached statistical significance [per 10-fold increase in prenatal DAPs: β = −3.5 points; 95% confidence interval (CI), −6.6 to −0.5; child DAPs: β = 2.4 points; 95% CI, 0.5 to 4.2]. Neither prenatal nor child DAPs were associated with PDI or CBCL attention problems, but both prenatal and postnatal DAPs were associated with risk of pervasive developmental disorder [per 10-fold increase in prenatal DAPs: odds ratio (OR) = 2.3, p = 0.05; child DAPs OR = 1.7, p = 0.04]. MDA and TCPy were not associated with any outcome. Conclusions We report adverse associations of prenatal DAPs with mental development and pervasive developmental problems at 24 months of age. Results should be interpreted with caution given the observed positive relationship with postnatal DAPs.

To GEE or not to GEE: comparing population average and mixed models for estimating the associations between neighborhood risk factors and health.

Two modeling approaches are commonly used to estimate the associations between neighborhood characteristics and individual-level health outcomes in multilevel studies (subjects within neighborhoods). Random effects models (or mixed models) use maximum likelihood estimation. Population average models typically use a generalized estimating equation (GEE) approach. These methods are used in place of basic regression approaches because the health of residents in the same neighborhood may be correlated, thus violating independence assumptions made by traditional regression procedures. This violation is particularly relevant to estimates of the variability of estimates. Though the literature appears to favor the mixed-model approach, little theoretical guidance has been offered to justify this choice. In this paper, we review the assumptions behind the estimates and inference provided by these 2 approaches. We propose a perspective that treats regression models for what they are in most circumstances: reasonable approximations of some true underlying relationship. We argue in general that mixed models involve unverifiable assumptions on the data-generating distribution, which lead to potentially misleading estimates and biased inference. We conclude that the estimation-equation approach of population average models provides a more useful approximation of the truth.

Diaphragm and lubricant gel for prevention of HIV acquisition in southern African women: a randomised controlled trial.

BACKGROUND Female-controlled methods of HIV prevention are urgently needed. We assessed the effect of provision of latex diaphragm, lubricant gel, and condoms (intervention), compared with condoms alone (control) on HIV seroincidence in women in South Africa and Zimbabwe. METHODS We did an open-label, randomised controlled trial in HIV-negative, sexually active women recruited from clinics and community-based organisations, who were followed up quarterly for 12-24 months (median 21 months). All participants received an HIV prevention package consisting of pre-test and post-test counselling about HIV and sexually transmitted infections, testing, treatment of curable sexually transmitted infections, and intensive risk-reduction counselling. The primary outcome was incident HIV infection. This study is registered with ClinicalTrials.gov, number NCT00121459. FINDINGS Overall HIV incidence was 4.0% per 100 woman-years: 4.1% in the intervention group (n=2472) and 3.9% in the control group (n=2476), corresponding to a relative hazard of 1.05 (95% CI 0.84-1.32, intention-to-treat analysis). The proportion of women using condoms was significantly lower in the intervention than in the control group (54%vs 85% of visits, p<0.0001). The proportions of participants who reported adverse events (60% [1523] vs 61% [1529]) and serious adverse events (5% [130] vs 4% [101]) were similar between the two groups. INTERPRETATION We observed no added protective benefit against HIV infection when the diaphragm and lubricant gel were provided in addition to condoms and a comprehensive HIV prevention package. Our observation that lower condom use in women provided with diaphragms did not result in increased infection merits further research. Although the intervention seemed safe, our findings do not support addition of the diaphragm to current HIV prevention strategies.

Some Surprising Results About Covariate Adjustment in Logistic Regression Models

Summary Results from classic linear regression regarding the effect of adjusting for covariates upon the precision of an estimator of exposure effect are often assumed to apply more generally to other types of regression models. In this paper we show that such an assumption is not justified in the case of logistic regression, where the effect of adjusting for covariates upon precision is quite different. For example, in classic linear regression the adjustment for a non-confounding predictive covariate results in improved precision, whereas such adjustment in logistic regression results in a loss of precision. However, when testing for a treatment effect in randomized studies, it is always more efficient to adjust for predictive covariates when logistic models are used, and thus in this regard the behavior of logistic regression is the same as that of classic linear regression.

Should people at low risk of cardiovascular disease take a statin

A review of statins for primary prevention of cardiovascular disease could alter guidance for those with a 10 year risk of less than 10%. John Abramson and colleagues argue that statins have no overall health benefit in this population and that prescribing guidelines should not be broadened

AIDS epidemiology : methodological issues

This is a collection of state-of-the-art reviews and original research work dealing with statistical and mathematical methodology relevant to the AIDS epidemic. It will be of interest and practical application to researchers and students in statistics, epidemology and public health.

Statistical analysis of HIV infectivity based on partner studies.

Partner studies produce data on the infection status of partners of individuals known or assumed to be infected with the human immunodeficiency virus (HIV) after a known or estimated number of contacts. Previous studies have assumed a constant probability of transmission (infectivity) of the virus at each contact. Recently, interest has focused on the possibility of heterogeneity of infectivity across partnerships. This paper develops parametric and nonparametric procedures based on partner data in order to examine the risk of infection after a given number of contacts. Graphical methods and inference techniques are presented that allow the investigator to evaluate the constant infectivity model and consider the impact of heterogeneity of infectivity, error in measurement of the number of contacts, and regression effects of other covariates. The majority of the methods can be computationally implemented easily with use of software to fit generalized linear models. The concepts and techniques are closely related to ideas from discrete survival analysis. A data set on heterosexual transmission is used to illustrate the methods.

Statistical models for prevalent cohort data

In prospective cohort studies individuals are sometimes recruited according to a certain cross-sectional sampling criterion. A prevalent cohort is defined as a group of individuals who have a certain disease at enrollment into the study. Statistical models for the analysis of prevalent cohort data are considered when the onset or diagnosis time of the disease is known. The incident proportional hazards model, where the time scale is duration with disease, is compared to the prevalent proportional hazards model, where the fundamental time scale is follow-up time. In certain cases the time of enrollment may coincide with another event (such as the initiation of treatment). This situation is also considered and its limitations highlighted. To illustrate the methodological ideas discussed in the paper, the analysis of data from an observational study of zidovudine (ZVD) in patients with the acquired immunodeficiency syndrome (AIDS) is presented.

Depuration of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) in breast milk from California first-time mothers (primiparae).

Background Little is known about the rates of loss (depuration) of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) from mothers during lactation. Depuration rates affect infant exposure to chemicals during breast-feeding, and fetal and lactational transfers during subsequent pregnancies. Objective Our objective in this study was to estimate depuration rates of PBDEs and PCBs using serial samples of breast milk. Method Nine first-time mothers (primiparae) each collected samples at 4, 6, 8, 12, 16, 20, and 24 weeks after birth. Nine additional primiparae each collected two samples at varying time intervals (18 to > 85 weeks after birth). Analytical precision was assessed to evaluate the accuracy of measured monthly percentage declines in PBDEs and PCBs. Results The four major PBDE congeners decreased 2 or 3% ± 1% per month over the 6-month period. These decreases were consistent over a 50-fold range (21–1,330 ng/g lipid weight) of initial PBDE concentrations in breast milk. The change in PCB-153 ranged from + 0.3% to –0.6% per month, with heterogeneous slopes and greater intraindividual variability. PBDE and PCB concentrations declined 1% per month over longer periods (up to 136 weeks). Conclusions Our data indicate that PBDEs and PCBs are not substantially (4–18%) reduced in primiparae after 6 months of breast-feeding. Consequently, the fetal and lactational exposures for a second child may not be markedly lower than those for the first. Participants were volunteers from a larger study population (n = 82), and were typical in their PBDE/PCB levels and in many demographic and lifestyle factors. These similarities suggest that our results may have broader applicability.

Association of housing disrepair indicators with cockroach and rodent infestations in a cohort of pregnant latina women and their children

Health burdens associated with poor housing and indoor pest infestations are likely to affect young children in particular, who spend most of their time indoors at home. We completed environmental assessments in 644 homes of pregnant Latina women and their children living in the Salinas Valley, California. High residential densities were common, with 39% of homes housing > 1.5 persons per room. Housing disrepair was also common: 58% of homes had peeling paint, 43% had mold, 25% had water damage, and 11% had rotting wood. Evidence of cockroaches and rodents was present in 60% and 32% of homes, respectively. Compared with representative national survey data from the U.S. Department of Housing and Urban Development, homes in our sample were more likely to have rodents, peeling paint, leaks under sinks, and much higher residential densities. The odds of rodent infestations in homes increased in the presence of peeling paint [odds ratio (OR) 2.1; 95% confidence interval (CI), 1.5–3.1], water damage (OR 1.9; 95% CI, 1.2–2.7), and mold (OR 1.5; 95% CI, 1.0–2.1). The odds of cockroach infestation increased in the presence of peeling paint (OR 3.8; 95% CI, 2.7–5.6), water damage (OR 1.9; 95% CI, 1.2–2.9), or high residential density (OR 2.1; 95% CI, 1.2–3.8). Homes that were less clean than average were more prone to both types of infestations. Pesticides were stored or used in 51% of households, partly to control roach and rodent infestations. These data indicate that adverse housing conditions are common in this community and increase the likelihood of pest infestations and home pesticide use. Interventions to improve housing and promote children’s health and safety in this population are needed.