Cohort Profile: The Health Outcomes and Measures of the Environment (HOME) study

Early life environmental factors may increase the risk of disease in childhood or adulthood. Well-established examples of early life environmental factors affecting later life health include the increased risk of vaginal clear cell carcinoma and several reproductive disorders following in utero diethylstilbestrol exposure, and cognitive decrements in children with prenatal mercury or childhood lead exposure.1–4 These studies indicate that the effect of some environmental factors may depend on the timing of exposure relative to sensitive developmental windows. There is concern that exposure to environmental chemicals during gestation, infancy or childhood may increase the risk of neurodevelopmental disorders, obesity or allergic diseases.5 However, we know little about the health effects of individual chemicals and even less about mixtures of environmental chemicals.6,7 Thus, prospective and longitudinal cohort studies with repeated exposure and outcome assessments offer the opportunity to determine whether and when human health is affected by environmental chemical exposures. We established the Health Outcomes and Measures of the Environment (HOME) Study, a prospective pregnancy and birth cohort in the greater Cincinnati OH metropolitan area, to determine whether early life environmental chemical exposures influence children’s health. The National Institute of Environmental Health Sciences and the United States Environmental Protection Agency originally funded the HOME Study through their Children’s Environmental Health Center Program to examine the relationship of prenatal lead, tobacco smoke, mercury, polychlorinated biphenyl (PCB) and pesticide exposures with children’s cognitive and behavioural development between birth and 3 years of age. We nested a randomized trial within the cohort to assess the efficacy of lead and injury hazard controls on children’s blood lead levels, cognitive and behavioural development and risk of household injury. Continued funding from the National Institutes of Health and various foundations has allowed us to measure a multitude of chemical exposures and conduct additional assessments through 8 years of age. Anticipated funding will allow us to conduct follow-up when children are 12 years old. Between March 2003 and January 2006, we recruited pregnant women to participate in a longitudinal pregnancy and birth cohort study. We identified women living in a nine-county region of the Cincinnati OH metropolitan area (Brown, Butler, Campbell, Clermont, Hamilton and Warren counties) and Northern Kentucky (Campbell, Brenton and Boone counties) using the medical scheduling systems of nine prenatal practices affiliated with three hospitals. Eligibility was determined using clinic records and phone interviews with women. Eligibility criteria included: living in the study region, < 19 weeks pregnant, > 18 years old, residing in a home built in or before 1978, not living in a mobile or trailer home, HIV-negative, not taking medications for seizures or thyroid disorders, planning to continue prenatal care and deliver at the collaborating clinics and hospitals, planning to live in the greater Cincinnati area for the next year, fluent in English and no diagnosis of diabetes, bipolar disorder, schizophrenia or cancer that resulted in radiation treatment or chemotherapy. To target children at increased risk of lead exposure, we enrolled women living in homes built before 1978 and stratified enrolment so that approximately 50%, 38% and 12% of women would be from the city of Cincinnati, surrounding suburbs and surrounding rural areas, respectively.8 We oversampled women who self-identified as Black (31%) to investigate potential health disparities. Eligibility and enrolment flowchart for the HOME Study. We compared HOME Study participants with women > 18 years of age who delivered infants in the study region from 2003 to 2004 (Table 1 Comparison of sociodemographic characteristics of HOME Study participants with a live birth (2003–06) and women in the study region with live births (2003–04) aStudy region originally included Brown, Butler, Hamilton, Clermont and Warren counties in Ohio, and Campbell, Brenton and Boone counties in Kentucky. We show study region data for Butler, Hamilton, Clermont and Warren counties since no participating women enrolled from Brown county and only two women enrolled from Kentucky. bSix women were missing sociodemographic information. cStudy region data come from the National Center for Health Statistics Birth Data. Data 2005–06 were not publicly available.32 Comparison of sociodemographic characteristics of HOME Study participants with a live birth (2003–06) and women in the study region with live births (2003–04) aStudy region originally included Brown, Butler, Hamilton, Clermont and Warren counties in Ohio, and Campbell, Brenton and Boone counties in Kentucky. We show study region data for Butler, Hamilton, Clermont and Warren counties since no participating women enrolled from Brown county and only two women enrolled from Kentucky. bSix women were missing sociodemographic information. cStudy region data come from the National Center for Health Statistics Birth Data. Data 2005–06 were not publicly available.32 We met women at their prenatal clinic appointments at an average of 16.0 and 26.4 weeks of gestation and at the hospital within 48 h of delivery (Table 2 Summary of measurements collected in HOME Study women and children (Cincinnati, OH, 2003–14) CV, clinic visit; HV, home visit; W, week; Y, year. aMeasures collected within 48 h of delivery. bMaternal urine. cWe collected both cord blood and maternal blood. dMaternal DNA collected from 26-week blood sample and child DNA collected from cord blood sample. e3 year visit only. Summary of measurements collected in HOME Study women and children (Cincinnati, OH, 2003–14) CV, clinic visit; HV, home visit; W, week; Y, year. aMeasures collected within 48 h of delivery. bMaternal urine. cWe collected both cord blood and maternal blood. dMaternal DNA collected from 26-week blood sample and child DNA collected from cord blood sample. e3 year visit only. Sociodemographic, perinatal and infant characteristics of HOME Study women who had a live birth (n = 398) and those who completed follow-up with their child at 8 years of age (n = 233)a aOf the 398 women with a live birth, 6 women dropped out before we collected sociodemographic information. Thus, 2 to 6 women are missing data for race, parity, prenatal vitamin use, and household income. bThe median, 25th percentile and 75th percentile of the number of visits was calculated by summing the number of follow-up visits a mother-child pair completed at 4 weeks and 1, 2, 3, 4, 5 and 8 years of age. We counted clinic and home visits conducted at 1, 2 and 3 years of age as one visit for each year, even for those completing both a clinic and home visit in a given year. cDetermined using serum cotinine concentrations at 16 or 26 weeks of gestation or at delivery. Women with a serum cotinine > 3 ng/ml at any visit were classified as smokers.33 Sociodemographic, perinatal and infant characteristics of HOME Study women who had a live birth (n = 398) and those who completed follow-up with their child at 8 years of age (n = 233)a aOf the 398 women with a live birth, 6 women dropped out before we collected sociodemographic information. Thus, 2 to 6 women are missing data for race, parity, prenatal vitamin use, and household income. bThe median, 25th percentile and 75th percentile of the number of visits was calculated by summing the number of follow-up visits a mother-child pair completed at 4 weeks and 1, 2, 3, 4, 5 and 8 years of age. We counted clinic and home visits conducted at 1, 2 and 3 years of age as one visit for each year, even for those completing both a clinic and home visit in a given year. cDetermined using serum cotinine concentrations at 16 or 26 weeks of gestation or at delivery. Women with a serum cotinine > 3 ng/ml at any visit were classified as smokers.33 Participating women who had a live birth were predominately non-Hispanic White (62%), 30 to < 35 years old at delivery (31%), multiparous (56%), non-smokers (86%) and had a bachelor’s degree or greater (51%) (Table 3). In baseline characteristics, women who returned to the clinic with their child at 8 years of age (n = 233 women, n = 237 children) were similar to the original cohort (Table 3). However, compared with the original cohort, a slightly greater proportion of women who were non-Hispanic White, well educated and had higher household income returned for visits from 1 to 5 years following delivery (results not shown). In addition, women with these characteristics returned to our study clinic for more visits (Table 3). All of the 407 live-born children are still eligible for follow-up. In addition, beginning with the 8-year visit, we attempted to re-engage children born to women who dropped out during the ‘run-in’ period. Threeof these children (one set of twins and one singleton) returned for the 8-year visit. We have extensive, multimodal and repeated assessments of environmental chemical exposures, child health and confounders in both mothers and children (Table 2). If possible, we used measures used by other birth cohorts to facilitate pooled analyses.9 We designed our biospecimen collectionto identify unique windows of vulnerability to environmental chemicals during gestation or childhood. In addition, we lot-tested our collection materials for contamination by chemicals that might be used in the secollection supplies (e.g. phthalates in diaper inserts).10 We collected > 37 000 biological specimens at multiple time points from women during pregnancy and the postpartum period and from children from 1 to 8 years of age (Table 2). Unique biospecimens include neonatal meconium, vernix from a subset of 122 infants, breast milk from mothers of breastfed infants and formula from weaned or formula-fed infants. We also collected clotted red blood cells and shed deciduous teeth from children at 8 years of age. To date, we have measured > 100 different chemicals in our participants’ biospecimens. These include lead, cadmium, mercury, arsenic, tobacco smoke metabolites, PCBs, organochlorine pesticides, polybrominated diphenyl ethers (PBDEs), phthalate metabolites, environmental phenols, perfluoroalkyl substances and organophosphorous/pyrethroid pesticide metabolites. For many chemicals, we have multiple measures during pregnancy and childhood (Supplementary Table 1, available as Supplementary data at IJE online). We measured whole-blood folate, urinary iodine and serum thyroid hormone concentrations during pregnancy, as well as cord blood thyroid hormone concentrations. Finally, we extracted DNA from maternal and child samples. We collected > 15 000 environmental samples from participants’ homes at 20 weeks of gestation and when children were 1, 2 and 3 years old. We collected outdoor soil samples at 20 weeks of gestation and dust wipes of floors, window troughs and windowsills during the first 3 years of life. We collected tap water at 20 weeks of gestation and floor or carpet dust with a high volume small surface sampler at all visits. We measured dust lead loadings in house dust and lead concentrations in soil, water and paint. We quantified organohalogen flame retardant concentrations in a subset of homes. We administered standardized questionnaires about pregnant women or children’s exposure to environmental chemicals. We initially designed our questionnaires to identify sources of exposure to pesticides, mercury, lead and tobacco smoke. We added questions about exposures to phthalates, bisphenol A, parabens, perfluoroalkyl substances and flame retardants when children were 5 and 8 years old. We used examiner-administered tests and parent-reported surveys to assess several neurobehavioural domains that might be affected by environmental chemical exposures.11 We developed our protocols to accommodate the unique aspects of conducting neurobehavioural assessments in children. All examiners underwent intensive training including discussions of test goals, psychometric properties, directions for proper administration and scoring, and strategies for testing young children. Examiners completed a prescribed number of practice examinations on age-appropriate pilot subjects and received frequent feedback before presenting for a ‘certification’ test. Once certified, our developmental psychologist (K.Y.) conducted quality checks every 6 months to ensure continued reliability. Quality control procedures also included observation of test administration and review of scoring forms. We conducted all neurobehavioural assessments at > 1 year of age in a clinic setting to reduce the influence of different environments or distractions on children’s performance. We assessed infant neurological and behavioural function within 48 h of delivery in the delivery hospital and at 4 weeks of age in the home using the NICU Network Neurobehavioral Scale. Between 1 and 8 years of age, we administered the Bayley Scales of Infant Development-II, Wechsler Preschool and Primary Scales of Intelligence-III and the Wechsler Intelligence Scales for Children-IV to assess mental and psychomotor development and cognitive abilities [i.e. intelligence quotient [IQ]). We also repeatedly assessed continuous and quantitative phenotypes of clinical disorders using parent-reported measures of child behaviour. These included the Social Responsiveness Scale to assess features of autism spectrum disorders and the Behavior Assessment System for Children-2 to assess features of attention-deficit/hyperactivity disorder (ADHD), conduct disorder, anxiety and depression. We assessed children’s executive function with the Behavior Rating Inventory of Executive Function, Conner’s Continuous Performance Task, Shape School, TRAILS-P, delay of gratification test and NEPSY. We assessed children’s language abilities, reading readiness and academic achievement with the Clinical Evaluation of Language Fundamentals, Woodcock-Johnson-III, and Wide Range Achievement Test-4, respectively. At 8 years of age, we assessed children’s play behaviours/preferences, gender identity and anxiety. Finally, we assessed children’s visual-spatial abilities using the Virtual Morris Water Maze, a computerized version of a rodent test used in toxicology studies. We abstracted neonatal anthropometry from medical records. Study staff measured weight, length/height and head circumference when children were 4 weeks and 1, 2, 3, 4, 5 and 8 years old. We measured weight using a digital scale with children dressed in undergarments or a dry diaper, except at 4 weeks when we weighed infants while they were naked. We measured recumbent length with a length board and standing height with a wall-mounted stadiometer. We assessed head circumference using a paper tape measure. At 4, 5 and 8 years of age, we measured waist circumference around a horizontal plane defined by the left and right iliac crests using a plastic measuring tape. Finally, we measured children’s body fat using a Tanita children’s body fat monitor at 8 years of age. We used standardized questions based on those from the National Health and Nutrition Examination Survey to assess wheeze, eczema, allergy and asthma symptoms at 6-month intervals between 6 months and 5 years of age and again when children were 8 years old. When children were 4 and 5 years old, we attempted to collect at least three acceptable forced expiratory volume (FEV)-1 measurements using a portable spirometer. At the 3-, 4- and 5-year visit, we measured exhaled nitric oxide, a marker of airway inflammation. We asked parents about child injuries in the home at between 3 months and 6 years of age. We provided parents with calendars to record events as they occurred. Medically attended injuries were those that prompted parents to call or visit a physician’s office, urgent care or emergency department. We classified injuries as modifiable if one of the installed interventions (e.g. wall-mounted stair gate) could have prevented the hazard or mechanism from causing the injury. We confirmed emergency room visits for residential injuries using the Hamilton County Injury Surveillance System. Because low-level lead exposure is a risk factor for intellectual deficits and behavioural problems in children, we conducted a randomized controlled trial in 355 participants to test the efficacy of lead abatement techniques on lead-contaminated house dust, blood lead concentrations and children’s neurodevelopment. Families were randomly assigned to receive either lead hazard controls (intervention group) before the child’s birth or equipment to reduce residential injuries (control group). The intervention consisted of reducing lead hazards (e.g, paint stabilization, window repair and extensive dust control). The control group received injury prevention devices (e.g. stair gates) or home modifications (e.g. reducing water heater temperature). Because numerous factors could confound the association between chemical exposures and child health, we collected an extensive set of covariates. We measured sociodemographic factors including maternal race/ethnicity, age, education, marital status, employment, insurance and household income. We assessed maternal depressive symptoms, ADHD, mental health and intelligence. We collected peri- and antenatal factors like maternal anthropometry, parity, breastfeeding and alcohol/drug/tobacco use. We abstracted clinical information from the mother and infant’s medical records. We measured some dietary behaviours during pregnancy (e.g. prenatal vitamin intake) and dietary/lifestyle factors in childhood (e.g. fresh fruit and vegetable consumption and physical/sedentary activity). Finally, we assessed parenting stress, caregiving environment, child sleep and child mouthing behaviours. Below we summarize some of our to between environmental chemicals and children’s health. of our be urinary bisphenol concentrations were with increased problems and decrements in executive function in not at 2 and 3 years of In addition, urinary concentrations during pregnancy were with cord blood thyroid hormone concentrations in concentrations were also with increased risk of and function at years of However, prenatal concentrations were not with infant behaviours at years of age or body at years of concentrations during pregnancy were with cognitive abilities, increased problems and increased behaviours in children at years of In addition, maternal concentrations during pregnancy were with increased maternal and concentrations in There were no between maternal and infant concentrations HOME Study women were as high as concentrations in pregnant women We that higher maternal concentrations during pregnancy were with increased in infants, increased at 8 years of age in a and between 2 and 8 years of and between maternal urinary or serum chemical concentrations and Social Responsiveness Scale in 4- and Cincinnati children. are as The are the in children born to women with of these chemicals. All other chemicals were as continuous that are by two their to on a scale to the include maternal age, race, marital status, education, parity, insurance during pregnancy, prenatal vitamin use, during pregnancy, household depressive symptoms during pregnancy, caregiving serum cotinine and all chemical concentrations. indicate the around the maternal serum cotinine concentrations during pregnancy, a of tobacco smoke exposure, were with in increased in infancy and increased body in children at 2 and 3 We that the injury intervention was with a in attended injuries through 2 years of intervention were the control of the randomized trial to reduce residential lead of the of the HOME Study is the longitudinal and multimodal of environmental chemical exposures during gestation and childhood. For many chemicals, we measured exposure to during pregnancy and to during childhood (Supplementary Table available as Supplementary data at IJE online). is the repeated of several child health us to examine the of any as well as of child health. Finally, we collected measures of many potential of the HOME Study is the sample reducing the to small effect and us from clinical (e.g. is not unique to the HOME Study, is to follow-up follow-up at 8 years of However, measured sociodemographic factors were not different participants that and not follow-up at 8 years of age. In addition, follow-up at years of age was to maternal chemical concentrations at Finally, our may not be to other since we in our cohort not of births in the study chemical concentrations women in our cohort are similar to those of pregnant women in the The HOME Study have in We with other and and to additional information about the HOME Study and a The HOME Study Data to review and ensure that they not with and are an of (e.g. cord in a The HOME Study is a prospective pregnancy and birth cohort designed to investigate the influence of environmental chemical exposures on children’s and Between March 2003 and January 2006, we recruited pregnant women from a region surrounding Cincinnati, follow-up included 1 home and 2 clinic visits during pregnancy, a visit within 48 h of and home or clinic visits when children were 4 weeks and 1, 2, 3, 4, 5 and 8 years of age. 407 live-born children, all are eligible for follow-up and 237 children completed follow-up at 8 years of age. Data include longitudinal and repeated measures of chemical exposures and child health. environmental chemical exposures measured in women and children, as well as repeated measures of child anthropometry, and allergy and in more about The HOME Study and Supplementary data are available at IJE The National Institutes of Health Institute of Environmental Health Environmental Protection of and Brown of Institute and provided funding for the HOME Study. of The no

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.