For the following questions, please refer to the posted article: Tehranifar, et al. (2008). Immigration and Risk of Chil
Posted: Mon Jul 11, 2022 2:21 pm
For the following questions, please refer to the posted article:Tehranifar, et al. (2008). Immigration
andRisk of Childhood Lead Poisoning: Findings from a Case-ControlStudy of New York Children.
American Journal of Public Health, 98(1), 92-97.
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Q ||| Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 1/8 84% + | 1 2 3 Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-Control Study of New York City Children Parisa Tehranifar, DPH, Jessica Leighton, PhD, Amy H. Auchincloss, PhD, MPH, Andrew Faciano, MPH, Howard Alper, PhD, Andrea Paykin, PhD, and Songmei Wu, DrPH Despite a dramatic decline in childhood lead poisoning in the United States, an estimated 1.6% of US children aged 1 to 5 years (ap- proximately 310000 children) have elevated blood lead levels of at least 10 µg/dL. Fur- thermore, the growing scientific evidence on cognitive impairments associated with blood lead levels below 10 µg/dL suggest that even a greater number of children are at risk of being adversely affected by low-level exposure to lead, 24 The most common high-dose source of lead exposure among US children is interior lead-based paint, but children who spend time outside the United States may be ex- posed to additional sources. The literature on lead poisoning among immigrant chil- dren living in the United States is scant. Sur- veillance and case studies have revealed a higher prevalence or incidence of elevated blood lead levels among refugee and inter- nationally adopted children living in the United States than among the general US population of children. However, these reports have lacked a comparison group and presented results not adjusted for confound- ing factors. To our knowledge, a recent national study involving Mexican American children is the only investigation in which individual- and family-level characteristics (e.g., age, family income, language spoken at home) and source of drinking water were controlled for in any examination of the association be tween foreign birthplace and lead poisoning. The results of that study revealed higher blood lead levels among foreign-born chil- dren than among US-born children of Mexi- can descent. In this study, we examined the associa tions of childhood lead poisoning with birth and residence in a foreign country among a multiethnic urban sample while accounting for child and family demographic characteris- tics, child behaviors, and current residential Objectives. We investigated whether foreign birthplace and residence were as- sociated with an increased risk of childhood lead poisoning. Methods. We conducted a matched case-control study among New York City children (mean age=3 years) tested for lead poisoning in 2002 (n=203 pairs). Chil- dren were matched on age, date of test, and residential area. Blood lead and hous- ing data were supplemented by a telephone survey administered to parents or guardians. Conditional logistic regression analysis was used to examine the rela- tionship of lead poisoning status to foreign birthplace and time elapsed since most recent foreign residence after adjustment for housing and behavioral risk factors. Results. Both foreign birthplace and time since most recent foreign residence had strong adjusted associations with lead poisoning status, with children who had lived in a foreign country less than 6 months before their blood test show- ing a particularly elevated risk of lead poisoning relative to US-born children with no foreign residential history before their blood test (odds ratio [OR] =10.9; 95% confidence interval [CI]-3.3, 36.5). Conclusions. Our findings demonstrate an increased risk of lead poisoning among immigrant children. (Am J Public Health. 2008;98:92-97. doi:10.2105/ AJPH.2006.093229) building characteristics. We hypothesized that lead poisoning would be positively associated with birth and residence in a foreign country after control for known risk factors. METHODS Design We used a matched case-control (1:1) de- sign, which allowed for a relatively low-cost study, a small sample size, and efficient con- trol for a complex web of demographic and environmental factors. Participants were re- cruited and data were collected during 2002 and 2003. Children were matched on age (mean age difference: 3 months; range: 0-9 months), date of lead poisoning test (mean test date difference-17 days; range-2-91 days), and residential area (defined via aggregated con- tiguous zip codes as established by the United Hospital Fund). We included these variables because the prevalence of children's mouthing behaviors (placing objects and hands in mouth) and their likelihood of un- dergoing blood lead testing decrease with 92 | Research and Practice | Peer Reviewed | Tehranifar et al. age, both blood lead testing and blood lead. levels exhibit seasonal trends, and building characteristics, blood lead levels, and likeli- hood of blood lead testing are associated with area of residence, 1839 Data Sources Children eligible to participate were youn- ger than 18 years, resided in New York City, had been tested for lead poisoning between May and December 2002, and had blood lead test results reported to the blood lead registry of the New York City Department of Health and Mental Hygiene's Lead Poisoning Prevention Program (LPPP; in New York State, laboratories are required to report all blood lead tests to the state health department). Case children were defined as those having a first- time blood lead level of 20 µg/dL. or greater from a venous sample or 2 tests with blood lead levels between 15 and 19 µg/dL. taken at least 3 months apart, with the second test in- volving a venous sample (LPPP's definition of lead poisoning cases requiring environmental interventions during the study period). Control children had blood lead levels of 5 µg/dL. or American Journal of Public Health | January 2008, Vol 98, No. 1 •= Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 2 /8 | - 84% + B below from a venous or capillary sample and no previous elevated blood lead level. All blood samples were analyzed by labo- ratories certified by the New York State De- partment of Health. We obtained information on blood test date, lead level, child's date of birth, and parent or guardian's name and ad- dress from the blood lead registry. We gath- ered information on building age and size from databases maintained by the city of New York. A 30-minute structured telephone ques- tionnaire, administered to the parents or guardians of both case children and controls, was used to collect data on family income, potential sources of lead paint exposure in the home, and child's gender, race/ethnicity, mouthing behaviors, country of birth, and for eign residential history. We pilot tested the questionnaire via a focus group and pretest interviews and translated it into Spanish. ||| 1 2 3 Data Collection A multilingual introductory letter was mailed to the parents or guardians of 1058 eligible children. At least 4 attempts were subsequently made to contact the parents or guardians by telephone. We were more successful in contact- ing parents or guardians of the 308 potential case children (78%) than the 750 potential control children (49%), primarily because, for 2 reasons, we had better contact information for the former. First, case children's blood test results were reported to LPPP by both labora- tories and health care providers, whereas con- trol children's results were reported by labora tories only. Second, LPPP visited the homes of case children as part of care coordination and environmental assessment activities. More families in the control group than in the case group refused participation in the study (22% of the control families vs 12% of the case families contacted). We were unable to assess differences in participation according to foreign birthplace or residence because this information was not available for any of the control families, nor was it available for some of the case families that were not interviewed. We interviewed 211 parents or guardians of case children and 288 parents or guardians of controls. Six case children were excluded because their matched controls were not suc- cessfully enrolled into the study. In addition, 2 case children were excluded as a result of missing or incorrect birth dates. When there were multiple controls, the first successfully in- terviewed control was selected. The final sam- ple included 203 case-control pairs. As a means of maximizing participation among immigrant families that spoke lan- guages other than English, trained bilingual English and Spanish speakers conducted the interviews, and professional interpreters made the initial contacts with the households and administered the questionnaire if needed. The questionnaire was administered in Spanish to 33% (135) of participants and in Haitian Creole to a single participant in the final sample. The distribution of inter- view language was similar for case and con- trol children. Initially, interviewers asked to speak with the mother or father of the child; if the mother and father were not available, they asked, in descending order, for the child's guardian or any available adult relative. For the most part (89% of case children and 95% of control children), parents rather than other adults were interviewed. Measures Children were categorized as US born or for- eign born (children born in Puerto Rico were classified as foreign born). Parents or guardians were asked whether their child had lived out- side of the United States for at least 2 months; if they answered yes, they were asked to pro- vide dates of residence. To capture the timing of potential lead exposures in foreign countries, we calculated the length of time between the most recent date of foreign residence and the time of blood lead testing among children who had lived outside of the United States. Because only a small percentage of chil- dren had lived in a foreign country (17%) and exploratory analyses suggested that the risk of lead poisoning dropped sharply approxi- mately 6 months after most recent foreign residence, we used a simple classification scheme in which 6 months was designated as the cutoff point. That is, children were grouped in the following categories for the final analysis: (1) lived in a foreign country less than 6 months before their blood test, (2) lived in a foreign country 6 or more months before their blood test, and (3) did not live in a foreign country before their blood test. January 2008, Vol 98, No. 1 | American Journal of Public Health Q ✩ A major pathway for lead exposure among urban children is ingestion of deteriorated leaded paint and household dust that have set- tled on their hands or on objects. Lead-based paint hazards are more likely to be present in buildings constructed before 1950, when use of lead paint was more widespread than in later years. We categorized time of building con- struction as before 1950 versus 1950 or after. Parents were asked about the presence of peeling or chipping paint in their home and were asked whether their children put their fingers in their mouths or ate any nonfood items (e.g., toys, crayons). Both of these measures were dichotomized for the pur- poses of our analysis. Because New York City laws aimed at preventing children's ex- posure to lead paint hazards in housing apply to buildings with 3 or more dwelling units, larger buildings may contain fewer lead paint hazards. Thus, we expected lead. poisoning to be inversely associated with building size, categorized according to number of dwelling units (fewer than 3 vs 3 or more). Finally, we asked about use of imported products known to contain lead (e.g., traditional Mexican folk remedies such as azarcon and greta). Data Analysis We used odds ratios (ORs) derived from conditional logistic regression analyses to ex- amine bivariate associations of lead poisoning status with foreign birthplace, time elapsed since most recent foreign residence, and other potential risk factors. In addition, we exam- ined associations between lead poisoning sta- tus and foreign birthplace and time since most recent foreign residence in separate con- ditional logistic regression models that ad- justed for potential confounders exhibiting statistically significant bivariate associations with lead poisoning status. RESULTS The mean age of the sample was 3 years (range: 0-17 years), and 75% of the children were younger than 5 years. As expected, foreign birthplace and time elapsed since most recent foreign residence were strongly correlated (Spearman rank correlation-0.80, P<.001); thus, we were not able to examine Tehranifar et al. | Peer Reviewed | Research and Practice | 93 = Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... ||| Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C.... 3 / 8 - 84% + their independent effects in the same model. Table 1 presents the characteristics of case and control children and the bivariate associ ations between each independent variable and lead poisoning status. 1 2 3 Twenty percent of case children were born outside of the United States, as compared. TABLE 1-Sociodemographic and Risk Characteristics of Participants, by Case-Control Status: New York City, 2002-2003 Place of birth Foreign country United States Time since most recent foreign residence <6 months 26 months Did not live in a foreign country Race/ethnicity Hispanic Asian Other African American Gender Boy Girl Family income level Below 150% of poverty level Above 150% of poverty level Peeling paint in home Yes No Time building of residence constructed Before 1950 1950 or after No. of units in building <3 23 Frequency at which child puts fingers in mouth Sometimes/always with 4% of control children (Table 1). Simi- larly, case children were more likely than were control children (27% vs 8%) to have lived outside of the United States for at least 2 months. In addition to 50 foreign-born chil- dren, 22 US-born children (15 case children, 7 control children) had lived in a foreign Never/rarely Child eats nonfood items Yes No Case Group, No. (%) 41 (20.1) 162 (79.8) 33 (164) 21 (10.4) 147 (73.1) 82 (40.4) 24 (11.9) 19 (9.4) 78 (384) 84 (41.4) 119 (586) 122 (739) 43 (26.1) 88 (44.0) 112 (56.0) 185 (91.6) 17 (8.4) 67 (33.0) 136 (67.0) 153 (76.1) 48 (23.9) 115 (56.7) 88 (43.3) Control Group No. (%) 94 | Research and Practice | Peer Reviewed | Tehranifar et al. 9(4.4) 194 (95.6) 4(2.0) 12 (5.9) 187 (92.1) 100 (49.2) 13 (6.4) 12 (5.9) 78 (38.4) 93 (45.8) 110 (542) 127 (89.8) 55 (302) 68 (33.5) 135 (66.5) 154 (76.6) 47 (234) 18 (8.9) 185 (91.1) 120 (59.1) 83 (40.9) 74 (36.5) 129 (63.5) Unadjusted OR (95% CI) 5.57 (2.49, 12.45) Reference 9.00 (3.16, 25.65) 2.26 (1.04,4.90) Reference 0.70 (0.42, 1.20) 2.18 (0.95,5.04) 1.65 (0.73, 3.72) Reference 0.84 (0.57,1.24) Reference 1.42 (0.84, 2.39) Reference 1.54 (1.03,2.30) Reference 3.23 (1.73.6.02) Reference 6.44 (3.19, 13.02) Reference 2.18 (1.39,3.40) Reference 2.32 (1.52, 3.53) Reference Note. OR-odds ratio; CI-confidence interval. The sample included 203 case-control pairs. "Foreign countries of birth included Haiti (n-12), Pakistan (n-7), the Dominican Republic (n-6), Mexico (n-5), Puerto Rico (n-3), India (n-2), Bangladesh (n-2), Liberia (n-2), Ghana (n-1), Macau (n-1), Israel (n-1), Russia (n-1), Yugoslavia (n-1), Jamaica (n-1), Trinidad and Tobago (n-1), El Salvador (n-1), Ecuador (n-1), Colombia (n-1), and Guyana (n-1). Time elapsed between most recent date of residence in a foreign country (for a minimum of 2 months) and date of blood lead testing. "Including White, American Indian, mixed racial background, and "don't know." Q ✩ country at some time before their blood test. Finally, 3 children (1 case child and 2 control children) had lived abroad for 2 months or longer more than once. There were differences in foreign countries of birth and residence between the case and control groups. Among case children, the main foreign countries of birth and residence. were Haiti, Pakistan, Mexico, and the Domini- can Republic; among control children, the pri- mary countries were Puerto Rico and the Do- minican Republic. There were no differences in use of imported products between the case group and the control group (data not shown). Both foreign birthplace and time since most recent foreign residence were strongly associated with lead poisoning status; children who had lived in a foreign country less than 6 months before their blood test were at par- ticularly elevated risk of lead poisoning rela- tive to US-born children with no foreign resi- dential history before their blood test (OR-9.0; 95% confidence interval [CI]-3.2, 25.7; Table 1). Building characteristics and child behavioral characteristics were associ- ated with lead poisoning status in the ex- pected directions. Case and control children were similar in terms of family income. Per- centages of children of African American and Hispanic descent were roughly similar in the 2 groups, although case children were dispro- portionately Asian. Table 2 summarizes the results of a series of conditional logistic regression models test- ing our hypothesis that birth and residence in a foreign country would be associated with an increased risk of lead poisoning. After ad- justment for housing and child behavioral risk factors, foreign-born children were 5 times more likely than were US-born children to be lead poisoned (OR-5.4; 95 % CI-2.2, 13.5; Table 2, model 1). A history of foreign residence was also strongly associated with lead poisoning status in adjusted models. Children living abroad for less than 6 months and for 6 or more months before their blood test, respectively, were 11 and 3 times more likely to to be lead poisoned than were US-born children who had not lived in a foreign country (Table 2, model 2). Reclassification of 3 con- trol children born in Puerto Rico as US bor strengthened the associations between American Journal of Public Health | January 2008, Vol 98, No. 1 = Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 4/8 | - 84% + E I 2 3 5 TABLE 2-Associations of Foreign Birthplace and Time Elapsed Since Most Recent Foreign Residence With Lead Poisoning Status, Adjusted for Building and Child Behavioral Characteristics: New York City, 2002-2003 Foreign born Time since most recent foreign residence <6 months before blood test 26 months before blood test Did not live in a foreign country (Ref) Building characteristics Peeling paint in home Residence in pre-1950 building Residence in building with fewer than 3 units Child behavioral characteristics Child puts fingers in mouth Child eats nonfood items foreign birth and residence and lead poison- ing status (data not shown). Other risk factors were also positively asso ciated with lead poisoning status in both mul- tivariate models. Residence in a building with fewer than 3 units (OR-5.6; 95% CI-2.5, 12.5; model 2) and residence in a building constructed before 1950 (OR-3.4; 95% CI-1.5, 7.8) were associated with an in- creased risk of lead poisoning. In addition, there remained a higher risk of lead poisoning among children who put their fingers in their mouth (OR=2.8; 95% CI-1.5, 5.2) and ate nonfood items (OR-2.0; 95% CI-1.2, 3.5) than among children who did not. DISCUSSION Model 1, OR (95% CI) 5.39 (2.15, 13.54) We found that, among children tested for lead poisoning, foreign-born children were 5 times more likely than were US-born children to have elevated blood lead levels after ad- justment for building characteristics and child behaviors. An even stronger association was found between lead poisoning and recent res- idence in a foreign country, with children liv- ing abroad for less than 6 months before their blood test being 11 times more likely to have elevated blood lead levels than US-born 1.18 (0.70, 1.98) 2.79 (1.29,6.03) 5.67 (259, 12.41) 2.24 (1.26,3.97) 2.10 (1.24, 3.56) Note. OR-odds ratio; C1-confidence interval 195 matched pairs. 193 matched pairs. "Time elapsed between most recent date of residence in a foreign country (for a minimum of 2 months) and date of blood lead testing. Model 2,"OR (95% CI) 10.94 (3.28, 36.46) 2.92 (1.12, 7.58) 1.00 1.17 (0.69, 2.00) 3.40 (1.47,7.84) 5.57 (2.48, 12.52) 2.80 (1.49, 5.24) 2.02 (1.16, 3.51) January 2008, Vol 98, No. 1 American Journal of Public Health children with no history of foreign residence before their blood test. We cannot draw definitive conclusions with respect to locations and sources of lead expo- sure in our sample, particularly because blood lead levels reflect recent exposures as well as past and chronic exposures to lead. The mean biological life of lead in the blood is approximately 30 days; however, after a period of lead exposure, lead is stored in the bone and other tissues in the body and may be released into the blood at a later point in time.21.22 Therefore, in cross-sectional studies such as ours, blood lead levels alone are not an accurate means of determining the timing and total duration of lead exposures among children. Despite this limitation, we address plausible explanations for and implications of our finding of increased risk associated with spending time in a foreign country. In the United States, patterns of childhood lead poisoning differ according to socioeco- nomic position. Similar patterns probably exist across nations, such that in economically depressed countries, where environmental lead is less regulated, children have more op- portunities for exposure to lead. In our study, the most common foreign countries of birth among children with lead poisoning were Q ✩ ☆ Mexico and countries of the Caribbean and South Asia. Results of an earlier study of ele- vated blood lead levels among refugee chil- dren showed higher prevalence rates among children born in Asian, African, Central American, and Caribbean countries than among children born in the United States or among refugee children from northern Eura- sia (e.g., Ukraine, Russia, Germany)." In many of these countries, children are likely to be exposed to widespread environmental sources of lead (e.g., leaded gasoline, indus- trial emissions), 23-26 Immigrant families' use of lead-contaminated products while living abroad or their contin ued use of these products in the United States can also increase their children's risk of expo- sure to lead. Such products include lead-glazed pottery and cookware and lead-contaminated foods, spices, medicines, and cosmetics. 27-32 We asked parents about their use of a limited number of known lead-contaminated prod- ucts, and only a few reported using them. However, in our experience of providing envi ronmental and educational services for chil- dren with lead poisoning, use of potentially contaminated products and food is often un- derreported, at least partly because of our and families' inadequate knowledge of what products may contain lead. These potential exposure sources are more difficult to identify than lead paint hazards in current housing but may be particularly important in attempt- ing to understand the excess risk observed among immigrant children. Consistent with the literature, other risk factors for lead poisoning in this study in- cluded housing and child behavioral charac teristics - One finding that is unique to our study is the strong association between building size and lead poisoning: New York City children living in buildings with fewer than 3 units were 5 times more likely to have elevated blood lead levels than were children living in larger buildings in the same neighbor- hoods. The city's local laws addressing lead paint hazards in the homes of young children, in existence since 1982, focus on hazards in large buildings and may be responsible for better control of lead paint hazards in these buildings than in smaller buildings. Additional research in jurisdictions without similar laws can help clarify this relationship. Regardless, Tehranifar et al. Peer Reviewed Research and Practice | 95 = Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Q ||| Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 3 4 ---- ww 5 6 5/8 84% + our findings suggest that building size may be a useful indicator for targeting primary pre- vention efforts in New York City. Limitations and Possible Biases Our results may have overestimated the increased risk of lead poisoning among immi- grant children if other unmeasured factors re- lated to immigration status increased or inter- acted with exposures to current lead paint hazards. Such factors could include cultural practices (eg., eating on the floor, leading to an increased possibility of ingesting leaded paint or dust) and preexisting nutritional deficiencies (leading to increased absorption of lead). We included 5 variables relating to chil- dren's household risk of exposure to lead paint hazards (presence of peeling paint in the home, children's mouthing behaviors and in- gestion of nonfood items, and building age and size), none of which reduced the magni- tude of the associations of lead poisoning with foreign birthplace and time elapsed since most recent foreign residence. Thus, although lead paint-based hazards are the most com- monly identified source of childhood lead poi- soning in New York City, our results point to the possibility that lead paint and its dust in current housing do not fully explain the in- creased risk of lead poisoning among immi- grant children. Our assessments of the presence of peeling paint in homes relied on parents' reports of deteriorated paint, and self-reports of such in formation may be less reliable than data col- lected through visual inspections by trained professionals. Potential measurement error may have resulted in inadequate control for the risk of lead paint hazards. Furthermore, LPPP inspectors visited the homes of children in the case group but not the control group. Thus, differential misclassification of exposure to lead paint hazards may have occurred if these visits occurred before the study inter- view, with the result that case families were more likely to know about or recall the pres- ence of deteriorated paint in their homes. Fi- nancial restrictions precluded the use of more-reliable methods (eg., visual inspection by a trained professional, collection of dust samples) of assessing this risk factor. We cannot rule out the possibility that the biases just described led to overestimation or underestimation of the adjusted relationship between foreign birthplace or foreign resi- dence and risk of lead poisoning. However, our use of a detailed questionnaire and our inclusion of building age as an additional con- trol variable may have mitigated the likeli- hood of any bias. To account for potential confounders, we used a pair-matched design and further ad- justed for measured confounding variables in our analyses. We chose to match children by their current neighborhood of residence be- cause building characteristics, blood lead lev- els, and likelihood of undergoing testing vary across New York City neighborhoods How- ever, because New York City neighborhoods exhibit considerable clustering according to sociodemographic characteristics (e.g., immi- gration, race/ethnicity, income), this matching strategy may have underestimated the rela- tionship of lead poisoning to these factors. Overmatching was evidenced by the fact that case and control children resembled each other in terms of socioeconomic status and racial/ethnic background, resulting in our in- ability to find any significant associations. However, these variables have generally been found to be related to an increased risk of lead poisoning.34-36 Despite this restriction, our results showed strong associations be- tween lead poisoning and foreign birthplace and recent immigration. Finally, we cannot rule out that our find- ings were influenced by selection bias result- ing from differences between children receiv ing and not receiving a blood lead test and between children participating and not partic- ipating in the study. Blood lead testing rates in New York City are high, with nearly 88% of children reaching their 3rd birthday in 2004 having undergone at least 1 blood test. New York City children 3 years or older are generally tested on the basis of their level of risk (as ascertained by their health care pro- viders) or in accordance with school require ments. Different results in an unknown direc tion may have been obtained if tested and untested children had differed in regard to important characteristics related to both their place of birth and residence and their blood lead levels. To lessen the probability of this bias, we matched children on age, neighbor- hood, and test date, allowing for a degree of 96 | Research and Practice | Peer Reviewed | Tehranifar et al. comparability in terms of the likelihood of undergoing a blood test. In addition, 78% of eligible case children but only 49% of eligible control children were enrolled in this study. A multilingual precall letter, multilingual interviewers, and repeated contacts with potential study partici- pants were used to protect against differences in participation rates between the 2 groups. Public Health Implications Our findings are consistent with those of other published studies reporting high per- centages of immigrants among children with lead poisoning. Taken together, these re- sults suggest a need for considering recent immigration as a risk factor for childhood. lead poisoning and allocating resources to identify and remove lead exposure sources in the immigrant communities at greatest risk. Public health practitioners and health care providers can contribute to these efforts by learning about common sources of exposure in the populations they serve and educating families of young children about ways to re- duce exposures. Although limited, our knowl edge of lead-contaminated foods, medicines, and other products is growing (and informa- tion and educational materials can be ob- tained through the Centers for Disease Con- trol and Prevention [http://www.cdc.gov/ nceh/lead/faq/FAQs.htm] and through state and local health departments). Equally important is educating immigrant families about methods of reducing children's exposure to lead paint hazards, given that paint may be an unfamiliar source of exposure for many immigrant communities. Finally, blood. lead testing of children who have recently en- tered the United States from foreign countries in which lead exposure rates are high can facili tate early identification of children exposed to lead and promote timely provision of environ- mental or medical interventions. About the Authors At the time of the study, the authors were with the Environ- mental Health Division of the New York City Department of Health and Mental Hygiene Requests for reprints should be sent to Parisa Tehrani- far, DrPH, Department of Epidemiology Mailman School of Public Health, Columbia University, 622 West 168th St, New York, NY 10032 (e-mail: [email protected]). This article was accepted February 26, 2007. American Journal of Public Health | January 2008, Vol 98, No. 1 •= Paused :
andRisk of Childhood Lead Poisoning: Findings from a Case-ControlStudy of New York Children.
American Journal of Public Health, 98(1), 92-97.
- For The Following Questions Please Refer To The Posted Article Tehranifar Et Al 2008 Immigration And Risk Of Chil 1 (370.26 KiB) Viewed 74 times
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 2 /8 | - 84% + B below from a venous or capillary sample and no previous elevated blood lead level. All blood samples were analyzed by labo- ratories certified by the New York State De- partment of Health. We obtained information on blood test date, lead level, child's date of birth, and parent or guardian's name and ad- dress from the blood lead registry. We gath- ered information on building age and size from databases maintained by the city of New York. A 30-minute structured telephone ques- tionnaire, administered to the parents or guardians of both case children and controls, was used to collect data on family income, potential sources of lead paint exposure in the home, and child's gender, race/ethnicity, mouthing behaviors, country of birth, and for eign residential history. We pilot tested the questionnaire via a focus group and pretest interviews and translated it into Spanish. ||| 1 2 3 Data Collection A multilingual introductory letter was mailed to the parents or guardians of 1058 eligible children. At least 4 attempts were subsequently made to contact the parents or guardians by telephone. We were more successful in contact- ing parents or guardians of the 308 potential case children (78%) than the 750 potential control children (49%), primarily because, for 2 reasons, we had better contact information for the former. First, case children's blood test results were reported to LPPP by both labora- tories and health care providers, whereas con- trol children's results were reported by labora tories only. Second, LPPP visited the homes of case children as part of care coordination and environmental assessment activities. More families in the control group than in the case group refused participation in the study (22% of the control families vs 12% of the case families contacted). We were unable to assess differences in participation according to foreign birthplace or residence because this information was not available for any of the control families, nor was it available for some of the case families that were not interviewed. We interviewed 211 parents or guardians of case children and 288 parents or guardians of controls. Six case children were excluded because their matched controls were not suc- cessfully enrolled into the study. In addition, 2 case children were excluded as a result of missing or incorrect birth dates. When there were multiple controls, the first successfully in- terviewed control was selected. The final sam- ple included 203 case-control pairs. As a means of maximizing participation among immigrant families that spoke lan- guages other than English, trained bilingual English and Spanish speakers conducted the interviews, and professional interpreters made the initial contacts with the households and administered the questionnaire if needed. The questionnaire was administered in Spanish to 33% (135) of participants and in Haitian Creole to a single participant in the final sample. The distribution of inter- view language was similar for case and con- trol children. Initially, interviewers asked to speak with the mother or father of the child; if the mother and father were not available, they asked, in descending order, for the child's guardian or any available adult relative. For the most part (89% of case children and 95% of control children), parents rather than other adults were interviewed. Measures Children were categorized as US born or for- eign born (children born in Puerto Rico were classified as foreign born). Parents or guardians were asked whether their child had lived out- side of the United States for at least 2 months; if they answered yes, they were asked to pro- vide dates of residence. To capture the timing of potential lead exposures in foreign countries, we calculated the length of time between the most recent date of foreign residence and the time of blood lead testing among children who had lived outside of the United States. Because only a small percentage of chil- dren had lived in a foreign country (17%) and exploratory analyses suggested that the risk of lead poisoning dropped sharply approxi- mately 6 months after most recent foreign residence, we used a simple classification scheme in which 6 months was designated as the cutoff point. That is, children were grouped in the following categories for the final analysis: (1) lived in a foreign country less than 6 months before their blood test, (2) lived in a foreign country 6 or more months before their blood test, and (3) did not live in a foreign country before their blood test. January 2008, Vol 98, No. 1 | American Journal of Public Health Q ✩ A major pathway for lead exposure among urban children is ingestion of deteriorated leaded paint and household dust that have set- tled on their hands or on objects. Lead-based paint hazards are more likely to be present in buildings constructed before 1950, when use of lead paint was more widespread than in later years. We categorized time of building con- struction as before 1950 versus 1950 or after. Parents were asked about the presence of peeling or chipping paint in their home and were asked whether their children put their fingers in their mouths or ate any nonfood items (e.g., toys, crayons). Both of these measures were dichotomized for the pur- poses of our analysis. Because New York City laws aimed at preventing children's ex- posure to lead paint hazards in housing apply to buildings with 3 or more dwelling units, larger buildings may contain fewer lead paint hazards. Thus, we expected lead. poisoning to be inversely associated with building size, categorized according to number of dwelling units (fewer than 3 vs 3 or more). Finally, we asked about use of imported products known to contain lead (e.g., traditional Mexican folk remedies such as azarcon and greta). Data Analysis We used odds ratios (ORs) derived from conditional logistic regression analyses to ex- amine bivariate associations of lead poisoning status with foreign birthplace, time elapsed since most recent foreign residence, and other potential risk factors. In addition, we exam- ined associations between lead poisoning sta- tus and foreign birthplace and time since most recent foreign residence in separate con- ditional logistic regression models that ad- justed for potential confounders exhibiting statistically significant bivariate associations with lead poisoning status. RESULTS The mean age of the sample was 3 years (range: 0-17 years), and 75% of the children were younger than 5 years. As expected, foreign birthplace and time elapsed since most recent foreign residence were strongly correlated (Spearman rank correlation-0.80, P<.001); thus, we were not able to examine Tehranifar et al. | Peer Reviewed | Research and Practice | 93 = Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... ||| Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C.... 3 / 8 - 84% + their independent effects in the same model. Table 1 presents the characteristics of case and control children and the bivariate associ ations between each independent variable and lead poisoning status. 1 2 3 Twenty percent of case children were born outside of the United States, as compared. TABLE 1-Sociodemographic and Risk Characteristics of Participants, by Case-Control Status: New York City, 2002-2003 Place of birth Foreign country United States Time since most recent foreign residence <6 months 26 months Did not live in a foreign country Race/ethnicity Hispanic Asian Other African American Gender Boy Girl Family income level Below 150% of poverty level Above 150% of poverty level Peeling paint in home Yes No Time building of residence constructed Before 1950 1950 or after No. of units in building <3 23 Frequency at which child puts fingers in mouth Sometimes/always with 4% of control children (Table 1). Simi- larly, case children were more likely than were control children (27% vs 8%) to have lived outside of the United States for at least 2 months. In addition to 50 foreign-born chil- dren, 22 US-born children (15 case children, 7 control children) had lived in a foreign Never/rarely Child eats nonfood items Yes No Case Group, No. (%) 41 (20.1) 162 (79.8) 33 (164) 21 (10.4) 147 (73.1) 82 (40.4) 24 (11.9) 19 (9.4) 78 (384) 84 (41.4) 119 (586) 122 (739) 43 (26.1) 88 (44.0) 112 (56.0) 185 (91.6) 17 (8.4) 67 (33.0) 136 (67.0) 153 (76.1) 48 (23.9) 115 (56.7) 88 (43.3) Control Group No. (%) 94 | Research and Practice | Peer Reviewed | Tehranifar et al. 9(4.4) 194 (95.6) 4(2.0) 12 (5.9) 187 (92.1) 100 (49.2) 13 (6.4) 12 (5.9) 78 (38.4) 93 (45.8) 110 (542) 127 (89.8) 55 (302) 68 (33.5) 135 (66.5) 154 (76.6) 47 (234) 18 (8.9) 185 (91.1) 120 (59.1) 83 (40.9) 74 (36.5) 129 (63.5) Unadjusted OR (95% CI) 5.57 (2.49, 12.45) Reference 9.00 (3.16, 25.65) 2.26 (1.04,4.90) Reference 0.70 (0.42, 1.20) 2.18 (0.95,5.04) 1.65 (0.73, 3.72) Reference 0.84 (0.57,1.24) Reference 1.42 (0.84, 2.39) Reference 1.54 (1.03,2.30) Reference 3.23 (1.73.6.02) Reference 6.44 (3.19, 13.02) Reference 2.18 (1.39,3.40) Reference 2.32 (1.52, 3.53) Reference Note. OR-odds ratio; CI-confidence interval. The sample included 203 case-control pairs. "Foreign countries of birth included Haiti (n-12), Pakistan (n-7), the Dominican Republic (n-6), Mexico (n-5), Puerto Rico (n-3), India (n-2), Bangladesh (n-2), Liberia (n-2), Ghana (n-1), Macau (n-1), Israel (n-1), Russia (n-1), Yugoslavia (n-1), Jamaica (n-1), Trinidad and Tobago (n-1), El Salvador (n-1), Ecuador (n-1), Colombia (n-1), and Guyana (n-1). Time elapsed between most recent date of residence in a foreign country (for a minimum of 2 months) and date of blood lead testing. "Including White, American Indian, mixed racial background, and "don't know." Q ✩ country at some time before their blood test. Finally, 3 children (1 case child and 2 control children) had lived abroad for 2 months or longer more than once. There were differences in foreign countries of birth and residence between the case and control groups. Among case children, the main foreign countries of birth and residence. were Haiti, Pakistan, Mexico, and the Domini- can Republic; among control children, the pri- mary countries were Puerto Rico and the Do- minican Republic. There were no differences in use of imported products between the case group and the control group (data not shown). Both foreign birthplace and time since most recent foreign residence were strongly associated with lead poisoning status; children who had lived in a foreign country less than 6 months before their blood test were at par- ticularly elevated risk of lead poisoning rela- tive to US-born children with no foreign resi- dential history before their blood test (OR-9.0; 95% confidence interval [CI]-3.2, 25.7; Table 1). Building characteristics and child behavioral characteristics were associ- ated with lead poisoning status in the ex- pected directions. Case and control children were similar in terms of family income. Per- centages of children of African American and Hispanic descent were roughly similar in the 2 groups, although case children were dispro- portionately Asian. Table 2 summarizes the results of a series of conditional logistic regression models test- ing our hypothesis that birth and residence in a foreign country would be associated with an increased risk of lead poisoning. After ad- justment for housing and child behavioral risk factors, foreign-born children were 5 times more likely than were US-born children to be lead poisoned (OR-5.4; 95 % CI-2.2, 13.5; Table 2, model 1). A history of foreign residence was also strongly associated with lead poisoning status in adjusted models. Children living abroad for less than 6 months and for 6 or more months before their blood test, respectively, were 11 and 3 times more likely to to be lead poisoned than were US-born children who had not lived in a foreign country (Table 2, model 2). Reclassification of 3 con- trol children born in Puerto Rico as US bor strengthened the associations between American Journal of Public Health | January 2008, Vol 98, No. 1 = Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 4/8 | - 84% + E I 2 3 5 TABLE 2-Associations of Foreign Birthplace and Time Elapsed Since Most Recent Foreign Residence With Lead Poisoning Status, Adjusted for Building and Child Behavioral Characteristics: New York City, 2002-2003 Foreign born Time since most recent foreign residence <6 months before blood test 26 months before blood test Did not live in a foreign country (Ref) Building characteristics Peeling paint in home Residence in pre-1950 building Residence in building with fewer than 3 units Child behavioral characteristics Child puts fingers in mouth Child eats nonfood items foreign birth and residence and lead poison- ing status (data not shown). Other risk factors were also positively asso ciated with lead poisoning status in both mul- tivariate models. Residence in a building with fewer than 3 units (OR-5.6; 95% CI-2.5, 12.5; model 2) and residence in a building constructed before 1950 (OR-3.4; 95% CI-1.5, 7.8) were associated with an in- creased risk of lead poisoning. In addition, there remained a higher risk of lead poisoning among children who put their fingers in their mouth (OR=2.8; 95% CI-1.5, 5.2) and ate nonfood items (OR-2.0; 95% CI-1.2, 3.5) than among children who did not. DISCUSSION Model 1, OR (95% CI) 5.39 (2.15, 13.54) We found that, among children tested for lead poisoning, foreign-born children were 5 times more likely than were US-born children to have elevated blood lead levels after ad- justment for building characteristics and child behaviors. An even stronger association was found between lead poisoning and recent res- idence in a foreign country, with children liv- ing abroad for less than 6 months before their blood test being 11 times more likely to have elevated blood lead levels than US-born 1.18 (0.70, 1.98) 2.79 (1.29,6.03) 5.67 (259, 12.41) 2.24 (1.26,3.97) 2.10 (1.24, 3.56) Note. OR-odds ratio; C1-confidence interval 195 matched pairs. 193 matched pairs. "Time elapsed between most recent date of residence in a foreign country (for a minimum of 2 months) and date of blood lead testing. Model 2,"OR (95% CI) 10.94 (3.28, 36.46) 2.92 (1.12, 7.58) 1.00 1.17 (0.69, 2.00) 3.40 (1.47,7.84) 5.57 (2.48, 12.52) 2.80 (1.49, 5.24) 2.02 (1.16, 3.51) January 2008, Vol 98, No. 1 American Journal of Public Health children with no history of foreign residence before their blood test. We cannot draw definitive conclusions with respect to locations and sources of lead expo- sure in our sample, particularly because blood lead levels reflect recent exposures as well as past and chronic exposures to lead. The mean biological life of lead in the blood is approximately 30 days; however, after a period of lead exposure, lead is stored in the bone and other tissues in the body and may be released into the blood at a later point in time.21.22 Therefore, in cross-sectional studies such as ours, blood lead levels alone are not an accurate means of determining the timing and total duration of lead exposures among children. Despite this limitation, we address plausible explanations for and implications of our finding of increased risk associated with spending time in a foreign country. In the United States, patterns of childhood lead poisoning differ according to socioeco- nomic position. Similar patterns probably exist across nations, such that in economically depressed countries, where environmental lead is less regulated, children have more op- portunities for exposure to lead. In our study, the most common foreign countries of birth among children with lead poisoning were Q ✩ ☆ Mexico and countries of the Caribbean and South Asia. Results of an earlier study of ele- vated blood lead levels among refugee chil- dren showed higher prevalence rates among children born in Asian, African, Central American, and Caribbean countries than among children born in the United States or among refugee children from northern Eura- sia (e.g., Ukraine, Russia, Germany)." In many of these countries, children are likely to be exposed to widespread environmental sources of lead (e.g., leaded gasoline, indus- trial emissions), 23-26 Immigrant families' use of lead-contaminated products while living abroad or their contin ued use of these products in the United States can also increase their children's risk of expo- sure to lead. Such products include lead-glazed pottery and cookware and lead-contaminated foods, spices, medicines, and cosmetics. 27-32 We asked parents about their use of a limited number of known lead-contaminated prod- ucts, and only a few reported using them. However, in our experience of providing envi ronmental and educational services for chil- dren with lead poisoning, use of potentially contaminated products and food is often un- derreported, at least partly because of our and families' inadequate knowledge of what products may contain lead. These potential exposure sources are more difficult to identify than lead paint hazards in current housing but may be particularly important in attempt- ing to understand the excess risk observed among immigrant children. Consistent with the literature, other risk factors for lead poisoning in this study in- cluded housing and child behavioral charac teristics - One finding that is unique to our study is the strong association between building size and lead poisoning: New York City children living in buildings with fewer than 3 units were 5 times more likely to have elevated blood lead levels than were children living in larger buildings in the same neighbor- hoods. The city's local laws addressing lead paint hazards in the homes of young children, in existence since 1982, focus on hazards in large buildings and may be responsible for better control of lead paint hazards in these buildings than in smaller buildings. Additional research in jurisdictions without similar laws can help clarify this relationship. Regardless, Tehranifar et al. Peer Reviewed Research and Practice | 95 = Paused :
Bb Immigration and Risk of Childhox + → CO learn-us-east-1-prod-fleet01-xythos.content.blackboardcdn.com/blackboard.learn.xythos.prod/5a3199fc4282a/14901830?X-Blackboard-Exp... Q ||| Immigration and Risk of Childhood Lead Poisoning: Findings From a Case-C... 3 4 ---- ww 5 6 5/8 84% + our findings suggest that building size may be a useful indicator for targeting primary pre- vention efforts in New York City. Limitations and Possible Biases Our results may have overestimated the increased risk of lead poisoning among immi- grant children if other unmeasured factors re- lated to immigration status increased or inter- acted with exposures to current lead paint hazards. Such factors could include cultural practices (eg., eating on the floor, leading to an increased possibility of ingesting leaded paint or dust) and preexisting nutritional deficiencies (leading to increased absorption of lead). We included 5 variables relating to chil- dren's household risk of exposure to lead paint hazards (presence of peeling paint in the home, children's mouthing behaviors and in- gestion of nonfood items, and building age and size), none of which reduced the magni- tude of the associations of lead poisoning with foreign birthplace and time elapsed since most recent foreign residence. Thus, although lead paint-based hazards are the most com- monly identified source of childhood lead poi- soning in New York City, our results point to the possibility that lead paint and its dust in current housing do not fully explain the in- creased risk of lead poisoning among immi- grant children. Our assessments of the presence of peeling paint in homes relied on parents' reports of deteriorated paint, and self-reports of such in formation may be less reliable than data col- lected through visual inspections by trained professionals. Potential measurement error may have resulted in inadequate control for the risk of lead paint hazards. Furthermore, LPPP inspectors visited the homes of children in the case group but not the control group. Thus, differential misclassification of exposure to lead paint hazards may have occurred if these visits occurred before the study inter- view, with the result that case families were more likely to know about or recall the pres- ence of deteriorated paint in their homes. Fi- nancial restrictions precluded the use of more-reliable methods (eg., visual inspection by a trained professional, collection of dust samples) of assessing this risk factor. We cannot rule out the possibility that the biases just described led to overestimation or underestimation of the adjusted relationship between foreign birthplace or foreign resi- dence and risk of lead poisoning. However, our use of a detailed questionnaire and our inclusion of building age as an additional con- trol variable may have mitigated the likeli- hood of any bias. To account for potential confounders, we used a pair-matched design and further ad- justed for measured confounding variables in our analyses. We chose to match children by their current neighborhood of residence be- cause building characteristics, blood lead lev- els, and likelihood of undergoing testing vary across New York City neighborhoods How- ever, because New York City neighborhoods exhibit considerable clustering according to sociodemographic characteristics (e.g., immi- gration, race/ethnicity, income), this matching strategy may have underestimated the rela- tionship of lead poisoning to these factors. Overmatching was evidenced by the fact that case and control children resembled each other in terms of socioeconomic status and racial/ethnic background, resulting in our in- ability to find any significant associations. However, these variables have generally been found to be related to an increased risk of lead poisoning.34-36 Despite this restriction, our results showed strong associations be- tween lead poisoning and foreign birthplace and recent immigration. Finally, we cannot rule out that our find- ings were influenced by selection bias result- ing from differences between children receiv ing and not receiving a blood lead test and between children participating and not partic- ipating in the study. Blood lead testing rates in New York City are high, with nearly 88% of children reaching their 3rd birthday in 2004 having undergone at least 1 blood test. New York City children 3 years or older are generally tested on the basis of their level of risk (as ascertained by their health care pro- viders) or in accordance with school require ments. Different results in an unknown direc tion may have been obtained if tested and untested children had differed in regard to important characteristics related to both their place of birth and residence and their blood lead levels. To lessen the probability of this bias, we matched children on age, neighbor- hood, and test date, allowing for a degree of 96 | Research and Practice | Peer Reviewed | Tehranifar et al. comparability in terms of the likelihood of undergoing a blood test. In addition, 78% of eligible case children but only 49% of eligible control children were enrolled in this study. A multilingual precall letter, multilingual interviewers, and repeated contacts with potential study partici- pants were used to protect against differences in participation rates between the 2 groups. Public Health Implications Our findings are consistent with those of other published studies reporting high per- centages of immigrants among children with lead poisoning. Taken together, these re- sults suggest a need for considering recent immigration as a risk factor for childhood. lead poisoning and allocating resources to identify and remove lead exposure sources in the immigrant communities at greatest risk. Public health practitioners and health care providers can contribute to these efforts by learning about common sources of exposure in the populations they serve and educating families of young children about ways to re- duce exposures. Although limited, our knowl edge of lead-contaminated foods, medicines, and other products is growing (and informa- tion and educational materials can be ob- tained through the Centers for Disease Con- trol and Prevention [http://www.cdc.gov/ nceh/lead/faq/FAQs.htm] and through state and local health departments). Equally important is educating immigrant families about methods of reducing children's exposure to lead paint hazards, given that paint may be an unfamiliar source of exposure for many immigrant communities. Finally, blood. lead testing of children who have recently en- tered the United States from foreign countries in which lead exposure rates are high can facili tate early identification of children exposed to lead and promote timely provision of environ- mental or medical interventions. About the Authors At the time of the study, the authors were with the Environ- mental Health Division of the New York City Department of Health and Mental Hygiene Requests for reprints should be sent to Parisa Tehrani- far, DrPH, Department of Epidemiology Mailman School of Public Health, Columbia University, 622 West 168th St, New York, NY 10032 (e-mail: [email protected]). This article was accepted February 26, 2007. American Journal of Public Health | January 2008, Vol 98, No. 1 •= Paused :