A longitudinal cohort study examining determinants of overweight and obesity in adulthood.
Previous investigations have highlighted early determinants of high BMI in adulthood, which include high birthweight, high pre-pregnancy maternal and post-birth childhood BMI, early menarche, high dietary intake, low socio-economic status and low levels of physical activity. (10,11) Feeding and parenting styles have also been suggested as determinants. (12) Additionally, investigations have implicated air quality to be an important modulator of long-term weight status. (13) For example, a recent animal study in pregnant rats exposed to Beijing's unfiltered air showed evidence of weight gain, and offspring were heavier pre-and postnatally with accompanying lung inflammation, systemic oxidative stress and enhanced proinflammatory markers. (13) Daily inhalation rates tend to be higher in overweight and obese human adults, (14) which may result in a greater intake of air contaminants. However, it remains unclear whether chronic exposures to air pollutants early in life have any association with BMI or psychosocial well-being in adulthood. Chronic exposures to air contaminants can have significant long-term health consequences, especially in children due to their developmental stage, underdeveloped immune system, proportion of time spent outdoors, and relatively smaller and immature lungs. (15)
Using a 30-year exposure and health data initially collected from a cohort of children residing in Hamilton (Ontario, Canada) during 1978-1986 and from follow-up research in 2006, we aimed to assess childhood and life-course determinants of weight status in adulthood. Complete details of the original program are described elsewhere. (16) Specifically, the original study took place in four geographically-distinct neighbourhoods in Hamilton--a city with a legacy of air-quality issues due to the industrial nature of its economy. The city experienced an air-quality gradient across the four neighbourhoods, with East Lower (EL) bearing the highest levels of air pollution due to its proximity to the industrial core relative to Hamilton's other three neighbourhoods: West Lower (WL), East Upper (EU) and West Upper (WU) (Figure 1). (16,17) WU region experienced relatively lower levels of air pollution. (17) This natural arrangement provided a unique opportunity to study the role of early air-contaminant exposures on weight status in adulthood. Furthermore, a combination of macro- and micro-environmental factors were examined in relation to weight status in adulthood, using the Life Course Health Development (LCHD) model as a guiding framework. (18) The LCHD model conceptualizes these macro- and micro-environmental contexts along with mechanistic and regulatory processes into an overall integrative framework (Figure 2). Using a longitudinal design with participants from distinct neighbourhoods of Hamilton, this study also incorporates the time and place elements of LCHD to identify predictors of weight status in adulthood.
Our study objectives were to: 1) apply the LCHD framework to assess childhood and life course determinants of weight status in adulthood, 2) evaluate associations between adulthood weight status and geographical differences in air quality, and 3) assess gender-specific factors associated with weight status in adulthood.
Data from childhood were acquired from the original research program (1978-1986), which included a cohort of 3202 children residing in four neighbourhoods of Hamilton. (16,19) In 2006, 929 cohort members were successfully retraced, 598 consented to participate and 395 returned completed surveys (response rate 66%), forming the basis of data from adulthood. (19) Of the 395 participants, 80 were excluded due to missing childhood and/or adulthood weight and/or height information required to calculate BMI. This resulted in a final sample size of 315. The McMaster University Research Ethics Board approved the survey and study.
Body mass index categorization at childhood and adulthood
BMI was calculated using self-reported weight and height information as: weight (kg)/height ([m.sup.2]). Each participant's childhood BMI was converted into BMI-for-age percentile score according to WHO Growth Charts adapted for Canada for boys (2-19 years) and girls (2-19 years). (20) These BMI-for-age percentile scores were then used to categorize participants into weight categories: wasted (<3rd percentile), normal weight (3rd-<85th), overweight (85th-97th) and obese (>97th). At adulthood, the calculated BMI values were used to classify participants as underweight (BMI: <18.5), normal weight (18.5-24.9), overweight (25-29.9) or obese ([greater than or equal to] 30). Weight status at adulthood was our primary outcome variable.
Assessment of ambient air quality in childhood
To assess environmental exposures to air pollutants, each participant was assigned annual exposure values for total suspended particulates (TSP) in 1980/1981, 1983/1984 and 1985/ 1986; particulate matter with aerodynamic diameter <3.3 [micro]m ([PM.sub.3.3]) in 1980/1981, 1981/1982, 1983/1984 and 1985/1986; and sulphur dioxide (S[O.sub.2]) in 1978/1979, 1980/1981, 1981/1982 and 1983/1984. Exposure estimates were modelled using location of schools the children attended during 1978-1986, and a large network of monitoring sites in Hamilton (details published previously). (16,19) Second, exposure was assessed by dichotomizing participants based on whether they were exposed to each air pollutant above or at/below median per sampling period (Table 1). Finally, exposure indices were created for TSP, [PM.sub.3.3] and S[O.sub.2] to indicate frequency of exposure above or at/below median across all sampling periods. As environmental exposures to air pollutants have previously been linked to respiratory, (19) non-respiratory, (17,19) and inflammatory disorders, (21) these air-quality indicators were analyzed as independent variables.
Macro-environmental variables at childhood and adulthood
In addition to ambient air-quality assessments, data related to demographic and residential domains were included as part of participants' macro-environment in childhood. Demographic variables included age, gender and household income dichotomized as below versus at/above low-income cut-off. Residential variables comprised neighbourhoods of residence in Hamilton, and whether participants were exposed to second-hand smoke. These variables captured early macro-environmental effects, which may impact adulthood BMI.
Several macro-environmental variables from adulthood were also included. Four dichotomized variables captured the adulthood demographical data: education, current household income, ethnicity and housing tenure; whereas, occupational information included employment status, type of occupation, and length of occupational exposures to gas/dust/contaminants. Residence in Hamilton (always or not) was dichotomized under residential history domain. Residential indoor air quality was surveyed using four binary variables: use of air conditioner/humidifier/filter, type of heating, air-duct cleaning and cooking method. One variable measured duration of residence (< versus [greater than or equal to] 5 years) in buildings constructed pre-1950 as an exposure indicator to harmful chemicals, such as lead and asbestos, commonly used in construction material in that era. Lead and asbestos exposures have previously been linked to changes in inflammatory markers. (22,23) Indicators of health care accessibility were evaluated in terms of whether access to regular family doctor and whether additional health insurance beyond provincially-funded coverage were available. Marital status and parental record of asthma
or respiratory problems (at least one versus none) were evaluated as part of family environment. Finally, information regarding the social environment was gathered using social contact and group participation scales of Social Health Battery. (19) Further details are presented in Table 1.
Micro-environmental variables in childhood and adulthood
Variables comprising participants' micro-environment at childhood were derived from behavioural and physiological domains. Smoking one or more cigarettes per day for more than 6 months captured the behavioural context; whereas, physiological domain included childhood weight status and the following binary determinants: presence of asthma in childhood, persistent morning or day/night cough, persistent wheeze and chest illnesses/colds before age 2. Airway obstruction was assessed as ratio of forced expiratory volume in first second ([FEV.sub.1]) over forced vital capacity (FVC) at/below median.
At follow-up, behavioural context was further explored by surveying whether the participant: was a current smoker, was ever a daily smoker, engaged in regular alcohol consumption and performed physical exercise exceeding 30 minutes for [greater than or equal to] 3 days/ week. Weight status and a diagnosis of high blood pressure were measured as physiological variables. Adulthood psychological context was captured by several factors that gathered binary data on stressful life events, concern over air pollution, emotional distress (scoring [greater than or equal to] 4 on General Health Questionnaire), self-rated health, and general feelings about income, health and life. These early and life course determinants were analyzed as independent variables. Further details are presented in Table 1.
Data were analyzed using SPSS v20. Descriptive statistics of outcome (dependent) and explanatory (independent) variables were calculated and analyzed. Chi-squared ([[chi].sup.2]) tests were used to compare weight categories at adulthood with each of the categorical independent variables. Those variables meeting a threshold of p < 0.10 at this bivariate analysis were entered into a entry procedure. Our first model generated best predictors series of multivariate logistic regression models, using forward-stepwise of adulthood weight status in the entire sample. To examine relative roles of ambient air quality, exposure indices of TSP, [PM.sub.3.3] and S[O.sub.2] were forced into this "best-fit" model to generate a second "exposure" model. Interactions of interest were examined. Finally, multivariate logistic regression models were created to analyze gender-specific determinants of adulthood weight status by stratifying the sample on gender. For each variable in logistic regression models, one category was selected as reference and other categories were compared against this reference. Significance, odds ratios (OR), 95% confidence intervals (CI) and pseudo-[R.sup.2] values were noted.
In childhood (mean age, 8 years), 72% of participants enjoyed a healthy weight; this declined to 33% in adulthood (mean age, 36 years) in the same cohort. Bivariate analyses indicated significant differences between weight status in adulthood and the following macro-environmental variables: gender, neighbourhood of residence in Hamilton during childhood, current exposure to smoking at home/work, type of occupation, duration >5 years of occupational exposures to gas/dust/contaminants, and additional health insurance coverage ([chi square], p < 0.05). Likewise, significant micro-environmental determinants included childhood weight status, asthma, high blood pressure in adulthood, self-rated health, and negative feelings about income, health and life ([[chi].sup.2], p < 0.05) (Table 1). Childhood weight categories predicted BMI values in adulthood ([F.sub.2,312] = 39.3, p < 0.001) (See Supplementary Figure 1, in the ARTICLE TOOLS section on the journal site).
Predictors of overweight and obesity in adulthood
The "best-fit" model showed that being overweight in adulthood was predicted by male gender (OR = 2.88) and duration >5 years of occupational exposures to gas/dust/contaminants (OR = 1.93). Additionally, residence in WU (OR = 0.33), EU (OR = 0.37) and WL (OR = 0.40) neighbourhoods in Hamilton during childhood was negatively associated with adulthood overweight, versus living in EL neighbourhood. Obesity in adulthood was associated with male gender (OR = 2.65), self-rated health at adulthood (OR = 7.23) as well as duration >5 years of occupational exposures to gas/dust/contaminants (OR = 3.06). Being overweight (OR = 3.39) or obese (OR = 31.22) during childhood predicted obesity in adulthood (Table 2).
The "exposure" model explained additional variance relative to "best-fit" model, as assessed by an increase in pseudo-[R.sup.2]. Exposure indices of TSP, [PM.sub.3.3] and S[O.sub.2] were, however, non-significant and not retained in the "best-fit" model (Table 2).
Gender-specific predictors of adulthood overweight and obesity
Among females, adulthood overweight was negatively associated with childhood residence in WU (OR = 0.17), EU (OR = 0.18) and WL (OR = 0.11) neighbourhoods compared to EL, but positively predicted by childhood overweight (OR = 4.64) and obesity (OR = 16.08). In comparison, adulthood obesity in females was not linked to residential neighbourhoods, but rather, associated strongly with childhood weight categories [overweight (OR = 4.11); obese (OR = 22.53)]. Other variables predictive of adulthood obesity in females were self-rated health at adulthood (OR = 10.54), having no additional health insurance coverage (OR = 0.27), having high blood pressure at adulthood (OR = 4.99), and always residing in Hamilton region (OR = 4.11) (Table 3).
Adulthood overweight in males was predicted by having no additional health insurance coverage (OR = 2.70), and duration >5 years of occupational exposures to gas/dust/contaminants (OR = 2.57); the latter variable was also associated with male obesity in adulthood (OR = 2.79) (Table 3).
As per our objectives, the study revealed global and gender-specific determinants of adulthood weight gain in a cohort longitudinally followed for approximately 30 years. The associations of overweight and obesity in adulthood with childhood residential neighbourhoods and occupational contaminant exposures are novel findings of this study.
Male gender and prolonged occupational exposures to harmful contaminants were associated with adulthood overweight and obesity. This gender disparity of male overweight and obesity is consistent with Canadian statistics as well as the existing literature that shows higher male prevalence of overweight in high-income countries belonging to the Organisation for Economic Cooperation and Development group. (1,24) The association of overweight and obesity with occupational exposures has rarely been reported.
Suadicani et al. found higher prevalence of obesity among men with blood phenotype "O" who had occupational exposures to airborne pollutants for at least 5 years (25)--a finding consistent with our study. This effect, however, was not examined among those who were overweight. (25) Jerrett and colleagues recently found traffic sources of air pollution positively associated with elevated BMI in children. (26) Other studies link higher BMI with a range of occupational variables, such as working extended hours per week, (27) and transport and production types of employments, (28) which may exacerbate workplace exposures to airborne contaminants. Although, we found a higher prevalence of overweight and obesity in relation to "manual" type of employment in this cohort, it is unclear whether manual employment exacerbated occupational exposures or contributed to elevated BMI by other mechanisms. As obesity is generally considered a low-grade inflammatory state, (29) it is possible that occupational exposures to pollutants interact synergistically with environmental and genetic risk factors to influence BMI via inflammatory pathways.
Undoubtedly, adulthood overweight and obesity are multifaceted conditions determined by complex biological and socio-cultural factors; (24) though some variables may show weight-class-specific associations, as observed in this study. In addition to male gender and duration of occupational exposures, childhood residential regions in Hamilton were significantly associated with adulthood overweight; whereas, childhood weight status was predictive of adulthood obesity (with no regional associations). Specifically, overweight and obese children showed ~3 and ~31 times greater odds of obesity in adulthood respectively. This finding suggests a high heritability index of obesity, estimated at greater than 0.70--a value comparable with other polygenic disorders, such as autism (0.91) and schizophrenia (0.81). (30) As adulthood overweight was not predictable by childhood weight status, but rather associated with residential neighbourhoods in our models, these findings highlight a multifactorial aetiology for adulthood overweight. This may suggest genetic determinants to be less or just as deterministic of adulthood overweight as regional or residence-based characteristics. Accordingly, WU, EU and WL regions were all found to be associated with lesser odds of adulthood overweight in comparison with the EL region, which sustained the poorest air quality. This implicates some role of regional air quality in influencing adulthood overweight, even though our "exposure model" did not highlight significant associations of TSP, [PM.sub.3.3] or S[O.sub.2] exposures. It is likely that living in a certain Hamilton neighbourhood may influence weight status indirectly, by affecting physical activity levels, food choices or region-specific factors; or directly via air quality-related inflammatory mechanisms. Our results support the notion that adulthood obesity carries a stronger genetic disposition, whereas overweight appears to be relatively heterogeneous with genetic as well as possibly modifiable environmental contributors. A recent study strengthens this complex multifactorial notion as McConnell et al. reported a synergistic effect of early exposure to tobacco smoke and near-roadway pollution with childhood obesity, (31) which in turn is a strong determinant of adulthood weight gain.
Gender-specific analysis revealed distinct factors among males and females. Prolonged occupational exposures in males and childhood weight status in females consistently predicted adulthood overweight and obesity. Association of childhood weight status among females may reflect strong biological and socio-cultural factors. It is recognized that females tend to have greater fat mass, smaller height, and lighter overall body weight, which vary based on gender-specific physiological conditions, such as pregnancy, menopause and overall metabolism. (32) Hence, it is plausible to find early weight status, determined in part by biological differences in fat metabolism, distribution and storage, to be associated with adulthood BMI among females. In addition to finding residential influences of air quality, females permanently residing in Hamilton since childhood were also four times more likely to be obese. Whether this implies a greater gender-specific susceptibility warrants further research. Higher blood pressure was also associated with female obesity. Interestingly, obese females and overweight males were less likely to own additional health insurance beyond provincially-funded coverage. In this cohort, a lower proportion of obese worked a "professional" type of occupation, but instead held more "manual" occupations, which may explain poorer health coverage as it is more often offered in professional positions. Additionally, duration of occupational exposures as a prime determinant of BMI among males may also be associated with job-related nature and/or tasks performed. A recent health report from Montreal found higher workplace exposures among males to motor vehicle exhaust, petroleum fractions, polycyclic aromatic hydrocarbons as well as building material and abrasive dust. (33)
Finally, overweight and obese individuals commonly reported negative feelings about health, income and life. However, the only factor that entered into our model was self-rated health reported as "fair or poor" among the obese adults, especially obese females. Such negative self-appraisals and psychosocial burden of obesity are confirmatory, which negatively influence quality of life, satisfaction, and overall psychological health of affected individuals. (9)
CONCLUSIONS, LIMITATIONS AND FUTURE RECOMMENDATIONS
Adulthood overweight and obesity are associated with childhood and life-course factors, including residential and occupational contaminant exposures, in a gender-specific manner. Longitudinal nature and inclusion of a multitude of variables guided by LCHD model are major strengths of this study. Certain limitations are noteworthy. Self-reported data have limitations, importantly the recall bias related to the recollection of contaminant exposures. Absence of air-quality data past childhood restricted our examination of chronic air-pollution exposures and prevented analyses on whether improved air quality has a restorative health effect. Data on physical activity, food intake levels, and physiological inflammatory indicators were unavailable. Finally, our reconstructed cohort had a reduced sample size. This loss to follow-up, however, was demonstrated to be unrelated to the health outcomes, residential neighbourhoods, or particulate exposures. (34) Moreover, as our findings are in line with existing literature, we believe that the participant selection process occurred fairly randomly and is unlikely to influence our results. Future studies with larger samples are warranted to replicate the findings and address limitations.
(1.) World Health Organization: Global Health Observatory Data. Available at: http://www.who.int/gho/ncd/risk_factors/obesity_text/en/ (Accessed March 24, 2016).
(2.) Statistics Canada: Overweight and Obese Adults (Self-Reported), 2014. Available at: http://www.statcan.gc.ca/pub/82-625-x/2015001/article/14185-eng.htm#n1 (Accessed March 24, 2016).
(3.) Dutton DJ, McLaren L. Explained and unexplained regional variation in Canadian obesity prevalence. Obesity (Silver Spring) 2011; 19(7):1460-68. PMID: 21253004. doi: 10.1038/oby.2010.339.
(4.) Twells LK, Gregory DM, Reddigan J, Midodzi WK. Current and predicted prevalence of obesity in Canada: A trend analysis. CMAJ Open 2014; 2(1): E18-26. PMID: 25077121. doi: 10.9778/cmajo.20130016.
(5.) Guh DP, Zhang W, Bansback N, Amarsi Z, Birmingham CL, Anis AH. The incidence of co-morbidities related to obesity and overweight: A systematic review and meta-analysis. BMC Public Health 2009; 9:88. PMID: 19320986. doi: 10.1186/1471-2458-9-88.
(6.) Luo W, Morrison H, de Groh M, Waters C, DesMeules M, Jones-McLean E, et al. The burden of adult obesity in Canada. Chronic Dis Can 2007; 27(4): 135-44. PMID: 17623559.
(7.) McLaren L. Socioeconomic status and obesity. Epidemiol Rev 2007; 29:29-48. PMID: 17478442. doi: 10.1093/epirev/mxm001.
(8.) Reilly JJ, Kelly J. Long-term impact of overweight and obesity in childhood and adolescence on morbidity and premature mortality in adulthood: Systematic review. Int J Obes (Lond) 2011; 35(7):891-98. PMID: 20975725. doi: 10.1038/ijo.2010.222.
(9.) Carr D, Friedman MA. Is obesity stigmatizing? Body weight, perceived discrimination, and psychological well-being in the United States. J Health Soc Behav 2005; 46(3):244-59. PMID: 16259147. doi: 10.1177/00221465 0504600303.
(10.) Laitinen J, Power C, Jarvelin MR. Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity. Am J Clin Nutr 2001; 74(3):287-94. PMID: 11522550.
(11.) Shumei SG, Wei W, William CC, Alex FR. Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence. Am J Clin Nutr 2002; 76(3):653-58. PMID: 12198014.
(12.) Shloim N, Edelson LR, Martin N, Hetherington MM. Parenting styles, feeding styles, feeding practices, and weight status in 4-12 year-old children: A systematic review of the literature. Front Psychol 2015; 6:1849. PMID: 26696920. doi: 10.3389/fpsyg.2015.01849.
(13.) Wei Y, Zhang JJ, Li Z, Gow A, Chung KF, Hu M, et al. Chronic exposure to air pollution particles increases the risk of obesity and metabolic syndrome: Findings from a natural experiment in Beijing. FASEB J 2016; 30(6):2115-22. PMID: 26891735. doi: 10.1096/fj.201500142.
(14.) Brochu P, Bouchard M, Haddad S. Physiological daily inhalation rates for health risk assessment in overweight/obese children, adults, and elderly. Risk Anal 2014; 34(3):567-82. PMID: 24147611. doi: 10.1111/risa.12125.
(15.) Schwartz J. Air pollution and children's health. Pediatrics 2004; 113(4 Suppl):1037-43. PMID: 15060197.
(16.) Kerigan AT, Goldsmith CH, Pengelly LD. A three-year cohort study of the role of environmental factors in the respiratory health of children in Hamilton, Ontario. Epidemiologic survey design, methods, and description of cohort. Am Rev Respir Dis 1986; 133:987-93.
(17.) Barakat-Haddad C, Elliott SJ, Pengelly D. Health impacts of air pollution: A life course approach for examining predictors of respiratory health in adulthood. Ann Epidemiol 2012; 22(4):239-49. PMID: 22463842. doi: 10.1016/j.annepidem. 2012.02.010.
(18.) Halfon N, Hochstein M. Life course health development: An integrated framework for developing health, policy, and research. Milbank Q 2002; 80(3):433-79. PMID: 12233246. doi:10.1111/1468-0009.00019.
(19.) Barakat-Haddad C, Elliott SJ, Pengelly D. Does chronic exposure to air pollution in childhood impact long-term respiratory health? ProfGeogr 2012; 64(3):446-63. doi:10.1080/00330124.2011.609775.
(20.) Dietitians of Canada: WHO Growth Charts Set 1. 2016. Available at: http:// www.dietitians.ca/Dietitians-Views/Prenatal-and-Infant/WHO-Growth-Charts/ WHO-Growth-Charts-Set-1.aspx (Accessed March 26, 2016).
(21.) Barakat-Haddad C, Elliott SJ, Pengelly D. Childhood exposure to air pollution as a potential contributor of chronic non-respiratory inflammatory disorders: A longitudinal prospective cohort study in Hamilton, Canada. J Environ Protect 2013; 4:779-88. doi: 10.4236/jep.2013.48091.
(22.) Sirivarasai J, Wananukul W, Kaojarern S, Chanprasertyothin S, Thongmung N, Ratanachaiwong W, et al. Association between inflammatory marker, environmental lead exposure, and glutathione S-transferase gene. Biomed Res Int 2013; 2013:1-6. PMID: 23484121. doi: 10.1155/2013/474963.
(23.) Haegens A, van der Vliet A, Butnor KJ, Heintz N, Taatjes D, Hemenway D, et al. Asbestos-induced lung inflammation and epithelial cell proliferation are altered in myeloperoxidase-null mice. Cancer Res 2005; 65(21):9670-77. PMID: 16266986. doi: 10.1158/0008-5472.CAN-05-1751.
(24.) Kanter R, Caballero B. Global gender disparities in obesity: A review. Adv Nutr 2012; 3(4):491-98. PMID: 22797984. doi: 10.3945/an.112.002063.
(25.) Suadicani P, Hein HO, Gyntelberg F. Airborne occupational exposure, ABO phenotype, and risk of obesity. Int J Obes (Land) 2005; 29(6):689-96. doi: 10. 1038/sj.ijo.0802956.
(26.) Jerrett M, McConnell R, Wolch J, Chang R, Lam C, Dunton G, et al. Traffic-related air pollution and obesity formation in children: A longitudinal, multilevel analysis. Environ Health 2014; 13:49. PMID: 24913018. doi: 10. 1186/1476-069X-13-49.
(27.) Shields M. Long working hours and health. Health Rep 1999; 11(2):33-48. PMID: 10618741.
(28.) Allman-Farinelli MA, Chey T, Merom D, Bauman AE. Occupational risk of overweight and obesity: An analysis of the Australian Health Survey. J Occup Med Toxicol 2010; 5:14. PMID: 20550716. doi:10.1186/1745-6673-5-14.
(29.) Johnson AR, Milner JJ, Makowski L. The inflammation highway: Metabolism accelerates inflammatory traffic in obesity. Immunol Rev 2012; 249(1):218-38. PMID: 22889225. doi: 10.1111/j.1600-065X.2012.01151.x.
(30.) Walley AJ, Blakemore AI, Froguel P. Genetics of obesity and the prediction of risk for health. Hum Mol Genet 2006; 15(2 Suppl):R124-30. PMID: 16987875. doi: 10.1093/hmg/ddl215.
(31.) McConnell R, Shen E, Gilliland FD, Jerrett M, Wolch J, Chang CC, et al. A longitudinal cohort study of body mass index and childhood exposure to secondhand tobacco smoke and air pollution: The Southern California Children's Health Study. Environ Health Perspect 2015; 123(4):360-66. PMID: 25389275.
(32.) Lovejoy JC, Sainsbury A; Stock Conference 2008 Working Group. Sex differences in obesity and the regulation of energy homeostasis. Obes Rev 2009; 10(2):154-67. PMID: 19021872. doi: 10.1111/j.1467-789X.2008. 00529.x.
(33.) Labreche F, Lacourt A, Lavoue J. Occupational Exposure to Chemical and Physical Contaminants: Sex-Differentiated Analysis. 2015. Available at: http://www.irsst. qc.ca/media/documents/PubIRSST/R-895.pdf (Accessed April 1, 2016).
(34.) Barakat-Haddad CP, Elliott SJ, Eyles J, Pengelly D. Predictors of locating children participants in epidemiological studies 20 years after last contact: Internet resources and longitudinal research. Eur J Epidemiol 2009; 24(8): 397-405. PMID: 19562490. doi: 10.1007/s10654-009-9364-5.
Received: July 8, 2016
Accepted: November 18, 2016
Caroline Barakat-Haddad, PhD,  Usman Saeed, (Hon) BSc,  Susan Elliott, PhD 
. Faculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, ON
. Institute of Medical Science, Faculty of Medicine, University of Toronto, Toronto, ON
. Department of Geography and Environmental Management, University of Waterloo, Waterloo, ON
Correspondence: Caroline Barakat-Haddad, PhD, Faculty of Health Sciences, UOIT, 2000 Simcoe Street North, Oshawa, ON L1H 7K4, Tel: 905-721-8668, ext. 2173,
Acknowledgement: We thank all the participants for their time and involvement in this study.
Conflict of Interest: None to declare.
Caption: Figure 1. Geographical map of the study area showing the four residential neighbourhoods in Hamilton, Ontario (Source: Barakat-Haddad et al. (17))
Caption: Figure 2. Graphical representation of the Life Course Health Development (LCHD) framework as conceptualized in this study (Modified from: Halfon and Hochstein (18)). Macro-environmental context includes health determinants related to demographic, residential, environmental, occupational, health care, family and social environments (7); whereas, micro-environmental context includes biobehavioural determinants \associated with functioning of behavioural, physiological and psychological systems, such as the immune and respiratory systems (2A). Mechanistic and regulatory processes depict developmentally programmed processes that can modulate micro-environmental context (2B). Macro-environmental factors interact with each other and with micro-environmental factors--which are in turn influenced by mechanistic and regulatory processes. These dynamic and multi-contextual interactions are modulated by environmental exposures and determine health outcomes over the life course (3).
Table 1. Distribution and prevalence of variables assessed at childhood and adulthood in this longitudinal cohort study Weight status at adulthood Variables Normal Overweight weight (n = 117) (n = 103) Macro-environmental variables in childhood and adulthood Demographic Gender, % male *** 37 62 Mean age in childhood (years)([dagger]) 8 8 Household income in childhood 14 11 (% below low-income cut-off)([dagger]) Mean age in adulthood (years) 36 36 Current household income 2 3 (% below low-income cut-off) Education (% completed high school) 99 99 Ethnicity (% Canadian) 65 56 Housing tenure (% homeowners) 85 89 Residential Neighbourhood of residence in Hamilton (%) *,([dagger]) East Lower 14 32 West Lower 19 20 East Upper 30 22 West Upper 37 26 Second-hand smoking in childhood 54 43 (% exposed)([dagger]) Always resided in Hamilton (%) 57 55 Resided in property built before 1950 19 29 (% for [greater than or equal to] 5 years) Residential exposure to gas/dust/ 91 92 contaminants (%) Type of heating (% gas/oil) 96 97 Frequency of air-duct cleaning 70 69 (% rarely/do not remember) Type of cooking method (% gas) 38 35 Residential usage of air conditioner (%) 89 82 Residential usage of air humidifier (%) 52 41 Residential usage of air filter (%) 41 33 Environmental Total suspended particulate (TSP) (% >median)([dagger]) 1980/1981 55 50 1983/1984 55 49 1985/1986 58 50 1980-1986 37 32 Particulate matter under 3.3 [micro]m diameter (PM3 3) (% >median)([dagger]) 1980/1981 59 49 1981/1982 51 46 1983/1984 41 44 1985/1986 45 43 1980-1986 46 36 Sulphur dioxide (SO2) (% >median) ([dagger]) 1978/1979 49 42 1980/1981 51 49 1981/1982 47 54 1983/1984 57 41 1978-1984 40 29 Current exposure to second-hand smoking 39 38 at home/work (%) * Occupational Employment (% full time) 71 72 Type of occupation (%) * Clerical 13 7 Manual 16 20 Managerial 17 13 Professional 54 55 Others 0 5 Occupational exposures to gas/dust/ 42 60 contaminants (% >5 years) ** Health care Access to regular family doctor (%) 93 89 Additional health insurance coverage 74 79 (%) ** Family Marital status (% married/common-law) 82 74 Parental record of asthma or 28 27 respiratory problems (%) Social Social contact subscale (% score of 14 9 [less than or equal to] 5) Group participation subscale (% score of 65 74 [less than or equal to] 4) Micro-environmental variables in childhood and adulthood Behavioural Experienced smoking in childhood 16 15 (% [greater than or equal to] 1 cigarette/day for >6 months)([dagger]) Current smoker (%) 19 25 Ever smoked on a daily basis (%) 28 33 Regular alcohol consumption (%) 10 15 Physical activity 46 46 (% [greater than or equal to] 3 times/week for >30 minutes) Physiological Asthma in childhood (%) *,([dagger]) 4 15 Persistent morning cough in childhood 9 8 (%)([dagger]) Persistent day/night cough in childhood 11 8 (%)([dagger]) Persistent wheeze in childhood 9 23 (%)([dagger]) Chest illnesses/colds before age 2 4 7 (% pneumonia or bronchitis)([dagger]) Index for pulmonary fraction 19 23 (% always above the median)([dagger]) Weight status in childhood (%) ***, ([dagger]) Normal weight 87 75 Overweight 11 19 Obese 2 6 High blood pressure in adulthood 6 7 (% diagnosed) ** Psychological Self-rated health (% fair/poor) *** 6 10 Concern over air pollution 54 50 (% moderate or extreme) Emotional distress (% score 6 14 [greater than or equal to] 4 on GHQ scale) Stressful life events 7 9 (% [greater than or equal to] 2) Feelings about income (% negative) * 7 11 Feelings about health (% negative) *** 0 8 Feelings about life (% negative) * 1 2 Weight Total status at sample adulthood (n = 315) Variables Obese (n = 95) Macro-environmental variables in childhood and adulthood Demographic Gender, % male *** 56 52 Mean age in childhood (years)([dagger]) 8 8 Household income in childhood 17 14 (% below low-income cut-off)([dagger]) Mean age in adulthood (years) 36 36 Current household income 8 4 (% below low-income cut-off) Education (% completed high school) 98 99 Ethnicity (% Canadian) 67 62 Housing tenure (% homeowners) 81 85 Residential Neighbourhood of residence in Hamilton (%) *,([dagger]) East Lower 17 22 West Lower 18 19 East Upper 35 29 West Upper 31 31 Second-hand smoking in childhood 46 48 (% exposed)([dagger]) Always resided in Hamilton (%) 64 58 Resided in property built before 1950 27 25 (% for [greater than or equal to] 5 years) Residential exposure to gas/dust/ 94 92 contaminants (%) Type of heating (% gas/oil) 97 96 Frequency of air-duct cleaning 69 69 (% rarely/do not remember) Type of cooking method (% gas) 38 37 Residential usage of air conditioner (%) 91 87 Residential usage of air humidifier (%) 36 43 Residential usage of air filter (%) 32 35 Environmental Total suspended particulate (TSP) (% >median)([dagger]) 1980/1981 48 51 1983/1984 58 54 1985/1986 50 53 1980-1986 33 34 Particulate matter under 3.3 [micro]m diameter (PM3 3) (% >median)([dagger]) 1980/1981 47 52 1981/1982 49 48 1983/1984 53 46 1985/1986 51 46 1980-1986 37 40 Sulphur dioxide (SO2) (% >median) ([dagger]) 1978/1979 46 45 1980/1981 49 50 1981/1982 48 50 1983/1984 50 49 1978-1984 30 33 Current exposure to second-hand smoking 55 44 at home/work (%) * Occupational Employment (% full time) 74 72 Type of occupation (%) * Clerical 12 10 Manual 28 21 Managerial 14 14 Professional 38 50 Others 8 4 Occupational exposures to gas/dust/ 67 56 contaminants (% >5 years) ** Health care Access to regular family doctor (%) 94 92 Additional health insurance coverage 60 71 (%) ** Family Marital status (% married/common-law) 71 76 Parental record of asthma or 39 31 respiratory problems (%) Social Social contact subscale (% score of 12 11 [less than or equal to] 5) Group participation subscale (% score of 61 67 [less than or equal to] 4) Micro-environmental variables in childhood and adulthood Behavioural Experienced smoking in childhood 15 15 (% [greater than or equal to] 1 cigarette/day for >6 months)([dagger]) Current smoker (%) 20 21 Ever smoked on a daily basis (%) 41 34 Regular alcohol consumption (%) 9 11 Physical activity 55 49 (% [greater than or equal to] 3 times/week for >30 minutes) Physiological Asthma in childhood (%) *,([dagger]) 13 10 Persistent morning cough in childhood 9 9 (%)([dagger]) Persistent day/night cough in childhood 18 12 (%)([dagger]) Persistent wheeze in childhood 14 16 (%)([dagger]) Chest illnesses/colds before age 2 3 5 (% pneumonia or bronchitis)([dagger]) Index for pulmonary fraction 38 26 (% always above the median)([dagger]) Weight status in childhood (%) ***, ([dagger]) Normal weight 51 72 Overweight 24 18 Obese 25 10 High blood pressure in adulthood 20 10 (% diagnosed) ** Psychological Self-rated health (% fair/poor) *** 29 15 Concern over air pollution 53 52 (% moderate or extreme) Emotional distress (% score 16 12 [greater than or equal to] 4 on GHQ scale) Stressful life events 14 10 (% [greater than or equal to] 2) Feelings about income (% negative) * 20 12 Feelings about health (% negative) *** 14 7 Feelings about life (% negative) * 6 3 Note: TSP = total suspended particulates; [PM.sub.3.3] = particulate matter with aerodynamic diameter <3.3 [micro]m; S[O.sub.2] = sulphur dioxide; GHQ = general health questionnaire. * p < 0.05; ** p < 0.01; *** p < 0.001. ([dagger]) Data collected at childhood. Table 2. Multivariate logistic regression models showing the associations of overweight and obesity in adulthood with childhood and life-course determinants in the entire cohort Outcome Variable Classification Best-fit model ([dagger]) (reference) (n = 314) ([double dagger]) OR 95% CI Overweight Gender (female) Male 2.88 *** 1.61-5.16 in Residence in West Upper 0.33 ** 0.15-0.74 adulthood childhood East Upper 0.37 * 0.16-0.88 (East Lower) West Lower 0.40 * 0.17-0.99 Weight status Overweight 2.29 1.00-5.27 at childhood Obese 4.54 0.87-23.66 (normal) Self-rated Fair/poor 1.80 0.62-5.24 health (good) Occupational >5 years 1.93 * 1.09-3.43 exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Index for TSP >median -- -- exposure ([less than or equal to] median) Index for >median -- -- [PM.sub.3.3] exposure ([less than or equal to] median) Index for >median -- -- S[O.sub.2] exposure ([less than or equal to] median) Obese in Gender (female) Male 2.65 ** 1.36-5.16 adulthood Residence in West Upper 0.80 0.30-2.09 childhood East Upper 1.09 0.40-2.93 (East Upper) West Lower 0.67 0.23-1.98 Weight status Overweight 3.39 ** 1.41-8.16 at childhood Obese 31.22 *** 6.61-147.54 (normal) Self-rated Fair/poor 7.23 *** 2.61-20.05 health (good) Occupational >5 years 3.06 ** 1.57-5.99 exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Index for TSP >Median -- -- exposure ([less than or equal to] median) Index for >Median -- -- [PM.sub.3.3] exposure ([less than or equal to] median) Index for >Median -- -- S[O.sub.2] exposure ([less than or equal to] median) Pseudo [R.sup.2] (Cox & -- 0.285; 0.321; Snell; Nagelkerke; McFadden) 0.154 Outcome Variable Classification Exposure model ([dagger]) (reference) (n = 312) ([section]) OR 95% CI Overweight Gender (female) Male 2.72 ** 1.50-4.91 in Residence in West Upper 0.31 ** 0.14-0.73 adulthood childhood East Upper 0.34 * 0.14-0.81 (East Lower) West Lower 0.34 * 0.14-0.86 Weight status Overweight 2.14 0.92-5.01 at childhood Obese 4.92 0.92-26.10 (normal) Self-rated Fair/poor 1.69 0.58-4.95 health (good) Occupational >5 years 1.99 * 1.11-3.56 exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Index for TSP >median 1.28 0.50-3.24 exposure ([less than or equal to] median) Index for >median 0.59 0.25-1.42 [PM.sub.3.3] exposure ([less than or equal to] median) Index for >median 0.74 0.36-1.50 S[O.sub.2] exposure ([less than or equal to] median) Obese in Gender (female) Male 2.46 ** 1.25-4.83 adulthood Residence in West Upper 0.82 0.30-2.20 childhood East Upper 0.99 0.36-2.73 (East Upper) West Lower 0.60 0.20-1.82 Weight status Overweight 3.18 * 1.31-7.75 at childhood Obese 34.58 *** 7.19-166.40 (normal) Self-rated Fair/poor 6.53 *** 2.34-18.24 health (good) Occupational >5 years 3.18 ** 1.62-6.24 exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Index for TSP >Median 1.37 0.48-3.93 exposure ([less than or equal to] median) Index for >Median 0.75 0.28-2.01 [PM.sub.3.3] exposure ([less than or equal to] median) Index for >Median 0.57 0.25-1.29 S[O.sub.2] exposure ([less than or equal to] median) Pseudo [R.sup.2] (Cox & -- 0.292; 0.329; Snell; Nagelkerke; McFadden) 0.158 Note: * p < 0.05; ** p < 0.01; *** p < 0.001. ([dagger]) Relative to "normal weight in adulthood" category. ([double dagger]) n for adulthood weight categories (best-fit model): normal = 102, overweight = 117, and obese = 95. ([section]) n for adulthood weight categories (exposure model): normal = 102, overweight = 116, and obese = 94. Table 3. Multivariate logistic regression models showing the gender-specific associations of overweight and obesity in adulthood with childhood and life course determinants Outcome Variable Classification Females (n = 139) ([dagger]) (reference) ([double dagger]) OR 95% CI Overweight Residence West Upper 0.17 ** 0.05-0.63 in in childhood East Upper 0.18 ** 0.05-0.62 adulthood (East Lower) West Lower 0.11 * 0.02-0.59 Weight status Overweight 4.64 * 1.39-15.51 at childhood (normal) Obese 16.08 * 1.63-158.37 Self-rated Fair/poor 4.41 0.79-24.72 health (good) Occupational >5 years -- -- exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Additional Yes 1.05 0.33-3.33 health insurance coverage (no) High blood Yes 0.36 0.05-2.83 pressure in adulthood (no) Always resided Yes 1.03 0.40-2.69 in Hamilton (no) Obese in Residence in West Upper 0.86 0.14-5.18 adulthood childhood East Upper 0.98 0.17-5.49 (East Lower) West Lower 1.65 0.23-12.19 Weight Overweight 4.11 * 1.14-14.87 status at childhood (normal) Obese 22.53 ** 2.40-211.23 Self-rated Fair/poor 10.54 ** 1.93-57.60 health (good) Occupational >5 years -- -- exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Additional Yes 0.27 * 0.09-0.83 health insurance coverage (no) High blood Yes 4.99 * 1.06-23.43 pressure in adulthood (no) Always resided Yes 4.11 * 1.26-13.46 in Hamilton (no) Pseudo [R.sup.2] (Cox & Snell; 0.417; 0.471; 0.249 Nagelkerke; McFadden) Outcome Variable Classification Males (n= 163) ([dagger]) (reference) ([section]) OR 95% CI Overweight Residence West Upper in in childhood East Upper -- -- adulthood (East Lower) West Lower -- -- Weight status Overweight -- -- at childhood (normal) Obese -- -- Self-rated Fair/poor -- -- health (good) Occupational >5 years 2.57 * 1.13-5.86 exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Additional Yes 2.70 * 1.11-6.53 health insurance coverage (no) High blood Yes -- -- pressure in adulthood (no) Always resided Yes -- -- in Hamilton (no) Obese in Residence in West Upper -- -- adulthood childhood East Upper -- -- (East Lower) West Lower -- -- Weight Overweight -- -- status at childhood (normal) Obese -- -- Self-rated Fair/poor -- -- health (good) Occupational >5 years 2.79 * 1.17-6.67 exposures to gas/dust/ contaminants ([less than or equal to] 5 years) Additional Yes 1.25 0.52-3.02 health insurance coverage (no) High blood Yes -- -- pressure in adulthood (no) Always resided Yes -- -- in Hamilton (no) Pseudo [R.sup.2] (Cox & Snell; 0.075; 0.085; 0.037 Nagelkerke; McFadden) * p < 0.05; ** p < 0.01;*** p < 0.001. ([dagger]) Relative to "normal weight in adulthood" category. ([double dagger]) n for adulthood weight categories (females model): normal = 58, overweight = 41, and obese = 40. ([section]) n for adulthood weight categories (males model): normal = 37, overweight = 73, and obese = 53.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||QUANTITATIVE RESEARCH|
|Author:||Barakat-Haddad, Caroline; Saeed, Usman; Elliott, Susan|
|Publication:||Canadian Journal of Public Health|
|Date:||Jan 1, 2017|
|Previous Article:||Understanding sexual orientation and health in Canada: who are we capturing and who are we missing using the statistics Canada sexual orientation...|
|Next Article:||Diabetes among non-obese Filipino Americans: findings from a large population-based study.|