For the second stage of the study, a questionnaire was mailed to a subsample of mothers of cases and controls to obtain more information on variables not available in databases. Balanced sampling of cells of the first-stage exposure-outcome cross-table was performed. This allowed the overrepresentation of small cells and an increase in statistical power 18 , The addresses of the mothers were obtained from the RAMQ for this subsample.
Only women whose mailing address was in Quebec, who were alive, and whose child was alive and had the same mailing address as the mother were contacted for the second stage of the study. The questionnaire included 40 questions on 15 pages and was available in both French and English. The questionnaire was pretested, and double-entry of questionnaire information was carried out independently by 2 research assistants in 2 Microsoft Access databases Microsoft Corporation, Seattle, Washington in order to ensure the quality of the data.
We conducted a systematic review in order to identify determinants to be considered.
A case-control study of risk factors for asthma in New Zealand children.
Data on 47 determinants were retrieved from either the administrative databases or the questionnaire. Their sources and definitions are detailed below. From the databases, we obtained information on maternal asthma during pregnancy, use of intranasal corticosteroids during pregnancy, mean number of antibiotic prescriptions filled per month during pregnancy, number of prenatal visits, obstetrician visits during pregnancy, mode of delivery, having had a pregnancy in the preceding year, chronic hypertension, pregnancy-induced hypertension, diabetes mellitus, and gestational diabetes.
From the questionnaire, we obtained information on history of allergy and atopy, maternal weight gain during pregnancy, smoking during pregnancy and prior to the index date , and duration of breastfeeding. For paternal risk factors, we obtained information on history of asthma, history of allergy and atopy, and paternal smoking during pregnancy and prior to the index date from the questionnaire. From the databases, we obtained information on male gender, being born small for gestational age, diagnosis of allergic rhinitis ICD-9 code or atopic dermatitis ICD-9 code prior to the index date, and having at least 1 prescription for antibiotics filled prior to the index date or within the first 6 months of life.
From the questionnaire, we obtained information on the development of atopy indicators prior to the index date reported allergies to acetylsalicylic acid, animals, dust mites, hay fever, pollen, other aeroallergens, mold, metal, or cutaneous allergies , and diagnosis of bronchopulmonary disease, administration of oxygen to the newborn after birth, and history of asthma, allergy, and atopy in the child's siblings.
From the databases, we obtained information on maternal age at conception, maternal receipt of social welfare in the year before or during pregnancy , maternal education at delivery, and living in a rural area. Crude rates of childhood asthma for children of asthmatic and nonasthmatic mothers were estimated from the cohort. For the first stage of the study, we obtained crude and adjusted rate ratios using conditional logistic regression and a backward selection strategy For the analysis using the subsample of cases and controls for which database and questionnaire information was available, we obtained crude and adjusted odds ratios using unconditional logistic regression.
Because 47 variables were evaluated, we used 4 steps to identify determinants of childhood asthma. Finally, estimates obtained for maternal asthma were corrected using sampling fractions and maternal asthma estimates from the first stage of the study, as described by Collet et al. The correction was done only for maternal asthma, because the second-stage selection had been made using the maternal asthma—childhood asthma cross-table and was not influenced by the prevalence of the other determinants investigated Questionnaire responders and nonresponders were compared using Pearson chi-square tests for the 20 variables for which information was available in databases.
All analyses were performed using SAS 9. The incidences of childhood asthma were From the cohort, 5, cases of childhood asthma were identified and , matched controls were selected for the first stage of the study. Table 1 displays characteristics available at the first stage for cases and controls, crude rate ratios, and adjusted rate ratios from the multivariate first-stage analysis. The most important determinants associated with the incidence of childhood asthma, in terms of rate ratio magnitude, were asthma during pregnancy, a previous diagnosis of allergic rhinitis and atopic dermatitis in the child prior to the index date, having at least 1 prescription for antibiotics filled prior to the index date or within the first 6 months of life, male gender, and use of antibiotics during pregnancy.
At the end of the 23 weeks allocated for the questionnaire collection process, After discarding 75 questionnaires that had been completed for the wrong child, we had questionnaire information on cases and controls available for use in the second-stage analysis cases and controls born to asthmatic mothers and cases and controls born to nonasthmatic mothers. Questionnaire responders and nonresponders were similar with regard to most of the 20 variables available in the databases, except for small statistically significant differences for receipt of social welfare, level of education, area of residence, mode of delivery, and pregnancy in the preceding year data available upon request.
Table 2 displays the distribution of second-stage potential determinants for questionnaire responders, along with crude odds ratios for childhood asthma. In final multivariate analyses Table 3 , several determinants were statistically significantly associated with increased risk of childhood asthma: a previous diagnosis of bronchopulmonary disease in the child prior to the index date, oxygen administration after birth, a prior diagnosis of atopic dermatitis, maternal receipt of social welfare, filling of 1 or more antibiotic prescriptions for the child prior to the index date or within the first 6 months of life, maternal asthma during pregnancy, maternal report of asthma in the child's father, male gender, asthma in the child's siblings, and use of antibiotics during pregnancy.
The estimate for asthma during pregnancy was corrected using sampling fractions and estimates from the first stage of the study, according to the method described by Collet et al. Since the hazard function may not be constant over age, we conducted subgroup analyses based on age at diagnosis.
Table 4 presents results of the first-stage subgroup analyses. Maternal asthma remained a strong risk factor in both age groups, as did child-related characteristics, but it was difficult to draw conclusions regarding the differences between models because of the difference in the statistical power of these analyses. The second-stage subgroup analyses had very low power, thereby preventing the presentation of informative comparative analyses.
- Allergies linked to increased risk for eosinophilic esophagitis?
- Association Between Obesity and Asthma among Adults: A Hospital Based Case-Control Study?
- the chrysanthemums literary analysis essay?
Overall, other sensitivity analyses conducted did not sizably modify the results obtained data available upon request. RRs were obtained using conditional logistic regression and were adjusted for all of the other variables presented in the table. The present study allowed for identification of the independent effects of numerous potential determinants of childhood asthma, obtained for different time periods of child development, including in utero.
To our knowledge, this is the largest study to date to have considered such a variety of potential determinants of this condition within a single study setting. Our results show that a prior diagnosis of atopic dermatitis, asthma during pregnancy, maternal report of asthma in the child's father, and asthma in the child's siblings were associated with an increased risk of childhood asthma 9 , 11 , 12 , 21— Diagnostic bias is also possible, since these well-known associations may cause a physician to diagnose this condition more quickly in children of asthmatic mothers, and those mothers could also be more likely to contact a physician when noticing asthma-like symptoms in their children.
Several additional determinants were associated with an increased risk of childhood asthma. A previous diagnosis of bronchopulmonary disease in the child prior to the index date was the determinant with the strongest association with childhood asthma. Some investigators report that early-life infections increase childhood asthma risk and that some of the symptoms could be due to early manifestations of asthma not yet diagnosed in the child 9 , 12 , 15 , 24 , Other risk factors include male gender 9 , 24 , 25 , 29 , maternal receipt of social welfare 30—32 , oxygen administration after birth 23 , 27 , filling of 1 or more antibiotic prescriptions for the child prior to the index date or within the first 6 months of life 8 , 15 , 24 , and use of antibiotics during pregnancy 13 , Protective effects were seen for having pets in the home during infancy, breastfeeding, and day-care attendance 5 , 6.
However, it is possible that part of the protective effect of the presence of pets in the home could be due to pet avoidance by families of at-risk children. Interestingly, having a wood-burning fireplace in the home between birth and the index date was associated with a reduction in childhood asthma risk. This could be an indicator of adequate ventilation, which would be generated by the presence of a chimney 34 , or a decline in household humidity levels and a subsequent reduction in mold exposure resulting from wood heating Use of intranasal corticosteroids during pregnancy was associated with a reduced risk of childhood asthma.
- Association between periodontitis and severe asthma in adults: A case–control study.
- Navigation menu;
- winning college application essay.
- Asthma - Wikipedia.
- Asthma - Wikipedia.
- bergen academies entrance exam essay.
- an essay on teachers day celebration;
This determinant could act as an indicator of maternal behaviors which would tend to limit exposure to environmental triggers of atopic manifestations. This could therefore enhance maternal asthma control, which has been associated with a reduced risk of childhood asthma Martel et al. Although we had adequate statistical power to evaluate the independent effects of a large number of determinants in this study, some usually recognized risk factors, such as parental cigarette smoking, were not found to be statistically significantly associated with the risk of childhood asthma in the multivariate analyses.
Crude odds ratios for maternal and paternal smoking during pregnancy and after childbirth were borderline-statistically significant. It may be that the population studied here was relatively homogeneous and did not allow for the distinction of a difference between the high proportions of smokers seen in this study. The higher proportion of female smokers is in accordance with the situation in Quebec, since women aged 25—44 years are those who smoke the most and a high prevalence of smoking has been reported in persons of lower socioeconomic status 3 , This study had strengths and limitations.
The large sample of children of asthmatic and nonasthmatic mothers, along with the combination of databases and the questionnaire, through the use of the 2-stage sampling method, allowed us to study multiple and diverse determinants of childhood asthma and disentangle their independent effects. Administrative databases provide information collected prospectively and independently of the outcome under study. The length of follow-up is also notable, as incident cases of asthma could be identified up to 10 years after birth.
The case definition for childhood asthma was designed to increase specificity and positive predictive value, since a child had to have received a diagnosis and used medication for asthma within a 2-year period. Although this does not entirely prevent residual confounding from affecting the results, the substantial number of variables included in the analyses greatly minimized its potential influence. Another key feature of this study is that for many determinants, the timing of the exposures was considered.
Several risk factors were measured for 2 distinct periods: during pregnancy and after the child's birth. To reduce the likelihood of collinearity, variables related to the maternal and child environments, such as variables for tobacco smoke exposure 37 , 38 , were evaluated in 2 distinct models.
Those models were also useful in attempting to distinguish different effects of a risk factor when the exposure occurred in 2 different time periods. For example, allergen exposure has been suggested to intervene prenatally in the neonatal switching from T-helper 1 lymphocytes to T-helper 2 lymphocytes, as well as postnatally via allergic sensitization 6 , In this study, stronger crude estimates were obtained for environmental determinants assessed for the period following the child's birth compared with assessment of the same determinants during pregnancy. In multivariate analysis, the only environmental determinants remaining in the final reduced model were measured for the period following childbirth.
However, despite the extent of the data collection efforts, information on some potential risk factors was not available. The data sources sometimes did not allow for precise data ascertainment; for example, use of antibiotics prior to the index date or within the first 6 months of life, used as a proxy for infections, could also reflect the reverse-causation phenomenon, since antibiotics could have been more often prescribed in cases of worsening asthmatic symptoms 8 , 15 , 24 , Recall bias in the questionnaire responses is also possible; but while this might have influenced the accuracy of some of the associations measured, information bias is potentially limited, since the crude and adjusted associations were consistent with the magnitude of the effects reported in the literature.
The questionnaire response rate was also not as high as expected, but selection bias, if present, is likely to have been minimal, since responders and nonresponders were similar with regard to most of the 20 database variables. Finer subgroups would also have been interesting to investigate, but the lower power that would have resulted from such segmentation of the cases and controls prevented this analysis. This is a complex task, and although genetic and environmental influences are important components to consider, the timing of exposures is also a key element for understanding this multifactorial disease 5—8 , Combining multiple sources of data provides the opportunity to draw the most complete picture of childhood asthma development.
This study allowed us to identify, within a single setting, the most influential risk factors for childhood asthma among a wide variety of determinants assessed for the prenatal, perinatal, and childhood periods. The authors are grateful to M. The authors also thank Karine Chouinard and Myriam Fortin for data entry and Karine Chouinard for helping with the logistics of the study.
Martel has received a doctoral research scholarship from the K. Malo has received support for clinical assays of various antiasthma drugs produced by several pharmaceutical companies Altana Pharma, GlaxoSmithKline, Hoffmann-LaRoche, and AstraZeneca. Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Article Navigation. Close mobile search navigation Article Navigation.
Volume Article Contents. Oxford Academic. Google Scholar. Jean-Luc Malo. Sylvie Perreault. Poverty can affect the etiology, exacerbation, perception and management of asthma. Studies carried out in different localities have produced conflicting results regarding the association between asthma and poverty. In England, a cross-sectional study involving approximately school children showed that the prevalence of wheezing was higher among children from the less privileged social classes.
This study was designed to test the hypothesis that poor children and adolescents are less likely to develop asthma. Approximately two thirds of its clientele come from the metropolitan area, and the remainder come from the countryside and from other states. Participants were admitted consecutively, on weekday mornings, from March to July of The study sample consisted of outpatients treated in the general pediatric department. Inclusion criteria were having been previously diagnosed with asthma or having had at least three previous episodes of wheezing and dyspnea.
Controls were selected from among outpatients treated in the following pediatric departments: neurology, orthopedics, surgery and ophthalmology. The choice of controls from these departments was aimed at minimizing selection biases, since controls from the general pediatric department could come in greater proportion from the metropolitan area of Recife and, unlike the cases, would likely have been treated for acute diseases.
Individuals with heart disease were excluded, as were those with immunodeficiency, malnutrition, lung diseases other than asthma or severe neurological diseases, as well as those with tuberculosis, systemic mycosis, psychomotor retardation and physical activity limitation caused by a condition other than asthma. One of the parents, or the legal guardian, of each participant gave written informed consent. A total of participants were studied, The vast majority of the participants The number of persons per household ranged from 2 to 15, with a mean and standard deviation of 4.
Maternal years of schooling ranged from 0 to 18, with a mean and standard deviation of 6. There was no statistically significant difference between the cases and controls in terms of any of the variables evaluated.
Ammonia, ammonium, and the risk of asthma: A register-based : Environmental Epidemiology
Together, these two groups accounted for No statistically significant association was found between poverty and asthma. The p value found in the univariate analysis demonstrated that adjusting for potential confounding factors would not change this result significantly, thereby rendering the multivariate analysis useless Table 1. Potential confounding factors were as follows: age; gender; place of origin; type of dwelling; number of rooms in the dwelling; maternal level of education; history of atopy; breastfeeding; eating pattern after weaning; vaccination schedule; medication use; viral diseases; previous parasitic infections or tuberculosis; passive smoking; and the presence of animals or aeroallergens in the household.
There is sufficient evidence that the prevalence of pediatric asthma is currently high in Brazil, at least in urban zones. The objective of the present study was to analyze the role that poverty plays in the development of asthma as a risk factor or as a protective factor , using the hygiene hypothesis as a theoretical reference. Another study 18 produced similar results. Although the authors found a higher prevalence of asthma symptoms in the upper socioeconomic groups, they did not mention how the cut-off points were determined.
A cross-sectional study of school children in Recife 8 produced results similar to those of the abovementioned study, 23 in which the prevalence of asthma symptoms was significantly higher in school children whose mothers had a higher level of education. Those three studies, however, were not designed to determine risk, and their findings might have been affected by biases or confounding factors.
Studies carried out in other countries have also produced discrepant results. In a survey conducted in Singapore using the International Study of Asthma and Allergies in Childhood ISAAC protocol, the authors found the prevalence of asthma to be higher in children from the upper socioeconomic classes. Worldwide, poverty is a burden for those who live in it. The determinants and characteristics of poverty vary among populations, even within Brazil.
According to one study, 11 poverty in Brazil results from the unequal distribution of wealth and unequal opportunities for socioeconomic inclusion. Poverty is more widespread in the northern and northeastern regions of the country. Even without scientific evidence, comparing different communities, it can be hypothesized that the determinants of poverty in Singapore and New Zealand, for example, are different from those observed in Brazil, and that the findings can therefore differ. It cannot be stated that the lack of a relationship between poverty and asthma found in our study resulted from the homogeneity of the sample studied, which consisted mostly of poor individuals, since subgroups were representative in all strata.
However, since social inequality in Brazil is high, 11 the external validity of this study is limited to patients treated via the SUS, and our findings cannot be extrapolated to the general population. This study has limitations that need to be addressed. First, all data were collected using a questionnaire.
The characterization of asthma based on the presence of three or more lifetime episodes of wheezing or on a previous diagnosis of asthma, although criteria that are widely used and accepted, 1,24 might have generated overdiagnosis of cases and selection bias. For studies aimed at determining asthma risk, some authors 25 recommend that pulmonary function tests and allergy tests be included.
Due to logistical constraints, it was not possible to carry out these complementary tests. Another potential limitation is memory bias, which might have resulted in an underestimation of the number of cases of asthma. In order to avoid this bias in determining poverty, we used the per capita income of the last month as a proxy for per capita income in the first years of life, assuming that the socioeconomic situation of the participants remained stable throughout their lives.
This might have also generated a non-differential selection bias. Despite these limitations, the study appears valid and, in a reasonably precise manner, confirms that poverty has no effect on asthma in children and adolescents treated via the SUS in the Northeast of Brazil. However, further studies are needed, especially cohort studies, in order to clearly explain this relationship in Brazil.
Global initiative for asthma. Global strategy for asthma management and prevention.
Related case control study on asthma
Copyright 2019 - All Right Reserved