Copyright 2004 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2004
Because obesity promotes inflammation and imposes mechanical constraints to the airways, a high birth weight may be a risk factor for asthma in childhood. However, to our knowledge, few studies have examined this potential relationship.
To determine the relationship between high birth weight and risk of emergency visits for asthma during childhood.
Population-based cohort study.
All neonates born at term (≥37 weeks) between April 1, 1985, and March 31, 1988, in Alberta (N = 83 595). We divided the cohort into birth-weight categories: low (<2.5 kg), normal (2.5-4.5 kg), or high (>4.5 kg). The cohort was observed prospectively for 10 years.
Main Outcome Measure
Comparison of risk of emergency visits for asthma over 10 years across the birth-weight categories.
Neonates born with a high birth weight had a significantly increased risk of emergency visits for asthma during childhood compared with neonates born with a normal birth weight (relative risk [RR], 1.16; 95% confidence interval [CI], 1.04-1.29). The relationship between birth weight and emergency visits for asthma beyond a birth weight of 4.5 kg was linear, such that every increment of 0.10 kg in birth weight was associated with an additional 10% (95% CI, 2%-19%) increase in the risk of emergency visits for asthma. Other factors associated with an elevated risk for emergency asthma visits during childhood included male sex (RR, 1.26; 95% CI, 1.22-1.30), aboriginal status (RR, 1.20; 95% CI, 1.11-1.29), and low-income status (RR, 1.11; 95% CI, 1.06-1.16).
A high, but not low, birth weight is a risk factor for increased emergency visits during childhood. The risk increases linearly beyond a birth weight of 4.5 kg.
CHILDHOOD ASTHMA IS A common chronic condition that is increasing in prevalence and severity, particularly among those younger than 5 years.1 Indeed, between 1987 and 1997, there was a 57% increase in the asthma hospitalization rate among those younger than 5 years.2 Interestingly, during this same interval, there was a substantial increase in the birth weight of neonates born in the Western world and an increased incidence of childhood obesity.3 While there is a paucity of studies evaluating lung function in high-birth-weight neonates, based on data from older children,4 increased body fat might be expected to produce lung function impairments in neonates that could increase their risk of asthma. One plausible explanation for the increased prevalence of early childhood asthma is the temporal increase in birth weight in the community.
We conducted a large population-based cohort study with the following aims: (1) to determine if low or high birth weight modifies the risk of childhood asthma independent of the effects of prematurity; (2) to describe the shape of the relationship between birth weight and childhood asthma; and (3) to determine if factors such as age, sex, race, and income status were also related to childhood asthma.
We identified all live births occurring in Alberta, Canada, between April 1, 1985, and March 31, 1988, using the Alberta Health Care Insurance Plan registry (N = 96 359). This registry contains information on sex, birth weight, gestational age of the neonate, and data concerning mother's age, race, marital status, and area of residence. We excluded 7627 children (7.9%) because their birth weight or gestational age was not recorded in the Alberta Health Care Insurance Plan registry. We also excluded children born at a gestational age of less than 37 weeks (n = 5137) to eliminate the influence of prematurity on the relationship between birth weight and risk of asthma. Neonates born at term (N = 83 595) were observed for 10 years after their birth.
From the physicians' claims database, we identified all physician encounters occurring in the emergency departments for which asthma was the principal diagnosis (as deemed by a physician) during the first 10 years of the children's lives. We used International Classification of Diseases, Ninth Revision, Clinical Modification code 493.x to capture asthma encounters.
The children in our cohort were grouped according to the birth weight reported in the Alberta Health Care Insurance Plan registry. Birth weights were categorized as follows: low (<2.5 kg), high (>4.5 kg), and normal (2.5-4.5 kg). χ2 Tests were used for comparison of sex, aboriginal status, and low-income status between birth-weight categories. Analysis of variance and post hoc comparisons (the Tukey studentized range test) were used to test the differences in gestational and maternal ages between birth-weight categories. We used a Poisson regression analysis using birth-weight categories as the independent variable and emergency visits as the dependent variable. To fit the data into a Poisson equation, person-years were log transformed and included in the model as an offset.5 To the baseline model that included only birth-weight categories and risk for asthma emergency visits, we added, one by one, maternal age, gestational age, sex of the child, low-income family, and race. We divided race into aboriginal and nonaboriginal subjects. Low-income families were defined as those that had a combined annual (single) family income of less than Can $27 210. This constitutes approximately 10% of Alberta families and is the threshold the provincial government uses to determine financial assistance for health care premiums.6
To determine the exact shape of the relationship between birth weight and risk of emergency visits for childhood asthma, we also plotted histograms of the frequency distribution for asthma emergency visits using birth weight as the independent variable. We then obtained an estimate of these distributions using smooth interpolation methods to achieve a good fit with minimal restriction on the functional form of this relationship. Finally, we performed a logistic regression analysis to determine the relationship between high birth weight and risk of emergency visits for childhood asthma. A 2-tailed P<.05 was considered statistically significant. All analyses were conducted using SAS statistical software, version 7.0 (SAS Institute Inc, Cary, NC). Data are presented as mean ± SD unless otherwise indicated.
There were 88 732 children born in Alberta between April 1, 1985, and March 31, 1988, with valid birth-weight data. Of these children, 5137 were born premature (<37 weeks' gestational age), leaving 83 595 term neonates for follow-up; these children were then followed up for 10 years. The mean birth weight for these children was 3.44 ± 0.48 kg, and the mean gestational age was 39.6 ± 1.2 weeks. Girls constituted 48.8% of the cohort (n = 40 794); 4.8% (n = 3976) of the children were aboriginal, and 14.6% (n = 12 205) were born into a low-income family. The mean maternal age was 27.0 ± 4.8 years. During follow-up, 5413 (6.5%) of the children had at least 1 emergency visit for asthma. The total number of emergency visits was 15 634.
The characteristics of children with low, high, and normal birth weights are described in Table 1. Most female neonates were born with a low birth weight; the least were born with a high birth weight. Although maternal age was similar between these birth-weight groups, children with low birth weights were more likely to come from a family with a low income compared with those with normal or high birth weights. In contrast, aboriginal children were more likely to have high birth weights compared with nonaboriginal children.
Compared with children with a normal birth weight, those with a high birth weight had a higher risk of emergency visits for asthma within the first 10 years of their lives (Table 2). When we controlled for gestational age, sex, maternal age, aboriginal status, and low-income status, a high birth weight was significantly associated with an increased risk of emergency visits for childhood asthma (relative risk [RR], 1.16; 95% confidence interval [CI], 1.04-1.29) compared with a normal birth weight (P = .01). This was a linear relationship, such that those with the highest birth weight had the greatest risk of experiencing emergency visits for asthma (Figure 1). For every 0.10-kg increase in birth weight above 4.5 kg, there was an additional 10% (95% CI, 2%-19%) increase in the risk of experiencing an emergency visit over 10 years. Adjustments for gestational age, sex, maternal age, aboriginal status, and low-income status did not change this relationship (RR increase of 10% [95% CI, 2%-19%] for every 0.10-kg increment in birth weight >4.5 kg). In contrast, we did not observe a significant relationship between low birth weight and risk of emergency visits for childhood asthma in children born at full term (RR, 1.00; 95% CI, 0.90-1.11) (P = .97). In a secondary analysis, we excluded emergency visits occurring within the first 2 years of life. The findings from the secondary analysis were similar to those from the main analysis (RR for high-birth-weight neonates, 1.17 [95% CI, 1.04-1.31]; and RR for low-birth-weight neonates, 1.07 [95% CI, 0.95-1.20]).
A spline interpolation curve demonstrating the relationship between risk of having at least 1 emergency visit for asthma and birth weight during the first 10 years of life.
Table 2 shows the crude and adjusted rates for emergency visits for childhood asthma in those with low, normal, and high birth weights across sex, income status, and aboriginal status. Boys consistently had a higher risk than girls for emergency visits for asthma. Similarly, aboriginal children had a higher risk for emergency visits for asthma. A similar but less striking relationship was observed for those children born in a family with a low income.
This large population-based study of childhood asthma and birth weight produced some important and novel observations. First, we observed that children born with a high birth weight (>4.5 kg) had the greatest incident risk of emergency visits for asthma. Second, the association between birth weight and emergency visits for childhood above a birth weight of 4.5 kg was linear such that those neonates having the highest birth weight had the greatest risk for asthma emergency visits. Although more boys and aboriginal children were found within the high-birth-weight category, adjustments for these and other confounding factors did not materially change this relationship. Third, a low birth weight was not an independent predictor of emergency visits for asthma during childhood.
The mechanism by which high birth weight results in an elevated incidence of emergency visits for childhood asthma is not well understood. However, it is known that heavy neonates at birth tend to remain obese throughout childhood and into adolescence.7- 10 This association is particularly pronounced when the birth weight is greater than 4.5 kg, such that the risk of continued obesity is 3 to 4 times that of those in lower-birth-weight categories.10 Because adiposity adversely affects lung function by decreasing expiratory flow rates and inducing premature closure of peripheral airways,4 obese children are more likely to develop asthma symptoms than age-matched normal-weight kids. Moreover, obesity can decrease respiratory muscle function and increase the work and energy cost of breathing, leading to worsening of asthma symptoms.11- 13 The proinflammatory state related to obesity may also contribute to the asthma burdens in the obese population. Biomedical experiments have demonstrated that adipocytes up-regulate the production of various proinflammatory cytokines, including leptin, interleukin 6, and tumor necrosis factor α, that can localize to the pulmonary system and exacerbate airway inflammation.14 Adipocytes may also increase the production and activation of mast cells in large airways, predisposing airways to bronchospasm.15
There remains considerable controversy surrounding the role of low birth weight in childhood asthma. Although some studies16- 18 suggest that low birth weight is a significant risk factor for the development of childhood asthma, others19- 23 do not. Part of the confusion may be related to the possible confounding effect of gestational age. Because prematurity and low birth weight are intrinsically linked,23 the independent contribution of low birth weight to the development of childhood asthma has not been fully established. An advantage of our study was that by excluding all premature neonates (gestational age, <37 weeks), we were able to determine the role of low birth weight without significant confounding by gestational age. Our observations are, therefore, consistent with the concept that low birth weight by itself does not predict incidence or severity of asthma later in childhood.19
Our observation that boys have a higher incidence of asthma emergency visits compared with girls is consistent with that of previous reports.24,25 Although the exact mechanism is not known, several studies25,26 suggest that female neonates have proportionately larger airways relative to their lung size than do male neonates, resulting in increased expired flow rates and airway conductance. We also found that aboriginal children had a higher risk of emergency asthma visits during childhood than nonaboriginal children. Although, from our study, we could not determine why this pattern of health care use for asthma was present among aboriginal children, a previous study27 suggests that aboriginal neonates have a substantially higher risk of early childhood infections, which may damage their airways early in life, predisposing them to the development of asthma during childhood. An alternate explanation is that aboriginal children may not be seeking or receiving optimal primary care in the community, leading to a heavy reliance on immediate rescue care in emergency departments.
A major strength of our study is the large sample size, which provided the opportunity to adequately control for the potential confounding effects of gestational age (ie, prematurity), sex, and other important sociodemographic factors to elucidate the independent association between birth weight and incident risk of childhood asthma. Adjustments for prematurity and sex were particularly important because they have been shown to be leading predictors of childhood asthma.17,25 Another advantage was that, because of the nature of our study design (a population-based cohort study), we were able to capture nearly all neonates born in Alberta during the study period. This avoided selection bias, which could be present in studies that use a highly selected cohort of patients. Furthermore, while many of the previous studies relied on self-report, we estimated the incidence and severity of asthma using emergency visits as case events. This has the advantage of avoiding recall bias, and allowed us to determine the risk of clinically significant asthma that brings patients to see their physicians for assistance and treatment.
Other childhood conditions may clinically mimic asthma. By relying on physician diagnosis of asthma, there was potential for diagnostic misclassification. However, for this bias to have made a significant difference to our findings, there had to have been a preferential diagnostic bias for asthma in children born with a high birth weight compared with those born with a low or normal birth weight. Because high birth weight has not been previously reported as a putative determinant of childhood asthma, to our knowledge, there is no material reason why clinicians would have preferentially labeled high-birth-weight children as "asthmatics," making this a remote possibility. Another limitation was that we did not capture symptomatic patients who did not have an emergency visit for their asthma. Therefore, our findings cannot be generalized to children who wheeze but do not seek emergency help.
In summary, our study findings indicate that there is a significant relationship between high, but not low, birth weight and emergency visits for childhood asthma. These data are consistent with other reports28 that indicate that individuals with severe asthma requiring emergency services tend to be more obese than those without asthma. Because there has been a gradual increase in birth weight in North America during the past 20 years,29 our observations may explain why there has been an increased incidence of severe childhood asthma during this time. Future work is needed for determining the mechanism by which high birth weight leads to increased incidence of emergency visits for childhood asthma. Our findings of increased emergency use for asthma among aboriginal children also highlight the need for future studies aimed at reducing the asthma burden in aboriginal communities.
Obesity is a proinflammatory state and adds to the asthma burden in the community. Children born with a high birth weight (>4.5 kg) have close to a 20% increased probability of emergency visits for asthma in the first 10 years of their lives than children born with a normal birth weight. For every 0.10-kg increase in birth weight above 4.5 kg, there is an additional 10% increase in the probability of having an emergency visit for asthma. A high birth weight, therefore, is an important risk factor for childhood asthma.
Corresponding author: Don D. Sin, MD, MPH, 2E4.29 Walter C. Mackenzie Centre, University of Alberta, Edmonton, Alberta, Canada T6G 2B7 (e-mail: email@example.com).
Accepted for publication June 19, 2003.
This study was supported by a New Investigator Award from the Canadian Institute of Health Research, Ottawa, Ontario (Dr Sin); and a Population Health Investigator Award from the Alberta Heritage Foundation for Medical Research, Edmonton (Dr Sin).
Sin DD, Spier S, Svenson LW, Schopflocher DP, Senthilselvan A, Cowie RL, Man SFP. The Relationship Between Birth Weight and Childhood AsthmaA Population-Based Cohort Study. Arch Pediatr Adolesc Med. 2004;158(1):60-64. doi:10.1001/archpedi.158.1.60