Abstract
Rationale: Allostatic load (AL), a novel measure of the physiologically dysregulated response of the body to stress, represents a biomarker of chronic stress exposure.
Objectives: To determine whether preadolescent children with high AL are more susceptible to asthma as adolescents.
Methods: This was a prospective evaluation of children recruited at 7 to 10 years of age in the nested case-control arm of the Study of Asthma, Genes and Environment and followed until 11 to 14 years of age. AL was measured using eight biomarkers: fasting glucose, total cholesterol, high-density lipoprotein cholesterol, dehydroepiandrosterone sulfate, cortisol, systolic and diastolic blood pressure, and waist-to-hip ratio. AL, created from the sum of biomarkers in a high-risk quartile, was related to prevalence and incidence of asthma using logistic regression.
Measurements and Main Results: Among 352 participants followed until 11 to 14 years of age, prevalent asthma was four times more likely in boys with high (>3) versus low (≤2) AL after adjusting for current asthma/atopy, age, ethnicity, parental history of asthma, and overweight status. Similar results were observed in the analysis of new-onset asthma in boys (adjusted odds ratio, 4.35; 95% confidence interval, 1.19–15.9). In girls, there were no associations between AL and asthma. In the analysis of a subset of biomarkers, combinations of total cholesterol, glucose, and cortisol were associated with similar or greater risk of asthma prevalence or onset in boys.
Conclusions: AL and its biomarkers are associated with an increased likelihood of asthma in adolescent boys. The observed association between AL and asthma may be attributable to a combined subset of AL biomarkers.
Clinicians use the Asthma Predictive Index in early life to predict onset of asthma in school-age children. However, there are limited data on which indices are useful for adolescent-onset asthma.
Recognizing the multisystem effects of allostatic load, this research presents evidence for candidate stress and metabolic biomarkers for investigation into their clinical utility as an index for adolescent-onset asthma.
New-onset asthma affects up to 10% of youth in the United States(1) and in adolescents imposes an extensive burden of morbidity and disease management onto a period of significant life changes. It is also a condition that persists into adulthood. Risk factors for early childhood asthma have been examined extensively; however, it remains challenging to predict which children will develop asthma as adolescents (2). Strong links have been reported between early life exposure to maternal distress, psychosocial stress, or depressive and anxiety symptoms in children and the development and prognosis of asthma (3–6). Few studies have demonstrated a connection between stressor exposure and asthma in adolescents, and the underlying biological links between these issues remain to be clarified (7).
Allostatic load (AL) is a concept developed to elucidate associations between chronic stress and adverse health outcomes. Allostasis refers to the adaptation of the body, through activation of the sympathetic nervous system, hypothalamus-pituitary-adrenal (HPA) axis, and metabolic system, to maintain physiological balance under stress (8). AL reflects a state of exhaustion after long-term stress exposure, whereby the normal allostasis response becomes dysfunctional in the following sequential manner. First, primary mediators like stress hormones (i.e., epinephrine, norepinephrine, and cortisol) and their antagonist, dehydroepiandrosterone sulfate (DHEA-S), become dysregulated. Second, metabolic (i.e., glucose and lipid metabolism) and cardiovascular (i.e., blood pressure and heart rate) systems adjust themselves in an attempt to compensate for primary mediators. Finally, resultant abnormal organ activities progress from subclinical stages to clinical disease (8). Various studies have measured the primary mediators and secondary outcomes of allostasis using a composite AL index to distinguish individuals at relatively high risk for clinical outcomes, including cardiovascular disease, diabetes, mood and cognitive disorders, and all-cause mortality (8–10). Studies have also examined subsets or individual components of the AL index in relation to adverse health outcomes. For example, subgroup analyses in one of the “MacArthur studies of successful aging” demonstrated that epinephrine, waist-to-hip ratio, and cortisol have greater roles in predicting the physical decline in adults, whereas diastolic blood pressure, epinephrine, and glycosylated hemoglobin are strongly associated with cognitive decline (11). Limited research has explored the impact of AL and its biomarkers on diseases in the pediatric years.
The biology that links stress and asthma is complex and requires consideration of multiple interactive pathways of influence. Previous studies have only examined these pathways individually, raising a need for an integrative method for cumulative biological risk assessment (4, 12). The AL index is a composite measure that enables the consideration of compound pathways simultaneously. This biologic measure also addresses the challenges in quantifying stress in epidemiological studies (4), ranging from measures of social circumstances to reports of stressful life events. Such studies have overlooked dramatic interpersonal differences in stress perception, which is influenced by many factors, such as personality and coping mechanisms, and by the effects of multiple environmental stressors (12, 13). The utility of the AL index is that it provides a biological construct to account for interpersonal variations in responses to different stressors (13). AL has been proposed as an explanation for the link between stress and asthma pathogenesis and/or morbidity (14) but has yet to be examined. The objective of the present study was to determine whether preadolescent children with higher levels of AL are more susceptible to asthma as adolescents. The secondary objective was to examine the association of the individual AL biomarkers with asthma and asthma phenotypes. Some of the preliminary results of this study have been previously reported in the form of an abstract (15).
This study was a prospective follow-up of the 1995 Study of Asthma, Genes and Environment (SAGE) birth cohort (16), approved by the University of Manitoba Research Ethics Board (H2002:078). From a mail-out survey to SAGE households, children with or without parent-reported asthma were randomly recruited from rural and urban areas into a nested case-control study at 7 to 10 years of age (mean, 9 yr) and were revisited at 9 to 11 years of age (mean, 10.5 yr) and 11 to 14 years of age (mean, 12.5 yr). At 9 and 12.5 years of age, asthma and atopy were confirmed by a pediatric allergist and a positive skin prick test to any of 14 common indoor and outdoor allergens, respectively (16, 17). At 10.5 years of age, an AL index was derived from these eight markers: early morning fasting blood sample determinations of serum glucose, total cholesterol (TC), high-density lipoprotein (HDL), DHEA-S (an HPA axis antagonist), and cortisol, as well as measurements of resting systolic and diastolic blood pressure and the waist-to-hip ratio, an indicator of central adiposity (8, 13, 18–23). Waist circumference was measured half way between the iliac crest and lowest rib, and hip circumference was measured at the maximum circumference over the buttocks.
Similar to other studies (8, 13, 18–23), a value of 1 was added to the AL score for each biomarker value that fell into the highest quartile for glucose, TC, systolic blood pressure, diastolic blood pressure, and waist-to-hip ratio, and the lowest quartile for HDL and DHEA-S. Because low cortisol levels represent a dysregulated HPA axis response to stress (24) and are valid indicators of AL in prepubertal children (25), a value of 1 was added for cortisol levels below or above the 12.5th percentile (13, 19). Sex-specific cut-off points were applied for each biomarker (Table 1). A total AL score of greater than 3 (highest quartile) was defined as high AL in both sexes.
| Cut-Off Points | ||
| Boys | Girls | |
| Systolic blood pressure, mm Hg | ≥116 | ≥116 |
| Diastolic blood pressure, mmHg | ≥68 | ≥68 |
| Waist-to-hip ratio | ≥0.89 | ≥0.87 |
| Fasting serum glucose, mmol/L | ≥5.20 (≥93.6 mg/dl) | ≥5.10 (≥91.8 mg/dl) |
| Serum total cholesterol, mmol/L | ≥4.54 (≥175.6 mg/dl) | ≥4.50 (≥174.0 mg/dl) |
| Serum HDL, mmol/L | ≤1.37 (≤53.0 mg/dl) | ≤1.33 (≤51.4 mg/dl) |
| Serum DHEA-S, nmol/L | ≤1.32 (≤3.6 ng/ml) | ≤1.20 (≤3.3 ng/ml) |
| Serum cortisol, nmol/L | ≤62.75 (≤1,731.3 μg/dl) or ≥257.31 (≥7,099.2 μg/dl) | ≤65.92 (≤1,818.7μg/dl) or ≥193.11 (≥5,327.9 μg/dl) |
| Total AL | >3 | >3 |
Several confounding factors (age, ethnicity [white vs. other], history of food allergy, atopy or asthma, postsecondary education of parent, and exposure to secondhand smoke) were derived from parent questionnaires (26–30). Using validated Tanner stage schematics (31, 32), puberty stage was self-reported at 10.5 years of age and dichotomized as prepubertal (Tanner < 2) or pubertal (Tanner ≥ 2). Overweight status was defined as a BMI z score greater than 1.04 at 10.5 years of age (33).
Purposeful stepwise logistic regression (34) was conducted using SPSS 18.0 to determine the association between high AL at 10.5 years of age and prevalent and incident asthma at 12.5 years of age. Parent educational level, pubertal stage, and secondhand smoke exposure did not meet our P < 0.10 criterion for initial variable selection and were not retained in final models. Separate models with overweight status as a covariate were run to assess AL independent from overweight. Analyses were stratified by sex subsequent to an interaction between AL and sex. To determine the association of AL with new-onset adolescent asthma, prevalence analyses were adjusted for asthma at 9 years of age, and incident analyses were conducted after excluding children with atopic asthma at 9 years of age because childhood atopic asthma is more likely to persist until adolescence (35).
A subset analysis was performed to examine the association of each individual AL biomarker with asthma and asthma phenotypes. Because studies have shown a decrease in glucose levels parallel to increases in insulin secretion and insulin resistance in girls who enter puberty, we also examined the association of high fasting insulin level with asthma in girls (36, 37). To identify the combination of biomarkers that potentially determined the associations between AL and asthma, we tested different combinations of biomarkers in relation to asthma and asthma phenotypes. The candidate set of biomarkers for these analyses was based on the level of significance for the individual associations (P < 0.10). The Hosmer and Lemeshow goodness-of-fit test was applied to all models.
A total of 439 children with an AL measurement at 10.5 years of age were included; from this sample, 160 (36.4%) had asthma, and 248 (56.5%) lived in urban areas. After 2 years, 352 children were revisited at 12.5 years of age (80.2%). The sex-specific distribution of the AL index is illustrated in Figure 1. No statistically significant differences were found between participants who did or did not attend the follow-up visit with respect to any of the study covariates, except for age and ethnicity. Children who were not present at the follow-up visit were on average 3.3 months older than their study peers (P < 0.001). Attrition at follow-up was lower for whites than for nonwhites (17.2 vs. 28.0%, respectively; P < 0.01). Although nonwhite families (primarily First-Nations) were recruited independently to maximize participation in this study (16), they accounted for the majority of the loss to follow-up.
Figure 1. Sex-specific distribution of the allostatic load score among study children.
[More] [Minimize]All 352 children who were present at both visits were included in the prevalence analysis. A total of 115 (32.7%) participants had prevalent asthma; of these, 70 (21.6%) and 33 (10.2%) had atopic and nonatopic asthma, respectively. Twelve participants with asthma did not consent to skin prick testing. Regarding physical development, 67.9% of boys and 41.2% of girls were categorized as prepubertal (Tanner stage < 2). In general, total asthma and both asthma phenotypes tended to be more prevalent in boys with high AL scores and girls with low AL scores, but differences were not statistically significant before adjustment for covariates (Figure 2). Table 2 shows the distribution of AL index and prevalent asthma among different covariates. High AL was less prevalent in white and normal-weight children (P < 0.05). Asthma, on the other hand, was more prevalent among children who were younger or atopic, who had a positive history of food allergy, or who had parental history of asthma. Similar associations were observed between AL scores and ethnicity or overweight among children in the incidence analyses. Incident asthma was more prevalent among children with a history of food allergy or parental history of asthma (P < 0.05; data not shown).
Figure 2. Prevalence and incidence of asthma among participants with high versus low allostatic score. Top, All children. Middle, Girls. Bottom, Boys. AL = allostatic load.
[More] [Minimize]| Prevalent Asthma (age 12.5 yr) | |||||||||||||
| High Allostatic Load Score | All Asthma | Atopic Asthma | Nonatopic Asthma | ||||||||||
| All Children n (%) | n | Prevalence (%) | P Value | n | Prevalence (%) | P Value | n | Prevalence (%) | P Value | n | Prevalence (%) | P Value | |
| Atopy (age 9) | |||||||||||||
| No | 175 (49.9) | 60 | 34.3 | 0.40 | 33 | 18.9 | <0.001* | 4 | 2.9 | <0.001* | 26 | 16.1 | 0.05* |
| Yes | 176 (50.1) | 53 | 30.1 | 81 | 46.0 | 65 | 43.0 | 7 | 7.5 | ||||
| Food allergy | |||||||||||||
| No | 310 (88.1) | 103 | 33.2 | 0.22 | 88 | 28.4 | <0.001* | 47 | 18.4 | <0.001* | 32 | 13.3 | 0.5 |
| Yes | 42 (11.9) | 10 | 29.3 | 27 | 64.3 | 23 | 63.9 | 1 | 7.1 | ||||
| Sex | |||||||||||||
| Female | 153 (43.5) | 54 | 29.6 | 0.26 | 52 | 34.0 | 0.64 | 30 | 24.6 | 0.86 | 15 | 10.4 | 0.16 |
| Male | 199 (56.5) | 59 | 35.3 | 63 | 31.7 | 40 | 23.7 | 18 | 16.4 | ||||
| Parental history of asthma | |||||||||||||
| No | 172 (55.8) | 62 | 36.0 | 0.17 | 48 | 27.9 | 0.03* | 28 | 19.4 | 0.05* | 18 | 13.4 | 0.98 |
| Yes | 136 (44.2) | 39 | 28.7 | 54 | 39.7 | 33 | 29.7 | 12 | 13.3 | ||||
| Ethnicity | |||||||||||||
| White | 274 (78.1) | 80 | 29.2 | 0.04* | 88 | 32.1 | 0.63 | 56 | 24.3 | 0.87 | 23 | 11.7 | 0.22 |
| Nonwhite | 77 (21.9) | 32 | 41.6 | 27 | 35.1 | 14 | 23.3 | 10 | 17.9 | ||||
| Overweight | |||||||||||||
| No | 241 (68.7) | 53 | 22.0 | <0.001* | 81 | 33.6 | 0.62 | 51 | 25.5 | 0.42 | 21 | 12.4 | 0.64 |
| Yes | 110 (31.3) | 59 | 53.6 | 34 | 30.9 | 19 | 21.1 | 12 | 14.5 | ||||
| Tobacco smoke exposure at home | |||||||||||||
| No | 260 (73.9) | 81 | 31.2 | 0.52 | 89 | 34.2 | 0.29 | 54 | 25.2 | 0.43 | 27 | 14.4 | 0.25 |
| Yes | 92 (26.1) | 32 | 34.8 | 26 | 28.3 | 16 | 20.8 | 6 | 9.0 | ||||
| Pubertal stage | |||||||||||||
| Early | 196 (56.5) | 57 | 29.1 | 0.23 | 63 | 32.1 | 0.65 | 40 | 23.8 | 0.78 | 18 | 12.3 | 0.63 |
| Late | 151 (43.5) | 53 | 35.1 | 52 | 34.4 | 30 | 25.2 | 15 | 14.4 | ||||
| Parental education (postsecondary) | |||||||||||||
| No | 17 (5.2) | 7 | 41.2 | 0.39 | 7 | 41.2 | 0.47 | 4 | 28.6 | 0.74 | 2 | 16.7 | 0.67 |
| Yes | 311 (94.8) | 97 | 31.2 | 102 | 32.8 | 64 | 24.6 | 28 | 12.5 | ||||
In boys, asthma at 12.5 years of age was more prevalent among those with high AL when they were 10.5 years of age, although this association was not statistically significant without adjustment (Table 3). Existing asthma or atopy and parental history of asthma at 9 years of age were also statistically significant predictors of asthma at 12.5 years of age. After adjusting for the covariates in model 1, there was a 33% increased likelihood of asthma for each single unit increase in the AL score (Table 3; continuous AL score). Also, boys who had high AL scores were 2.7 times more likely to have asthma in adolescence compared with those in the low-AL group. These associations were statistically significant and independent of age, ethnicity, parental history of asthma, and existing asthma or atopy at 9 years of age. Existing asthma and atopy, as well as parental history of asthma, remained significant predictors for asthma at 12.5 years of age (data not shown). After additional adjustment for overweight status (model 2), we observed a 63% increased chance of having asthma for each unit AL score increase (Table 3; continuous AL score). Additionally, the association between high AL and asthma increased to an odds ratio of 4. These associations were independent of all covariates, including overweight status. In girls, high AL was not associated with asthma at 12.5 years of age, and this finding was consistent at baseline and after the addition of potential covariates in regression models.
| Boys | Girls | |||||
| Crude OR | Adjusted OR Model 1 | Adjusted OR Model 2 | Crude OR | Adjusted OR Model 1 | Adjusted OR Model 2 | |
| All asthma* (n = 352) | ||||||
| AL score continuous | 1.11 (0.90–1.36) | 1.33 (1.00–1.83)† | 1.63 (1.12–2.37)† | 0.88 (0.68–1.13) | 0.72 (0.50–1.03) | 0.70 (0.48–1.03) |
| AL low | 1 | 1 | 1 | 1 | 1 | 1 |
| AL high | 1.29 (0.68–2.46) | 2.70 (1.01–7.23)† | 4.01 (1.37–11.8)† | 0.74 (0.36–1.50) | 0.33 (0.11–1.04) | 0.32 (0.10–1.02) |
| Atopic asthma‡ (n = 319) | ||||||
| AL low | 1 | 1 | 1 | 1 | 1 | 1 |
| AL high | 1.11 (0.51–2.41) | 2.30 (0.62–8.56) | 4.58 (1.04–20.2)† | 1.04 (0.45–2.41) | 0.63 (0.17–2.36) | 0.55 (0.14–2.13) |
| Nonatopic asthma§ (n = 282) | ||||||
| AL low | 1 | 1 | 1 | 1 | 1 | 1 |
| AL high | 2.26 (0.76–6.69) | 3.72 (1.03–13.5)† | 3.82 (0.91–16.1) | 0.44 (0.14–1.46) | 0.45 (0.11–1.78) | 0.33 (0.07–1.52) |
| Incident asthma‖ (n = 258) | ||||||
| AL low | 1 | 1 | 1 | 1 | 1 | 1 |
| AL high | 1.54 (0.59–4.01) | 2.06 (0.73–5.83) | 4.35 (1.19–15.9)† | 0.39 (0.13–1.12) | 0.48 (0.16–1.48) | 0.44 (0.14–1.39) |
Boys who had high AL scores were 4.6 times more likely to have atopic asthma at 12.5 years of age after controlling for covariates in model 2 (Table 3). Nonatopic asthma was significantly more prevalent among boys with high AL in model 1, independent of asthma and atopy status at 9 years of age. Nevertheless, after overweight status was added in model 2, this association became less significant (Table 3). Atopic asthma at 9 years of age, history of food allergy, and parental asthma were statistically associated with atopic asthma at 12.5 years of age, whereas nonatopic asthma at 9 years of age was the only predictor for nonatopic asthma in adolescent years (data not shown). In contrast, high AL scores among girls did not show a significant association with atopic or nonatopic asthma in either of the models (Table 3).
Excluding children with atopic asthma at 9 years of age, the incidence analysis included 258 children with 49 new cases of asthma (Table 3). In boys, high AL score increased the likelihood of asthma development in adolescence more than 4-fold after adjusting for the covariates in model 2. History of food allergy was another significant predictor of asthma development in model 2 (data not shown).
In girls, a high AL score was not predictive of asthma in both models (Table 3); however, as in boys, history of food allergy was associated with asthma development (data not shown). We were unable to perform the incidence analysis for each asthma phenotype due to small sample sizes.
After adjustment for model 2 covariates (Table 4), abnormal cortisol response, as a single biomarker, was more likely to be present in boys with prevalent nonatopic asthma. High TC was significantly associated with new-onset asthma in boys. In girls, on the other hand, high serum glucose had a statistically significant inverse association with prevalent nonatopic asthma (Table 5). Table 6 shows the final results of the subset analyses for combined AL biomarkers. In summary, boys who possessed high levels of at least two biomarkers (elevated TC, elevated glucose, and abnormal cortisol levels) were 3.0 and 3.3 times more likely to have prevalent and new-onset asthma as adolescents, respectively. In the analysis of asthma phenotypes, having elevated levels of TC and glucose increased the likelihood of having atopic asthma by approximately five times in boys. As with the AL index, elevations in TC (clinical cut-off ≥ 5.2 mmol/ml) (38) and glucose (clinical cut-off ≥ 5.6–6.9 mmol/l) (39, 40) were at subclinical levels. Nonatopic asthma, on the other hand, was 8.2 times more prevalent among boys who had elevated levels of glucose and cortisol. In girls, no biomarker combination predicted asthma in adolescence.
| Prevalent Asthma | ||||
| Asthma (n = 199)* | Atopic Asthma (n = 169)† | Nonatopic Asthma (n = 144)‡ | Incident Asthma (n = 140)§ | |
| Systolic blood pressure | 1.48 (0.49–4.46) | 1.05 (0.24–4.53) | 0.98 (0.22–4.25) | 1.73 (0.46–6.54) |
| Diastolic blood pressure | 1.38 (0.49–3.86) | 3.22 (0.79–13.2) | 0.62 (0.11–3.32) | 1.13 (0.35–3.64) |
| Waist-to-hip ratio | 1.78 (0.48–6.44) | 0.44 (0.06–3.29) | 2.71 (0.62–11.9) | 2.49 (0.61–10.1) |
| Glucose | 2.18 (0.86–5.50) | 3.07 (0.83–11.3) | 3.35 (0.95–11.9) | 1.86 (0.65–5.28) |
| Total cholesterol | 2.26 (0.88–5.78) | 2.30 (0.64–8.22) | 1.54 (0.41–5.79) | 2.03 (1.03–8.92)** |
| HDL | 1.22 (0.44–3.42) | 0.64 (0.15–2.79) | 2.16 (0.55–8.42) | 1.28 (0.37–4.41) |
| DHEA-S | 0.83 (0.32–2.19) | 0.93 (0.26–3.31) | 0.61 (0.13–2.92) | 0.88 (0.27–2.88) |
| Cortisol | 2.25 (0.83–6.10) | 1.92 (0.19–4.34) | 3.89 (1.00–15.5)** | 1.58 (0.52–4.83) |
| Prevalent Asthma | ||||
| Asthma (n = 153)* | Atopic Asthma (n = 122)† | Nonatopic Asthma (n = 110)‡ | Incident Asthma (n = 119)‖ | |
| Systolic blood pressure | 0.60 (0.20–1.85) | 1.58 (0.32–7.68) | 0.49 (0.11–2.27) | 1.02 (0.48–4.44) |
| Diastolic blood pressure | 0.74 (0.26–2.13) | 0.75 (0.16–3.52) | 0.65 (0.18–2.41) | 0.89 (0.32–2.48) |
| Waist-to-hip ratio | 0.63 (0.17–2.33) | 0.95 (0.16–5.56) | 0.53 (0.10–2.72) | 0.49 (0.14–1.75) |
| Glucose | 0.44 (0.15–1.23) | 1.32 (0.34–5.20) | 0.10 (0.02–0.65)** | 0.41 (0.13–1.21) |
| Total cholesterol | 1.21 (0.43–3.39) | 2.24 (0.60–8.36) | 0.82 (0.20–3.46) | 1.19 (0.51–3.37) |
| HDL | 1.01 (0.32–3.19) | 0.61 (0.12–2.99) | 0.72 (0.16–3.20) | 0.69 (0.21–2.34) |
| DHEA-S | 1.08 (0.38–3.05) | 0.88 (0.22–3.49) | 1.58 (0.43–5.80) | 0.80 (0.26–2.41) |
| Cortisol | 0.63 (0.18–2.16) | 0.60 (0.07–4.85) | 2.22 (0.29–17.1) | 0.62 (0.16–2.39) |
| Insulin | 2.57 (0.65–10.3) | 0.67 (0.10–4.49) | 13.7 (1.27–14.8)** | 1.64 (0.49–5.44) |
| Crude OR | Adjusted OR | ||
| Prevalent asthma* | Total cholesterol, glucose, and cortisol | 1.67 (0.81–3.45) | 3.02 (1.04–8.78)† |
| Prevalent atopic asthma‡ | Total cholesterol and glucose | 1.58 (0.65–3.82) | 4.76 (1.10–20.5)† |
| Prevalent non-atopic asthma§ | Glucose and cortisol | 3.37 (1.02–11.1)† | 8.23 (1.75–38.7)† |
| Incident asthma‖ | Total cholesterol, glucose, and cortisol | 1.84 (0.64–5.27) | 3.36 (1.02–11.2)† |
In this follow-up study of 352 children, we for the first time report a link between AL and asthma susceptibility in adolescence. Boys who had high AL at 10.5 years of age were four times more likely to have prevalent or incident asthma at 12.5 years of age. Each additional unit increase in AL score increased the likelihood of prevalent asthma by 60% in boys. High AL was associated with atopic and nonatopic asthma in boys. All associations were independent of genetic risk and observed at subclinical elevations of several biomarkers. Upon further testing of the AL biomarker subsets, elevations in the TC and glucose combination were found to increase the likelihood of atopic asthma, whereas the glucose and cortisol subset was associated with nonatopic asthma in adolescent boys. These subset associations were comparable to, if not of a greater magnitude than, those observed with the total AL score. In girls, high AL scores were not associated with prevalent (atopic or nonatopic) or incident asthma.
First introduced by Sterling and Eyer (41), the concept of allostasis was operationalized in the 1990s and established as a measure of the collective impacts of long-term stress exposure (42). The majority of studies on AL have been conducted in adult populations and report on cardiovascular, metabolic, psychological, and senile cognitive outcomes (8). Studies considering AL in pediatric populations are limited and fall under three themes. In one grouping of publications, the AL theory is merely proposed as an explanation for psychological and neuro-developmental outcomes after stressful life circumstances in children and adolescents (43, 44). Comprising a second theme, the AL index is applied as a tool to assess cumulative stress in children (8, 22, 23, 45), demonstrating long-lasting effects of childhood stressors on AL markers continuing into adulthood (46). Finally, a few pediatric studies have examined the AL index in childhood in relation to adverse outcomes, such as school days missed due to illness (47) or impaired working memory in early adulthood (9). Our study builds on this literature to demonstrate that AL, as an index representing incremental effects of long-term stress, is linked to asthma in adolescents.
Isolated AL components, such as impaired cortisol responses (48) and high body adiposity (49), have shown positive associations with asthma in youth. More recently, insulin resistance (50) and abnormal lipid metabolism (51) have been linked to childhood asthma. The metabolic syndrome, a proinflammatory state characterized by central obesity, high blood pressure, dyslipidemia, and insulin resistance (39), has recently been reported in children and adults with asthma (39, 52). The present findings enhance our understanding of the roles played by these biomarkers in adolescent boys with asthma. Initially, we found a 4-fold association with AL scores in the highest quartile, showing that multibody system biomarkers are significantly associated with adolescent asthma. By also testing the AL index as a continuous measure (Table 3), we discovered evidence for a linear association between incremental increases in the AL score and asthma likelihood. Next, results from the incident asthma analysis confirmed that change in the biomarkers preceded asthma onset.
However, examining subset combinations of the AL biomarkers, which included metabolic and cardiovascular risk factors (TC and glucose) and a stress hormone (cortisol) in relation to asthma (Table 6), we documented associations with asthma that were improved over the corresponding biomarkers in isolation (Tables 4 and 5) and that were equally comparable to those of the total AL index (Table 3). These findings suggest that the reported links between total AL index and asthma are attributable to the subset combinations of biomarkers. These results provide evidence for multisystem biomarkers in adolescent-onset asthma that are not solely derived from metabolic, cardiovascular, or HPA axis activity and confer that all AL biomarkers are not uniformly involved in this process.
Multiple interactive biomarkers were linked to asthma in this study. High cholesterol (39, 51) and insulin resistance (39, 50) can promote a proinflammatory state. Lipid and glucose metabolism are also affected by stress. Stress-induced elevations in TC levels have been observed for as long as 3 years after exposure (53). Also, hyperresponsiveness to stress has been prospectively linked with elevated fasting glucose and insulin resistance (54). Chronic sympathetic nervous system overactivity is posited to explain these observations, likely through lipolytic and glycogenolytic effects of epinephrine and norepinephrine and their inhibitory effect on glucose uptake (53, 54). It is equally plausible that elevated levels of TC and glucose in adolescent boys with atopic asthma are the result of stress-induced sympathetic nervous system overactivity. We did not have a direct measure of environmental stressors or the sympathetic nervous system activity to test this explanation. We were able to measure cortisol levels, and our findings point to a direct role for HPA activity and insulin resistance in nonatopic asthma among boys.
Several of the studied biomarkers, including blood pressure, lipid profile, glucose metabolism, and waist-to-hip ratio, can be affected by total body adiposity (40). Because overweight status has been linked to stress-induced unhealthy eating behavior (55) and to asthma in adolescents (49), it may confound the association between AL or a subset of its biomarkers with asthma. When overweight was added as a separate measure to models, the association between AL and atopic asthma was unmasked in boys, revealing an effect of AL on atopic asthma that was independent of body adiposity. In nonatopic asthma, adjustment for overweight also increased the odds ratio for AL but widened the confidence interval to include the value of 1. In contrast, the association between the cortisol-glucose subset of AL and nonatopic asthma remained independent of total body adiposity. Because being overweight is common in adolescents with nonatopic asthma (49), the last finding suggests that overweight confounded the association between nonatopic asthma and components of AL other than cortisol and glucose.
Sex differences were clearly evident in our findings. Consistent with a greater effect of parent stress on the development of wheezy symptoms in boys but not girls (56), AL and its biomarkers were associated with prevalent and incident asthma in boys only. In girls, crude odds ratios indicated a null association between AL and prevalent asthma, but an inverse association was uncovered after adjusting for current asthma. Because pubertal changes start at an earlier age in girls compared with boys, we speculate that fluctuating hormones, such as elevations in the DHEA/cortisol ratio (57) or insulin resistance observed during midpuberty in girls (36, 37, 57), would affect the AL score. Early-onset menarche is speculated to both increase (29) and decrease (58) the risk of developing asthma in adolescent girls. Adjustment for self-reported pubertal stage did not change model results in our study. Thus, puberty, which was present in 60% of study girls, has unclear effects on the AL and asthma association in our study. Because almost 70% of study boys were considered prepubertal, it is unlikely that our findings were influenced by pubertal-induced hormonal changes in boys.
Of the eight individual components of AL, few biomarkers were associated with asthma phenotype; elevated cortisol in boys and insulin levels in girls increased the likelihood for nonatopic asthma. Similar to the overall AL score, a significant inverse association was observed between relatively high glucose levels and prevalent nonatopic asthma in adolescent girls. Initially, this finding appeared to be incongruent with our insulin results and with literature documenting a higher prevalence of insulin resistance among children and adults who have asthma (50, 59). However, in contrast to the parallel rise in glucose and insulin seen in adults before the development of type 2 diabetes, recent studies are consistently showing a slight drop in glucose that is concurrent with insulin resistance in pubertal girls (36, 37). This new literature supports our observations of lower glucose and higher insulin levels with nonatopic asthma in adolescent girls in puberty.
There are a number of strengths to our research: 1) the total number of participants was quite large for the prevalence study, 2) asthma and asthma phenotypes were diagnosed by a pediatric allergist based on established guidelines (17), and 3) allostatic load was measured using biomarkers similar to previous studies in children and adolescents.
A few limitations were also apparent in this study. No direct measure of sympathetic nervous system activity (epinephrine or norepinephrine) was available for the composition of AL score. Also, we did not use a stress questionnaire to identify specific environmental (either psychological or physical) stressors. Evidence behind the theory that the composite AL score and its components are a reflection of long-term stress exposure formed the basis of our study. Without inquiring into stressful life events, we can only speculate that a high AL score and elevated biomarker levels were the outcome of chronic stress. Future study is required to compare and contrast AL biomarkers versus paper and pencil stress measures to predict the development of asthma. Furthermore, the 2-year follow-up time period may be insufficient to evaluate the impact of AL biomarkers. However, our study was conducted over the critical age period when adolescent asthma first develops, and our analyses were adjusted for preexisting asthma. In addition, we also performed an incidence analysis where only new cases of asthma were studied. Larger cohorts of children followed over a longer period of time are needed to investigate the causal association between AL and its biomarkers with new-onset atopic and nonatopic asthma.
Asthma is one of the most common chronic diseases in adolescents and affects the lives of youth and their families (60). The AL concept enabled us to assess a set of multisystem, stress-responsive biomarkers that had not been considered collectively in relation to asthma in this age group. Our results show that even subclinical, elevated levels of stress and metabolic biomarkers increase the likelihood of adolescent asthma and differ by gender and asthma phenotype. Clinicians currently use the Asthma Predictive Index (61) to predict asthma in young schoolchildren, an index not validated in older children and composed of risk factors that are nonmodifiable. AL and its components are amendable to change in later childhood despite originating from stressors that may occur in early life and go undetected. We reported associations with the AL composite measure and a subset of its components. It is possible that the association between AL index and asthma was predominantly derived from a smaller subset of biomarkers. Although this finding would promote convenience and cost savings in the clinical setting, our findings warrant replication in other studies before the AL composite measure of multisystem biomarkers is dismissed.
The authors thank the study families and the SAGE research team. They also thank Y. Zeng for computer analysis support.
| 1. | Akinbami L. The state of childhood asthma, United States, 1980–2005. Adv Data 2006;381:1–24. |
| 2. | Mandhane PJ, Greene JM, Cowan JO, Taylor DR, Sears MR. Sex differences in factors associated with childhood- and adolescent-onset wheeze. Am J Respir Crit Care Med 2005;172:45–54. |
| 3. | Bahreinian S, Ball GD, Colman I, Becker AB, Kozyrskyj AL. Depression is more common in girls with nonatopic asthma. Chest 2011;140:1138–1145. |
| 4. | Wright RJ. Epidemiology of stress and asthma: From constricting communities and fragile families to epigenetics. Immunol Allergy Clin North Am 2011;31:19–39. |
| 5. | Chida Y, Hamer M, Steptoe A. A bidirectional relationship between psychosocial factors and atopic disorders: a systematic review and meta-analysis. Psychosom Med 2008;70:102–116. |
| 6. | Kozyrskyj AL, Mai XM, McGrath P, Hayglass KT, Becker AB, Macneil B. Continued exposure to maternal distress in early life is associated with an increased risk of childhood asthma. Am J Respir Crit Care Med 2008;177:142–147. |
| 7. | Turyk ME, Hernandez E, Wright RJ, Freels S, Slezak J, Contraras A, Piorkowski J, Persky VW. Stressful life events and asthma in adolescents. Pediatr Allergy Immunol 2008;19:255–263. |
| 8. | Juster RP, McEwen BS, Lupien SJ. Allostatic load biomarkers of chronic stress and impact on health and cognition. Neurosci Biobehav Rev 2010;35:2–16. |
| 9. | Evans GW, Schamberg MA. Childhood poverty, chronic stress, and adult working memory. Proc Natl Acad Sci USA 2009;106:6545–6549. |
| 10. | Karlamangla AS, Singer BH, Seeman TE. Reduction in allostatic load in older adults is associated with lower all-cause mortality risk: Macarthur studies of successful aging. Psychosom Med 2006;68:500–507. |
| 11. | Karlamangla AS, Singer BH, McEwen BS, Rowe JW, Seeman TE. Allostatic load as a predictor of functional decline: Macarthur studies of successful aging. J Clin Epidemiol 2002;55:696–710. |
| 12. | Sexton K, Linder SH. Cumulative risk assessment for combined health effects from chemical and nonchemical stressors. Am J Public Health 2011;101:S81–S88. |
| 13. | Gallo LC, Jimenez JA, Shivpuri S, Espinosa de los Monteros K, Mills PJ. Domains of chronic stress, lifestyle factors, and allostatic load in middle-aged Mexican-American women. Ann Behav Med 2011;41:21–31. |
| 14. | Wright RJ. Health effects of socially toxic neighborhoods: the violence and urban asthma paradigm. Clin Chest Med 2006;27:413–421, v. |
| 15. | Bahreinian S, Ball GD, McNeil BJ, HayGlass KT, Becker AB, Kozyrskyj AL. Does chronic stress predict the development of asthma in pre-adolescents? Am J Respir Crit Care Med 2011;183:A1905. |
| 16. | Kozyrskyj AL, HayGlass KT, Sandford AJ, Pare PD, Chan-Yeung M, Becker AB. A novel study design to investigate the early-life origins of asthma in children (SAGE study). Allergy 2009;64:1185–1193. |
| 17. | Becker A, Lemiere C, Berube D, Boulet LP, Ducharme F, Fitzgerald M, Kovesi T. 2003 Canadian asthma consensus guidelines executive summary. Allergy Asthma Clin Immunol 2006;2:24–38. |
| 18. | Goodman E, McEwen BS, Huang B, Dolan LM, Adler NE. Social inequalities in biomarkers of cardiovascular risk in adolescence. Psychosom Med 2005;67:9–15. |
| 19. | Bellingrath S, Weigl T, Kudielka BM. Chronic work stress and exhaustion is associated with higher allostastic load in female school teachers. Stress 2009;12:37–48. |
| 20. | Maloney EM, Boneva R, Nater UM, Reeves WC. Chronic fatigue syndrome and high allostatic load: results from a population-based case-control study in Georgia. Psychosom Med 2009;71:549–556. |
| 21. | Hasson D, Von Thiele Schwarz U, Lindfors P. Self-rated health and allostatic load in women working in two occupational sectors. J Health Psychol 2009;14:568–577. |
| 22. | Evans GW, Kim P, Ting AH, Tesher HB, Shannis D. Cumulative risk, maternal responsiveness, and allostatic load among young adolescents. Dev Psychol 2007;43:341–351. |
| 23. | Evans GW. A multimethodological analysis of cumulative risk and allostatic load among rural children. Dev Psychol 2003;39:924–933. |
| 24. | Miller GE, Chen E, Zhou ES. If it goes up, must it come down? Chronic stress and the hypothalamic-pituitary-adrenocortical axis in humans. Psychol Bull 2007;133:25–45. |
| 25. | Badanes LS, Watamura SE, Hankin BL. Hypocortisolism as a potential marker of allostatic load in children: associations with family risk and internalizing disorders. Dev Psychopathol 2011;23:881–896. |
| 26. | Kwon JW, Kim BJ, Song Y, Seo JH, Kim TH, Yu J, Kim HB, Lee SY, Kim WK, Kim KW, et al.. Changes in the prevalence of childhood asthma in Seoul from 1995 to 2008 and its risk factors. Allergy Asthma Immunol Res 2011;3:27–33. |
| 27. | Caron J, Liu A. A descriptive study of the prevalence of psychological distress and mental disorders in the Canadian population: comparison between low-income and non-low-income populations. Chronic Dis Can 2010;30:84–94. |
| 28. | Kewalramani A, Bollinger ME. The impact of food allergy on asthma. J Asthma Allergy 2010;3:65–74. |
| 29. | Al-Sahab B, Hamadeh MJ, Ardern CI, Tamim H. Early menarche predicts incidence of asthma in early adulthood. Am J Epidemiol 2010;173:64–70. |
| 30. | Juhn YJ, Sauver JS, Katusic S, Vargas D, Weaver A, Yunginger J. The influence of neighborhood environment on the incidence of childhood asthma: a multilevel approach. Soc Sci Med 2005;60:2453–2464. |
| 31. | Marshall WA, Tanner JM. Variations in pattern of pubertal changes in girls. Arch Dis Child 1969;44:291–303. |
| 32. | Marshall WA, Tanner JM. Variations in the pattern of pubertal changes in boys. Arch Dis Child 1970;45:13–23. |
| 33. | Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, Mei Z, Curtin LR, Roche AF, Johnson CL. CDC growth charts: United States. Adv Data 2000;314:1–27. |
| 34. | Mark W. Epidemiology study design and data analysis. New York: Chapman & Hall/CRC; 1999. |
| 35. | Illi S, von Mutius E, Lau S, Niggemann B, Gruber C, Wahn U. Perennial allergen sensitisation early in life and chronic asthma in children: a birth cohort study. Lancet 2006;368:763–770. |
| 36. | Reinehr T, Toschke AM. Onset of puberty and cardiovascular risk factors in untreated obese children and adolescents: a 1-year follow-up study. Arch Pediatr Adolesc Med 2009;163:709–715. |
| 37. | Frontini MG, Srinivasan SR, Berenson GS. Longitudinal changes in risk variables underlying metabolic syndrome x from childhood to young adulthood in female subjects with a history of early menarche: the Bogalusa Heart Study. Int J Obes Relat Metab Disord 2003;27:1398–1404. |
| 38. | National Cholesterol Education Program (NCEP). Highlights of the report of the expert panel on blood cholesterol levels in children and adolescents. Pediatrics 1992;89:495–501. |
| 39. | Del-Rio-Navarro BE, Castro-Rodriguez JA, Garibay Nieto N, Berber A, Toussaint G, Sienra-Monge JJ, Romieu I. Higher metabolic syndrome in obese asthmatic compared to obese nonasthmatic adolescent males. J Asthma 2010;47:501–506. |
| 40. | Zimmet P, Alberti KG, Kaufman F, Tajima N, Silink M, Arslanian S, Wong G, Bennett P, Shaw J, Caprio S. The metabolic syndrome in children and adolescents: an IDF consensus report. Pediatr Diabetes 2007;8:299–306. |
| 41. | Sterling P, Eyer J. Handbook of life stress, cognition and health. New York: John Wiley & Sons; 1988. |
| 42. | Stewart JA. The detrimental effects of allostasis: allostatic load as a measure of cumulative stress. J Physiol Anthropol 2006;25:133–145. |
| 43. | Grassi-Oliveira R, Ashy M, Stein LM. Psychobiology of childhood maltreatment: effects of allostatic load? Rev Bras Psiquiatr 2008;30:60–68. |
| 44. | Worthman CM, Costello EJ. Tracking biocultural pathways in population health: the value of biomarkers. Ann Hum Biol 2009;36:281–297. |
| 45. | Worthman CM, Panter-Brick C. Homeless street children in Nepal: use of allostatic load to assess the burden of childhood adversity. Dev Psychopathol 2008;20:233–255. |
| 46. | Danese A, Moffitt TE, Harrington H, Milne BJ, Polanczyk G, Pariante CM, Poulton R, Caspi A. Adverse childhood experiences and adult risk factors for age-related disease: depression, inflammation, and clustering of metabolic risk markers. Arch Pediatr Adolesc Med 2009;163:1135–1143. |
| 47. | Johnston-Brooks CH, Lewis MA, Evans GW, Whalen CK. Chronic stress and illness in children: the role of allostatic load. Psychosom Med 1998;60:597–603. |
| 48. | Dreger LC, Kozyrskyj AL, HayGlass KT, Becker AB, MacNeil BJ. Lower cortisol levels in children with asthma exposed to recurrent maternal distress from birth. J Allergy Clin Immunol 2010;125:116–122. |
| 49. | Visness CM, London SJ, Daniels JL, Kaufman JS, Yeatts KB, Siega-Riz AM, Calatroni A, Zeldin DC. Association of childhood obesity with atopic and nonatopic asthma: results from the National Health and Nutrition Examination Survey 1999–2006. J Asthma 2010;47:822–829. |
| 50. | Arshi M, Cardinal J, Hill RJ, Davies PS, Wainwright C. Asthma and insulin resistance in children. Respirology 2010;15:779–784. |
| 51. | Cottrell L, Neal WA, Ice C, Perez MK, Piedimonte G. Metabolic abnormalities in children with asthma. Am J Respir Crit Care Med 2011;183:441–448. |
| 52. | Agrawal A, Mabalirajan U, Ahmad T, Ghosh B. Emerging interface between metabolic syndrome and asthma. Am J Respir Cell Mol Biol 2011;44:270–275. |
| 53. | Steptoe A, Brydon L. Associations between acute lipid stress responses and fasting lipid levels 3 years later. Health Psychol 2005;24:601–607. |
| 54. | Flaa A, Aksnes TA, Kjeldsen SE, Eide I, Rostrup M. Increased sympathetic reactivity may predict insulin resistance: an 18-year follow-up study. Metabolism 2008;57:1422–1427. |
| 55. | Martyn-Nemeth P, Penckofer S, Gulanick M, Velsor-Friedrich B, Bryant FB. The relationships among self-esteem, stress, coping, eating behavior, and depressive mood in adolescents. Res Nurs Health 2009;32:96–109. |
| 56. | Milam J, McConnell R, Yao L, Berhane K, Jerrett M, Richardson J. Parental stress and childhood wheeze in a prospective cohort study. J Asthma 2008;45:319–323. |
| 57. | Belgorosky A, Baquedano MS, Guercio G, Rivarola MA. Expression of the IGF and the aromatase/estrogen receptor systems in human adrenal tissues from early infancy to late puberty: implications for the development of adrenarche. Rev Endocr Metab Disord 2009;10:51–61. |
| 58. | Meyer HW, Wurtz H, Suadicani P, Valbjorn O, Sigsgaard T, Gyntelberg F. Molds in floor dust and building-related symptoms among adolescent school children: a problem for boys only? Indoor Air 2005;15:17–24. |
| 59. | Al-Shawwa BA, Al-Huniti NH, DeMattia L, Gershan W. Asthma and insulin resistance in morbidly obese children and adolescents. J Asthma 2007;44:469–473. |
| 60. | Farnik M, Pierzchala W, Brozek G, Zejda JE, Skrzypek M. Quality of life protocol in the early asthma diagnosis in children. Pediatr Pulmonol 2010;45:1095–1102. |
| 61. | Castro-Rodriguez JA, Holberg CJ, Wright AL, Martinez FD. A clinical index to define risk of asthma in young children with recurrent wheezing. Am J Respir Crit Care Med 2000;162:1403–1406. |
Supported by AllerGen NCE, Inc., and the Canadian Institutes of Health Research new emerging team and operating grants, by a studentship from the Women and Children’s Health Research Institute in Edmonton (S.B.), by Population Health Investigator Awards from Alberta Innovates–Health Solutions and New Investigator Awards from CIHR (G.D.C.B. and I.C.), and by the Women and Children’s Health Research Institute/Stollery Children's Hospital Foundation (A.L.K.).
Author Contributions: S.B. had complete access to all study data and assumes responsibility for its integrity and accuracy. A.L.K. and G.D.C.B. were cosupervisors for S.B.; made substantial contributions to conception and design, acquisition of data, analysis, and interpretation of data; revised the submitted article critically for important intellectual content; and provided final approval of the version to be published. T.K.V.L. was a thesis committee member for S.B.; made substantial contributions to the analysis and interpretation of data; revised the submitted article critically for important intellectual content; and provided final approval of the version to be published. I.C. was a thesis committee member for S.B.; made substantial contributions to the analysis and interpretation of data; revised the submitted article critically for important intellectual content; and provided final approval of the version to be published. B.J.M. is one of the investigators of the SAGE study on which the manuscript was based; made substantial contributions to acquisition and interpretation of data; revised the submitted article critically for important intellectual content; and provided final approval of the version to be published. A.B.B. is the principal investigator for the SAGE study on which the manuscript was based; made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data; revised the submitted article critically for important intellectual content; and provided final approval of the version to be published.
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.201201-0025OC on September 6, 2012
Author disclosures are available with the text of this article at www.atsjournals.org.
