- Top of page
Background: It has been hypothesized that obesity and insulin resistance may play a role in the development of asthma and allergy. The aim of the study was to examine the association of obesity and insulin resistance with asthma and aeroallergen sensitization.
Methods: Cross-sectional population-based study of 3609 Danish men and women aged 30–60 years. Aeroallergen sensitization was defined as positive levels of specific IgE against a panel of inhalant allergens. Asthma was defined as self-reported physician diagnosed asthma. Allergic asthma was defined as the presence of both asthma and aeroallergen sensitization. The homeostasis model assessment of insulin resistance was used to estimate the degree of insulin resistance. Body mass index, waist-to-hip ratio, and waist circumference were used as measures of obesity. Data were analyzed by multiple logistic regression analyses.
Results: Obesity was associated with increased risk of aeroallergen sensitization as well as allergic and nonallergic asthma. Insulin resistance was asssociated with aeroallergen sensitization and allergic asthma, but not nonallergic asthma. The associations of obesity with aeroallegen sensitization and allergic asthma became nonsignificant after adjustment for insulin resistance, whereas the association of obesity with nonallergic asthma was unaffected. No sex-differences were observed.
Conclusion: Obesity may be related to an increased risk of aeroallergen sensitization and allergic asthma through mechanisms also involved in the development of insulin resistance.
The prevalence of aeroallergen sensitization (atopy) and asthma has increased in adults during recent decades (1–3). The specific factors responsible for this increase are unknown. Changes towards a westernized, urban, and affluent lifestyle appear to be followed by an increasing prevalence of aeroallergen sensitization and atopic diseases such as asthma (4–7). Research on the causes of this epidemic has to a large extent been focusing on environmental factors, such as exposure to allergens, microorganisms, and air pollution. However, in most countries the increase in aeroallergen sensitization and asthma has been accompanied by a marked increase in overweight and obesity. Furthermore, an increasing number of epidemiological studies support that obesity is associated with increased risk of developing asthma (8–11), but the underlying biological mechanisms are not known. The epidemiological association of obesity with IgE sensitization is less clear (8). Insulin resistance is considered to be a major contributing factor to the close relationship between obesity and the development of type 2 diabetes. Recently, it has been hypothesized that insulin resistance may play a role in the causation of asthma and allergy (12–14). The possible links between insulin resistance and allergic diseases may be that they share some common inflammatory pathways or mediators of immune responses e.g. adipokines and cytokines secreted from adipose tissue (15–17). As yet, there are no epidemiological population-based studies on the relationship of insulin resistance to asthma and allergy.
The aim of this study was to examine the association of obesity measures and insulin resistance with aeroallergen sensitization and asthma in a population-based sample of 3609 Danish men and women aged 30–60 years.
- Top of page
Characteristics of the study population are shown in Table 1. The overall prevalence of aeroallergen sensitization and self-reported asthma was 25.8% and 8.3%, respectively. The prevalence of aeroallergen sensitization and insulin resistance was higher among men than among women, whereas the prevalence of asthma was higher among women than among men. The prevalence of obesity ranged from 7.2% to 23.4% depending on sex and the anthropometric measure used.
Table 1. Characteristics of the study population by sex¶
|Characteristics||All participants (Ntotal = 3609)†||Men (Ntotal = 1743)†||Women (Ntotal = 1866)†|
| Men||48.3 (1743)‡||–||–|
| Younger age (30 + 40 years)||46.3 (1669)‡||43.9 (765)‡||48.5 (904)‡**|
| Low socioeconomic status (social class 1 + 2)||12.4 (409)‡||10.7 (172)‡||14.0 (137)‡**|
| Daily smoker||34.9 (1250)‡||35.8 (619)‡||34.1 (631)‡|
| Body mass index (kg/m2)||25.6 (19.4–37.5)§||26.3 (20.5–36.1)§||24.7 (18.8–39.3)§***|
| Waist-to-hip ratio||0.85 (0.70–1.02)§||0.91 (0.80–1.05)§||0.79 (0.69–0.94)§***|
| Waist circumference (cm)||85 (64–114)§||92 (75–116)§||78 (62–109)§***|
| ObesityBMI (BMI ≥ 30 kg/m2)||17.6 (634)‡||17.8 (310)‡||17.4 (324)‡|
| ObesityWHR (♂: WHR >1.00, ♀: WHR >0.85)||12.4 (448)‡||7.2 (125)‡||17.4 (323)‡***|
| ObesityWC (♂: WC ≥102 cm, ♀: WC ≥88 cm)||21.1 (760)‡||18.7 (325)‡||23.4 (435)‡**|
| HOMA insulin resistance index||1.15 (0.37–4.09)§||1.28 (0.38–4.44)§||1.08 (0.36–3.90)§***|
| Diabetes or insulin resistance (HOMA-IR ≥ 1.77)||29.7 (978)‡||36.0 (572)‡||23.8 (406)‡***|
|Asthma and aeroallergen sensitization|
| Aeroallergen sensitization||25.8 (865)‡||28.0 (468)‡||23.6 (397)‡**|
| Asthma||8.3 (290)‡||6.8 (114)‡||9.8 (176)‡**|
| Allergic asthma||4.5 (146)‡||4.5 (72)‡||4.6 (74)‡|
| Nonallergic asthma||3.7 (121)‡||2.1 (34)‡||5.3 (87)‡***|
Tables 2 and 3 shows that the prevalence of aeroallergen sensitization and asthma was higher in the obese and in the insulin resistant participants compared to the nonobese and noninsulin resistant participants, respectively. Additionally, obesity (BMI, WHR, and WC) and insulin resistance were associated with a significantly higher risk of aeroallergen sensitization as well as asthma when examined by multiple regression analyses (Tables 2 and 3). Adjustment for age, sex, smoking, and socioeconomic status did not change these associations substantially.
Table 2. Risk of aeroallergen sensitization associated with obesity measures and insulin resistance
|Risk factors||Prevalence of aeroallergen sensitization % (n/ntotal) ||Risk of aeroallergen sensitization OR (95% CI)|
|Obesity measures||Crude||+Confounders||+Confounders + HOMA-IR|
|Body mass index|
| Underweight (<18.5 kg/m2)||18.9 (7/37)||0.90 (0.39–2.11)*||0.92 (0.39–2.15)†|
| Normal (≥18.5–25 kg/m2)||23.4 (309/1318)||1.00||1.00|
| Overweight (≥25–30 kg/m2)||27.0 (317/1176)||1.18 (0.98–1.42)*||1.13 (0.93–1.36)†|
| Obese (≥30 kg/m2)||28.7 (154/537) N = 3068¶||1.33 (1.05–1.68)* Ptrend = 0.01||1.11 (0.86–1.44)† Ptrend = 0.28|
|BMI continuous (per 5 kg/m2)|| ||1.10 (1.00–1.20)* P = 0.04||1.01 (0.91–1.12)† P = 0.90|
| Normal||25.1 (674/2687)||1.00||1.00|
| Obese (♂≥1.00, ♀≥0.85)||29.8 (111/373) N = 3060¶||1.55 (1.20–1.98)* P = 0.00||1.38 (1.07–1.80)† P = 0.02|
| WHR continuous (per 0.1)|| ||1.31 (1.15–1.51)* P = 0.00||1.22 (1.06–1.41)† P = 0.01|
| Normal||24.7 (427/1731)||1.00||1.00|
| Overweight (♂: ≥94–102 cm, ♀: ≥80–88 cm)||26.6 (183/689)||1.15 (0.93–1.41)*||1.09 (0.88–1.34)†|
| Obese (♂: ≥102 cm, ♀: ≥88 cm)||27.3 (175/642) N = 3062¶||1.25 (1.01–1.55)* Ptrend = 0.03||1.06 (0.84–1.34)† Ptrend = 0.94|
| WC continuous (per 5 cm)|| ||1.05 (1.01–1.09)* P = 0.01||1.02 (0.98–1.06)† P = 0.45|
|Insulin resistance||Crude||+Confounders||+Confounders + BMI|
|HOMA insulin resistance index|
| Noninsulin resistant||23.6 (510/2165)||1.00||1.00|
| Insulin resistant (HOMA-IR ≥ 1.77 μU/ml)||30.9 (222/718)||1.37 (1.13–1.66)*||1.32 (1.07–1.63)‡|
| Diabetic§||29.7 (55/185) N = 3068¶||1.59 (1.13–2.25)* Ptrend = 0.00||1.55 (1.08–2.21)‡ Ptrend = 0.00|
| HOMA-IR continuous (per 1.0)|| ||1.15 (1.06–1.24)* P = 0.00||1.15 (1.05–1.25)‡ P = 0.00|
Table 3. Risk of asthma associated with obesity measures and insulin resistance
|Risk factors||Prevalence of asthma % (n/ntotal)||Risk of asthma OR (95% CI)|
|Obesity measures||Crude||+Confounders||+Confounders + HOMA-IR|
|Body mass index|
| Underweight (<18.5 kg/m2)||11.1 (4/36)||1.31 (0.45–3.82)*||1.33 (0.46–3.88)†|
| Normal (≥18.5–25 kg/m2)||8.0 (109/1364)||1.00||1.00|
| Overweight (≥25–30 kg/m2)||6.9 (83/1207)||0.92 (0.68–1.25)*||0.87 (0.64–1.19)†|
| Obese (≥30 kg/m2)||12.3 (69/561) N = 3168¶||1.61 (1.16–2.23)* Ptrend = 0.02||1.38 (0.95–2.00)† Ptrend = 0.25|
| BMI continuous (per 5 kg/m2)|| ||1.24 (1.10–1.40) P = 0.00||1.20 (1.04–1.38) P = 0.01|
| Normal||7.6 (211/2769)||1.00||1.00|
| Obese (♂≥1.00, ♀≥0.85)||13.6 (53/391) N = 3160¶||1.68 (1.20–2.35)* P = 0.00||1.51 (1.06–2.15)† P = 0.02|
| WHR continuous (per 0.1)|| ||1.26 (1.03–1.54) P = 0.03||1.18 (0.96–1.47) P = 0.12|
| Normal||6.8 (120/1778)||1.00||1.00|
| Overweight (♂: ≥94–102 cm, ♀: ≥80–88 cm)||8.6 (61/711)||1.32 (0.95–1.83)*||1.28 (0.92–1.78)†|
| Obese (♂: ≥102 cm, ♀: ≥88 cm)||12.3 (83/673) N = 3162¶||1.85 (1.37–2.50)* Ptrend = 0.00||1.69 (1.20–2.38)† Ptrend = 0.00|
| WC continuous (per 5 cm)|| ||1.10 (1.1.05–1.16) P = 0.00||1.09 (1.03–1.16) P = 0.00|
|Insulin resistance||Crude||+Confounders||+Confounders + BMI|
|HOMA insulin resistance index|
| Noninsulin resistant||7.4 (166/2232)||1.00||1.00|
| Insulin resistant (HOMA-IR ≥1.77 μU/ml)||10.6 (79/745)||1.52 (1.14–2.03)*||1.35 (0.98–1.87)‡|
| Diabetic§||10.5 (20/191) N = 3168¶||1.39 (0.83–2.30)* Ptrend = 0.01||1.20 (0.71–2.05)‡ ptrend = 0.17|
| HOMA-IR continuous (per 1.0)|| ||1.14 (1.02–1.27) P = 0.02||1.07 (0.94–1.21) P = 0.32|
However, in analyses with aeroallergen sensitization as outcome, the effects of obesity measures were attenuated towards the null value and became nonsignificant (except for WHR) after adjusting for insulin resistance (Table 2). In contrast, the association between insulin resistance and aeroallergen sensitization was neither affected by adjustment for BMI (Table 2) nor WHR or WC (data not shown).
In analyses with asthma as outcome, mutually adjustment for insulin resistance and obesity did not seem to affect the associations to a great extent, except for the association between insulin resistance and asthma became statistically nonsignificant (Table 3).
Table 4 shows the results of the regression analyses using allergic and nonallergic asthma as outcomes, respectively. The obesity measures were associated with increased risk of allergic asthma as well as nonallergic asthma. However, only the associations with nonallergic asthma remained statistically significant, while the associations with allergic asthma attenuated toward the null value after adjustment for insulin resistance (Table 4). In contrast, insulin resistance was associated with a higher risk of allergic asthma, but not with nonallergic asthma (Table 4). The association with allergic asthma remained statistically significant after adjustment for BMI (Table 4), Similar results were obtained when including WHR or WC instead of BMI in the model (data not shown).
Table 4. Risk of allergic and nonallergic asthma associated with obesity measures and insulin resistance
|Risk factors||Risk of allergic asthma OR (95% CI)||Risk of nonallergic asthma OR (95% CI)|
|Obesity measures||+Confounders||+Confounders + HOMA-IR||+Confounders||+Confounders + HOMA-IR|
|Body mass index (n = 2974¶)|
| Underweight (<18.5 kg/m2)||1.78 (0.41–7.76)*||1.84 (0.42–8.05)†||1.03 (0.24–4.48)*||1.02 (0.23–4.44)†|
| Normal (18.5–25 kg/m2)||1.00||1.00||1.00||1.00|
| Overweight (≥25–30 kg/m2) ||1.03 (0.67–1.56)*||0.94 (0.61–1.45)†||0.89 (0.56–1.40)*||0.89 (0.56–1.40)†|
| Obese (≥30 kg/m2)||1.73 (1.10–2.72)* Ptrend = 0.05||1.31 (0.78–2.18)† Ptrend = 0.50||1.35 (0.83–2.21)* Ptrend = 0.36||1.38 (0.79–2.42)† Ptrend = 0.44|
| BMI continuous (per 5 kg/m2)||1.23 (1.04–1.46) P = 0.02||1.11 (0.91–1.35) P = 0.30||1.21 (1.02–1.45) P = 0.04||1.27(1.03–1.57) P = 0.03|
|Waist-to-hip ratio (n = 2966¶)|
| Obese (♂≥1.00, ♀≥0.85) ||1.82 (1.14–2.91)* P = 0.02||1.51 (0.92–2.47)† P = 0.11||1.59 (0.99–2.55)* P = 0.06||1.63 (0.99–2.68)† P = 0.06|
| WHR continuous (per 0.1)||1.29 (0.97–1.70) P = 0.08||1.13 (0.84–1.52) P = 0.43||1.24 (0.92–1.67) P = 0.16||1.27 (0.92–1.74) P = 0.15|
|Waist circumference (n = 2968¶)|
| Overweight (♂: ≥94–102 cm, ♀: ≥ 80–88 cm)||1.38 (0.89–2.14)*||1.28 (0.81–2.00)†||1.27 (0.78–2.06)*||1.29 (0.79–2.10)†|
| Obese (♂: ≥102 cm, ♀: ≥88 cm) ||1.78 (1.16–2.71)* Ptrend = 0.01||1.42 (0.88–2.29)† Ptrend = 0.14||1.69 (1.08–2.65)* Ptrend = 0.02||1.85 (1.11–3.08)† Ptrend = 0.02|
| WC continuous (per 5 cm)||1.10 (1.02–1.18) P = 0.01||1.06 (0.98–1.14) P = 0.18||1.10 (1.03–1.18) P = 0.01||1.13 (1.04–1.23) P = 0.01|
|Insulin resistance||+Confounders||+Confounders + BMI||+Confounders||+Confounders + BMI|
|HOMA insulin resistance index (n = 2974¶)|
| Noninsulin resistant||1.00||1.00||1.00||1.00|
| Insulin resistant (≥1.77 μU/ml) ||1.73 (1.17–2.54)*||1.58 (1.03–2.43)‡||1.26 (0.80–1.96)*||1.10 (0.66–1.82)‡|
| Diabetic§||2.01 (1.05–3.86)* Ptrend = 0.00||1.79 (0.90–3.56)‡ Ptrend = 0.03||0.67 (0.26–1.69)* Ptrend = 0.96||0.57 (0.21–1.50)‡ ptrend = 0.49|
| HOMA-IR continuous (per 1.0)||1.22 (1.06–1.40) P = 0.01||1.18 (1.01–1.38) P = 0.05||1.10 (0.92–1.31) P = 0.31||0.98 (0.79–1.23) P = 0.88|
Finally, we tested whether any of the associations of obesity measures and insulin resistance with aeroallergen sensitization and asthma outcomes were modified by sex. No statistically significant interactions were observed (results not shown).
- Top of page
All obesity measures (BMI, WHR, and WC) examined in this study were associated with increased risk of aeroallergen sensitization and asthma including allergic and nonallergic asthma. Insulin resistance was asssociated with aeroallergen sensitization and allergic asthma, but not with nonallergic asthma.
The association between obesity and allergic disease has been examined in several epidemiological studies. In agreement with our results, a positive association between BMI and asthma has relatively consistently been demonstrated in both cross-sectional and longitudinal studies (8, 9). However, most studies on the relation between obesity and allergic disease have considered only BMI and not alternative obesity measures, such as WHR and WC as examined in this study. Although the prevalence of obesity was lower using the WHR criteria compared to the BMI or WC criteria the associations with aeroallergen sensitization and asthma generally seemed to be in the same order of magnitude suggesting that all measures can be used. However, WHR and WC are probably better measures of visceral obesity, which is hypothesized to be of importance regarding insulin resistance. Future studies may consider other measures of obesity, such as body fatness (bio-impedance) and visceral fatness as assessed by ultrasound.
Some studies have reported a stronger relation between BMI and asthma among women than among men. We did not observe any effect modification by sex for any of the obesity measures examined.
Studies examining the relationship of obesity with allergic and nonallergic asthma have been inconsistent. In agreement with our observations, some studies have found an association between obesity and both allergic and nonallergic asthma (24). In contrast, other studies have observed an association between obesity and nonallergic asthma but not allergic asthma (25, 26).
The relation between BMI and aeroallergen sensitization (atopy) has been examined in a few studies, and in contrast with our results, no association or an association only among females was found (8).
To our knowledge this is the first epidemiological population-based study on the relationship of insulin resistance with asthma and aeroallergen sensitization. However, our results are supported by a recent study in morbidly obese children also reporting an association between insulin resistance and asthma (14).
We also reported that the associations of obesity measures with aeroallegen sensitization and allergic asthma (except for WHR and aeroallergen sensitization) were weakened and became nonsignificant after adjusting for insulin resistance, suggesting that obesity is related to aeroallergen sensitization and allergic asthma through mechanisms that may also be involved in the development of insulin resistance.
Obesity and Insulin resistance are causal factors of major importance to the incidence of type 2 diabetes and cardiovascular disease. Cardiovascular disease has been associated with self-reported asthma along with several surrogate measures of allergic disorders, such as total IgE levels, eosinophilia, positive skin prick test results, daily pollen burden, and measurements of lung function (27–31). This supports that asthma, allergy, and cardiovascular disease may share central effector pathways and some common risk factors. At present, cardiovascular disease and type 2 diabetes are considered as low-grade inflammatory diseases (32, 33). Accordingly, asthma and allergy have been associated not only with local, but also with systemic inflammatory markers e.g. C-reactive protein and fibrinogen, and should probably also be considered as systemic inflammatory diseases (15, 16). The possible links between asthma, allergy, cardiovascular disease, and diabetes may be that they share some common inflammatory pathways or mediators of immune responses. Interestingly, it has been proposed that the relationship of obesity to asthma and allergy may be on account of decreased immunological tolerance to allergens as a consequence of immunological changes induced by adipokines, immunological mediators secreted by adipose tissue (17).
Because of the cross-sectional study design, cause, and effect cannot be determined, as the presence of obesity/insulin resistance and asthma/aeroallergen sensitization were determined simultaneously. Thus, reverse causation cannot be excluded. For example, it is possible that subjects with asthma have adopted a more sedentary lifestyle and thus have gained more weight. Additionally, certain medical treatments for asthma e.g. steroids may induce metabolic changes. Unfortunately, we did not have data on asthma medications in the present study. However, in Denmark steroids for asthma are mainly administered by the inhalant route, possibly resulting in only minor systemic effects.
The definition of asthma was based on self-reported asthma, having been diagnosed by a physician anytime earlier. This may have led to some misclassification of asthma. For example, childhood asthma without current asthma may be a specific phenotype with an etiology different from asthma in adulthood. Additionally, other diseases such as chronic obstructive pulmonary disease may have been misclassified as asthma. Unfortunately, specific information on current asthma symptoms was not available in the present study. This is clearly a limitation in the interpretation of the results and future studies should address the role of the asthma phenotype in more detail. However, this type of misclassification is likely to be nondifferential, and thus would tend to attenuate the observed associations toward the null value. Moreover, in the analyses with allergic and nonallergic asthma as outcomes, an association between insulin resistance and asthma was only observed for allergic asthma, and not for nonallergic asthma, indicating that obesity and insulin resistance are in fact related to the allergic phenotype. This finding may also suggest that the effect of insulin resistance on asthma risk is mainly driven by an effect on atopy. However, obesity measures were also strongly associated with nonallergic asthma, which is consistent with misclassification of asthma in the nonallergic asthma group, as respiratory symptoms, such as cough and breathlessness occur more frequently among obese persons than others (34, 35) On the other hand, obesity may be causally related to nonallergic asthma by mechanisms not related to insulin resistance. It has been suggested that the effects of obesity on asthma may be mediated through alterations in airway mechanics, immune responses, hormonal influences, and genetic factors (8).
In conclusion, in this cross-sectional population-based study, obesity, and insulin resistance were associated with an increased risk of aeroallergen sensitization and asthma. These findings raise the hypothesis that obesity is related to aeroallergen sensitization and allergic asthma through mechanisms that may also be involved in the development of insulin resistance. Prospective studies preferably using a more strict asthma diagnosis are nessesary to confirm the results. Also studies that include other metabolic biomarkers e.g. adiponectin, leptin, IL-6 are important to further examine the individual contributions of these markers and insulin resistance to the possible increased risk of allergy and asthma.