In this sample of 1994 youth with diabetes, the prevalence of asthma was 10.9%. This estimate is slightly higher than population-based prevalence estimates based on self-report, published by the Centers for Disease Control and Prevention in 2001, that 8.7% of all youth younger than 18 years in the United States had asthma.28
When we restricted our sample to youth younger than 18 years at the time of the study visit for comparison purposes (n
= 1891), 10.8% (95% CI: 9.4%–12.2%) had asthma. Consistent with reports from other studies, the higher prevalence of asthma observed in the SEARCH study population may be indicative of an increasing incidence of asthma among youth with diabetes compared with the general population.4,5
We observed a higher prevalence of asthma among youth with type 2 (16.1% [95% CI: 12.0%–20.2%]), than those with type 1 diabetes (10.0% [95% CI: 8.6%–11.4%]). Nonoverlapping CIs around the estimates in each group indicate a significantly higher prevalence of asthma among youth with type 2 diabetes; however, this may be due to the fact that youth receiving medical care for asthma may be more likely to have previously undiagnosed type 2 diabetes recognized. Youth with type 1 diabetes, however, are typically diagnosed at younger ages, irrespective of other comorbidities.
When we examined demographic characteristics associated with asthma, we observed a strong association between race/ethnicity and asthma among youth with type 1 diabetes. The prevalence of asthma was highest among Native American (22.2% [95% CI: 0%–49%]), Asian/Pacific Islander (19.2% [95% CI: 7.9%–30.4%]), and black (17.3% [95% CI: 11.7%–23.0%]) youth and lowest among non-Hispanic white (8.1% [95% CI: 6.6%–9.6%]) youth. Although rates of asthma observed in non-Hispanic white and black subjects were relatively similar to those observed in other large cross-sectional studies conducted among children and adolescents in the United States,29,30
asthma prevalence among Asian/Pacific Islanders was higher. These findings may suggest that rates of asthma among individuals with type 1 diabetes vary according to specific racial/ethnic groups but may also be indicative of higher asthma prevalence reported in specific Asian/Pacific Islander subgroups.31
BMI status was also associated with asthma among youth with type 1 diabetes; nearly 43% of those with asthma were overweight or obese, compared with 32% of those without asthma. The association between BMI and asthma observed in our study is consistent with the findings of several population-based studies of asthma and obesity in children and adolescents.6–8
As expected, a high proportion of youth with type 2 diabetes were overweight or obese (90.6%); although there were marginally more youth who were overweight or obese among those with asthma than those without asthma, the difference was not statistically significant. This is likely due to the strong correlation between childhood obesity and development of type 2 diabetes, which limited the variability of BMI in this group. However, we also observed a significant association between BMI and asthma among youth with type 1 diabetes, which may suggest a role for the inflammatory processes underlying obesity and the development of asthma, among those with autoimmune-related diabetes.
Although the absolute difference was small, hemoglobin A1c was significantly higher in participants with asthma than those without asthma, with and without adjustment for demographic and clinical covariates. We also observed an association between asthma and poor glycemic control among youth with type 1 diabetes. This relationship was significant after adjustment for age and gender but was attenuated after controlling for race/ethnicity and BMI. The SEARCH study previously reported a strong association between glycemic control and race/ethnicity as well as BMI.32
The relationship between use of asthma medications and glycemic control was significant. Youth with type 1 diabetes who had a diagnosis of asthma in the medical record but did not have asthma pharmacotherapy indicated in the medical record or self-reported at the time of the study visit were more likely to have poor glycemic control than those treated for asthma. It is interesting to note that those whose asthma was treated with leukotriene modifiers, alone or in combination with rescue inhalers or other inhaled medications, had the lowest prevalence of poor glycemic control. In fact, 72% of these participants had good glycemic control, representing the highest proportion of good glycemic control among all groups of medication users. A possible explanation for this finding is that these medications block leukotriene synthesis or interfere with leukotriene-receptor binding, thereby reducing airway inflammation and mucus secretion, and improving mucociliary clearance and lung function.33
The anti-inflammatory action of these medications may also ameliorate systemic inflammation present in both obesity and diabetes.
We acknowledge several limitations of this study. The cross-sectional and observational nature of these data did not allow us to assess changes in asthma severity, which may be associated with glycemic control. The medical record abstraction and health questionnaire did not query medication dosages or dispense dates. This lack of information precluded us from evaluating medication adherence or treatment duration, which may have affected glycemic control. In addition, although oral steroids may negatively impact glycemic control, only 6 youth with type 1 diabetes and asthma had oral steroids noted in the medical record; none reported taking oral steroids at the time of the study visit when hemoglobin A1c was measured. As such, it is likely that the impact of oral steroid use on glycemic control was minimal.
Despite extensive efforts to optimize recruitment, only 62% of eligible youth from the 2002–2005 incident cohorts completed the SEARCH study visit. To determine if our prevalence estimates may have been biased by selection into the study, we examined medical records of youth in the 2002–2005 cohorts that are available for all youth regardless of whether they completed a study visit, and identified those with asthma diagnoses and medications. Youth who did not complete a study visit (n = 1307) had an asthma prevalence of 9.3% (95% CI: 7.8%–10.9%), which is comparable to the 9.7% (95% CI: 8.4%–11.0%) prevalence observed among the youth included in these analyses (n = 1994). The proportion of youth with asthma based solely on medical record diagnoses and medication prescriptions is slightly lower than the proportion identified by combination of medical record data and self-report. This is likely due to the time difference between medical record abstraction and the in-person study visit; youth who reported asthma and concomitant asthma medication use on the health questionnaire who did not have a diagnosis of asthma in the medical record, may have developed asthma >6 months after diabetes diagnosis.
The strengths of our study include its large sample of racially/ethnically diverse youth with type 1 and type 2 diabetes, and the ability to assess glycemic control based on hemoglobin A1c values collected using a common protocol and tested at a central laboratory. Unlike many studies of asthma that rely on self-report alone,1,6,7,9,28–31
we were able to ascertain asthma diagnoses and pharmacologic treatment from medical records, supplemented by self-report, to estimate its prevalence and examine its relation to glycemic control. Of those participants who completed the study visit, 99.2% completed the health questionnaire, of which 95% also had medical record data available.