O2043: The Effect of Zip Code on Racial Disparities in Exercise Capacity for Pediatric Patients

Background: Peak oxygen consumption (VO2peak) is a marker of cardiorespiratory fitness (CRF). Prior pediatric studies have shown racial differences in CRF with Black patients and Latinx patients having lower VO2peak than Caucasian patients. These studies did not evaluate the effect of local built and social environments on CRF. The aims of this study are to determine whether racial disparities in CRF exist in a large pediatric cohort and to evaluate whether a patient’s locality augments this disparity.

Methods: The institutional cardiopulmonary exercise testing (CPET) database was retrospectively analyzed. All patients  18 years of age without a diagnosis of cardiac disease and with a maximal effort CPET (respiratory exchange ratio  1.10) were included. Patients were divided into three self-identified racial categories: Caucasian, Black, and Latinx. Patients with multiple self-identified races, other races, or those who declined to provide racial identification were excluded. The percent predicted VO2peak from each patient was matched to the patient’s zip code. The percent predicted VO2peak from each zip code was averaged to determine a zip code’s mean CRF. The zip codes were arranged by mean CRF and the top and bottom tertiles were identified. All patients within the top and bottom tertile performing zip codes were separately re-evaluated. Zip codes with 4 or fewer CPETs were excluded from sub-analysis. ANOVA and student’s T test were used to determine significance.

Results: A total of 1,757 patients (12.0% Black, 32.3% Latinx) met inclusion criteria. The racial distribution, mean age, body mass index, and measures of exercise capacity are seen in Table 1. For entire cohort, Caucasian patients had higher percent predicted VO2peak compared to both Black and Latinx patients (p < 0.001 for both comparisons). For the lower performing zip codes, Caucasian and Latinx patients had better predicted VO2peak compared to Black patients, but the difference did not reach statistical significance (p = 0.168 and p = 0.214, respectively) (Figure 1). For the higher performing zip codes, there were no racial differences in predicted VO2peak (p = 1.0 for all comparisons). There are significant differences in racial distribution between the top and bottom tertiles. The proportion of Caucasian patients is much higher in the top than bottom tertile (85.3% vs. 27.0%, p < 0.01) and there were fewer Black and Latinx patients in the top compared to the bottom tertile (5.2% vs. 17.4% and 10.3% vs. 55.6%, respectively, p< 0.01 for both comparisons).

Conclusion: In a large pediatric cohort, Caucasian patients have higher percent predicted VO2peak compared to both Black and Latinx patients. Zip codes with lower predicted VO2peak have significantly more Black and Latinx patients while zip codes with higher predicted VO2peak are predominantly Caucasian. A patient’s local built environment may have a small effect on measures of exercise capacity.