METHODS

Study design

Details of the randomised double blind antioxidant supplementation study have recently been published.² In brief, 158 asthmatic children were recruited through the allergy clinic of a large public hospital in Mexico City and followed over a 12 week period with two spirometric tests per week. At baseline, parents completed an interviewer administered questionnaire on diet and risk factors for asthma. Randomisation to receive the supplement (vitamin C 250 mg/day and vitamin E 50 mg/day) or placebo was conducted in a double blind manner. Children provided blood samples for measurements of vitamins C and E at baseline and at several points during the follow up period to verify compliance.² Pulmonary function was measured at each visit with a spirometer (Medifacts Pneumotachograph, San Clemente, CA, USA) according to American Thoracic Society specifications. The best of three technically acceptable manoeuvres was used. Institutional review boards of the collaborating institutions approved the study and parents gave written informed consent for their child’s participation.

Measurements of ambient ozone and climate variables were obtained from the Mexican government’s air monitoring stations. All children resided within 5 km of a monitoring station whose values were assigned to that child. We also classified exposure to particles less than 10 μm and nitrogen dioxide but we found no evidence for an effect of antioxidant intervention in response to these agents² and thus they are not considered in this analysis.

DNA was extracted from whole blood or buffy coat using a standard Gentra Puregene protocol (Gentra Systems, Minneapolis, MN, USA). GSTM1 genotype was determined as previously described.⁶ The method distinguishes subjects with homozygous deletion of GSTM1 (GSTM1 null genotype) from those with either one or two copies of the gene (GSTM1 positive genotype).

Statistical analysis

The effect of ozone exposure on the day before spirometric tests was examined using the maximum 1 hour concentration.² Generalised estimating equation models (GEEs) were used to assess the relation between ozone exposure and changes in lung function tests (FEF_25–75) by GSTM1 genotype (null versus positive) within the two treatment groups (placebo and antioxidant supplemented). Confounding was evaluated as previously described.²

We first compared the effect of ozone on lung function by GSTM1 genotype within the two supplementation groups, and then compared the effect of supplementation on the relation between ozone and lung function by genotype. Regression coefficients were compared using a t test to estimate the modifying effect of GSTM1 polymorphisms/or supplementation groups on ozone mediated responses. We also tested the significance of interaction terms between genotypes, supplementation groups, and ozone. These analyses were conducted in all children and in the subsample of children with moderate and severe asthma. All analyses used Stata Release 7.0 (Stata Corporation, College Park, TX, USA).

RESULTS

The GSTM1 null genotype was present in 39% of children. Although more children with moderate and severe asthma had a GSTM1 null genotype (p = 0.03), baseline pulmonary function did not differ and other characteristics were balanced between the two GSTM1 genotype groups (male 63.5% versus 66.1%; mean age 9.2 years in both groups; mother smoking 17.0% versus 18.3%). The number of spirometric tests per child was similar between the supplement and placebo groups (mean 22, range 12–28). All but three participants had at least one positive skin test for common antigens. The 1 hour maximum ozone level during the study period ranged from 12 to 309 ppb with a mean (SD) of 102 (47) ppb.

In the placebo group ozone levels were significantly and inversely associated with FEF_25–75 in children with the GSTM1 null genotype; no significant decrement was observed in GSTM1 positive children (table 1). Expressed as a percentage of baseline FEF_25–75, for an increase of 50 ppb in 1 h maximum ozone concentration, this decrement in children with the GSTM1 null genotype taking placebo corresponded to −2.9% (95% CI −5.2 to -0.6, p = 0.01). This effect was stronger in children with moderate to severe asthma (−4.7%, 95% CI -7.7 to −1.7; p = 0.002). The difference in the relation between ozone and FEF_25–75 by genotype among all asthmatic children receiving placebo was of borderline statistical significance (p = 0.10). Children receiving antioxidant supplementation had no statistically significant ozone related decrement in FEF_25–75, regardless of genotype. When we considered the effect of antioxidant supplementation by genotype, the beneficial effect was seen primarily in the GSTM1 null individuals (difference of 2.7% between the placebo and supplement groups, p = 0.09) rather than in the GSTM1 positive children (difference of 0.9% between placebo and supplement groups, p = 0.49). Among GSTM1 null children with moderate and severe asthma, the effect of supplementation was enhanced (4.4%, p = 0.04; table 2). In a general model, interactions between genotypes and ozone and supplement group and ozone were marginally significant (p = 0.14 and p = 0.093, respectively).

DISCUSSION

The findings are consistent with the role of GSTM1 in oxidative stress. GSTM1, along with other GSTs, appears to be critical in protecting against oxidative stress by catalysing the conjugation of glutathione to reactive oxygen species.³GSTM1 null genotype appears to contribute to increased levels of biomarkers of oxidative stress after ozone exposure.⁴ In our population, as in other populations of Mexican descent,^5,7 nearly 40% of the children had a GSTM1 null genotype. However, given our sample size, we could not stratify by other genotypes that might interact with GSTM1 in relation to the response to ozone. A larger intervention study would be necessary for this type of analysis.

Our data provide preliminary evidence that asthmatic children who may be genetically impaired in their ability to handle oxidative stress, by virtue of deletion of the GSTM1 gene, are more susceptible to the impact of ozone exposure on small airways function. Furthermore, supplementation with the antioxidants vitamins C and E above the minimum daily requirement might compensate for this genetic susceptibility.