Defined as the number of days from start of treatment up to the first date when an asthma exacerbation becomes serious. A serious asthma exacerbation was one that resulted in hospitalization, intubation or death.

Full analysis set: all randomized patients who received at least one dose of study medication.

The number of patients with at least one serious asthma exacerbation over the course of the study. A serious asthma exacerbation was one that resulted in hospitalization, intubation or death.

Full analysis set: all randomized patients who received at least one dose of study medication.

Number of patients requiring treatment with systemic corticosteroids (oral or parenteral) over the course of the study (up to 21 months).

Full analysis set: all randomized patients who received at least one dose of study medication.

The criterion for asthma worsening were: decrease in peak expiratory flow (PEF) >= 20% from mean baseline on >= 3 consecutive days, nighttime symptom score >= 2 on >= 2 consecutive nights, decrease in forced expiration volume in 1 second (FEV1) >=20% from baseline at evening visits, daytime symptom score of 3 or 4 on >= 2 consecutive days, requiring an urgent unscheduled visit for medical care, 24 hour rescue medication use >= 8 puffs on >= 2 consecutive days, and any other clinically important symptoms (pre-specified MedDRA preferred terms).

Full analysis set: all randomized patients who received at least one dose of study medication.

Spirometry was conducted according to internationally accepted standards. Trough FEV1 was measured 15 minutes before dosing; measurements within 6 hours of rescue medication use were set to missing. Repeated measures of analysis of covariance model: change from baseline to trough FEV1 = treatment + visit + treatment*visit interaction + baseline FEV1 + region + asthma related hospitalization in the last 12 months + asthma worsening in the last 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and final visit were included in this analysis.

Spirometry was conducted according to internationally accepted standards. Change from baseline at final visit. FEV1 data taken within 6 hours of rescue medication was excluded from the analysis. Repeated measures of analysis of covariance model: change from baseline to final visit FEV1 = treatment + visit + treatment*visit interaction + baseline FEV1 + region + asthma related hospitalization in the last 12 months + asthma worsening in the last 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and final visit were included in this analysis.

Spirometry was conducted according to internationally accepted standards. Change from baseline at final visit. FVC data taken within 6 hours of rescue medication was excluded from the analysis. Repeated measures of analysis of covariance model: change from baseline to final visit FVC = treatment + visit + treatment*visit interaction + baseline FVC + region + asthma related hospitalization in the last 12 months + asthma worsening in the last 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and final visit were included in this analysis.

PEF was performed every morning and evening prior to study medication use except evenings on the day of clinic visits.

Baseline is average over the last 14 days prior to start of treatment. Analysis of covariance model: change from baseline in PEF = treatment + baseline PEF + region + history of asthma related hospitalization in the past 12 months + history of asthma worsening in the past 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and at final visit were included in this analysis.

Baseline = the last 14 days prior to start of treatment. Analysis of covariance model: Change from baseline = treatment + baseline value + region + history of asthma related hospitalization in past 12 months + history of asthma worsening in past 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and final visit were included in this analysis.

Total asthma symptom score = morning symptoms (0, 1) + daytime score (0-4) + nighttime score (0-4). The range is from 0 to 9. A lower number indicates improvement. Baseline = the last 14 days prior to start of treatment.

Analysis of covariance model: Change from baseline = treatment + baseline value + region + history of asthma related hospitalization in past 12 months + history of asthma worsening in past 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and endpoint were included in this analysis.

24 hours consists of both 12 hour daytime and 12 hour nighttime. Baseline = the last 14 days prior to start of treatment. Analysis of covariance model: Change from baseline = treatment + baseline value + region + history of asthma related hospitalization in past 12 months + history of asthma worsening in past 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and final visit were included in this analysis.

The Asthma Control Questionnaire score ranges from 0 (good control of asthma) to 6 (poor control of asthma). A negative change in score indicates improvement in asthma control.

Repeated measures of analysis of covariance model: change from baseline in ACQ score = treatment + visit + treatment*visit interaction + baseline ACQ score + region + asthma related hospitalization in the last 12 months + asthma worsening in the last 12 months + African American patient.

Full analysis set: all randomized patients who received at least one dose of study medication. Participants with observations at both baseline and final visit were included in this analysis.