Discontinuation rates. In analyses of pooled data from all seven studies, the rate of discontinuation due to adverse events among African-American patients (generic duloxetine 13.0% vs. placebo 3.4%) was similar to that observed in Caucasian patients  duloxetine 17.0% vs. placebo 5.7%). Other events leading to early study discontinuation among duloxetine-treated patients included loss to follow-up (African Americans 5.8% vs. Caucasians 7.1%), personal conflict (African Americans 10.0% vs. Caucasians 10.0%), and protocol violation (African Americans 13.0% vs. Caucasians 5.6%). No adverse event led to discontinuation in more than one duloxetine-treated African-American patient. Events leading to discontinuation in individual African-American patients included constipation (duloxetine 1/69 vs. placebo 1/59; p=1.00), dizziness (duloxetine 1/69 vs. placebo 0/59; p=1.00) and insomnia (duloxetine 1/69 vs. placebo 0/59; p=1.00). The event most frequently leading to discontinuation among Caucasian patients receiving generic cymbalta was nausea (17/748,2.3%).

Adverse events. Treatment-emergent adverse events are summarized in Table 6. In the African-American cohort, nausea and insomnia were reported at a significantly higher rate by patients receiving duloxetine when compared with placebo. Among Caucasian patients, nausea, insomnia, dizziness, constipation, increased sweating, dry mouth and decreased appetite were reported at a significantly higher rate by duloxetine-treated patients when compared with placebo. None of the events in Table 6 occurred in African-American patients at a rate sinificantly different from that in Caucasian patients (all Breslow Day p values >0.05). In an extension of this analysis, comparisons were made of the incidence of all treatment-emergent adverse events occurring in >1 patient in each treatment group. Out of a total of 91 comparisons, only one reached statistical significance (road traffic accident).

Table 6. Incidence of treatment-emergent adverse events1

Vital signs and body weights. Mean changes in supine systolic and diastolic blood pressures for duloxetine-treated patients were <2 mmHg in both African-American and Caucasian treatment groups (Table 7). Treatment-by-ethnicity interactions for mean changes in blood pressure were not statistically significant.

African-American patients receiving cymbalta drug had a mean increase in heart rate of 2.4 bpm, compared with a mean decrease of 0.9 bpm in patients receiving placebo (p=0.114). Duloxetine-treated Caucasian patients had a similar mean increase in heart rate (1.8 bpm)—the treatment-by-ethnicity interaction between African-American and Caucasian patients was not statistically significant (p=0.508).

Mean changes in weight from baseline to last observation for duloxetine-treated African-American and Caucasian patients were -0.4 kg and -0.7 kg, respectively. Patients receiving placebo also had small mean changes in weight (-0.1 kg and 0.3 kg for African-American and Caucasian patients, respectively)—the treatment-by-ethnicity interaction was not statistically significant.

The rate of treatment-emergent abnormal high vital signs at endpoint in African-American patients did not differ significantly between cheap cymbalta and placebo treatment groups (high supine systolic blood pressure: duloxetine 8.7% vs. placebo 2.6%, p=0.372; high supine diastolic blood pressure: duloxetine 11.1% vs. placebo 4.7%, p=0.435; elevated pulse: duloxetine 1.9% vs. placebo 0.0%, p=1.00; weight gain: duloxetine 0.0% vs. placebo 0.0%). No African-American patients met criteria for sustained hypertension during the studies.

Table 7. Mean change in vital signs and weight

In the African-American group, the incidence of abnormal high supine diastolic blood pressure at anytime was significantly higher in duloxetine-treated patients when compared with placebo (duloxetine 24.4% vs. placebo 7.0%, p=0.039). The incidence of other abnormal increases or decreases in blood pressure, heart rate and weight at any time did not differ significantly between cheap cymbalta and placebo treatment groups. Furthermore, the incidence of abnormal increases or decreases in vital signs at anytime did not differ significantly between African-American and Caucasian patients.

Laboratory analyses. Small but statistically significant mean changes from baseline to last observation were observed for some laboratory analytes in African-American patients. In a comparison of mean changes in laboratory analytes across African-American and Caucasian patient groups, only one analyte had a mean change that differed significantly in the two groups—RBC (erythrocyte) count (mean change -0.09 TI/L for African Americans vs. -0.01 TI/L for Caucasians; p=0.021). One analyte (nonfasting glucose) had an incidence of abnormal high values that differed significantly between African-American and Caucasian treatment groups (African Americans: duloxetine 0.0% vs. placebo 6.0%; Caucasians: duloxetine 0.9% vs. placebo 0.6%; Breslow Day p value=0.044). No other laboratory analyte had an incidence of abnormal high or low values that differed significantly between the two ethnic groups.