Fifteen trials were included (2,086 participants, range 44 to 524). Twelve trials were classed as low quality (did not report the randomisation process or allocation concealment, blinding was not used or follow-up was not conducted). Three trials were classed as high quality. There were some discrepancies between the text and figures in the paper; these results were taken from the text.
Antihypertensive effects: Decreases in blood pressure after four, six, eight, 32 and 40 weeks of treatment were not significantly different between (S)-amlodipine and racemic amlodipine. No significant differences were seen at four and eight weeks when only high-quality trials were included in the analysis. Funnel plots indicated some evidence of publication bias at four and eight weeks, but the numbers of trials on each plot were small.
Changes in heart rate: After four weeks of treatment (S)-amlodipine was associated with a smaller change in heart rate compared with racemic amlodipine (MD -0.99, 95% CI -1.70 to -0.28; four trials with I2=0%). There was no difference after eight weeks. No significant difference between treatments was seen in the single high-quality trial at four and eight weeks.
Adverse events: Fewer people experienced an adverse event with (S)-amlodipine compared with racemic amlodipine (RD -0.04, 95% CI -0.06 to -0.02; 13 trials with I2=0%). When only the two high-quality trials were included no difference was seen.
Incidence of oedema was significantly lower with (S)-amlodipine than with racemic amlodipine (RD -0.02, 95% CI -0.03 to 0.00; 13 trials with I2=0%). When only the two high-quality trials were included no difference was seen. There was some evidence of publication bias for oedema.
One high-quality trial found a significant increase in aspartate and alanine aminotransferase with (S)-amlodipine compared with racemic amlodipine (RD 0.08, 95% CI 0.01 to 0.05; error in the paper's results).
There were no differences between treatments for 13 other types of adverse events including flushing and headache.