Nineteen RCTs (n=1,463 patients, range 12 to 354) were included in the review. The quality of four of 14 trials with hydrochlorothiazide dose 12.5mg to 25 mg were classified as low risk of bias; the rest were classified as high risk of bias. The results of quality assessment for five trials with hydrochlorothiazide dose 50 mg were not reported. Follow-up duration ranged from four to 95 weeks.
Antihypertensive efficacy (14 RCTs)
Hydrochlorothiazide treatment of 12.5mg to 25mg reduced systolic ambulatory blood pressure by 6.5mmHg (95% CI 5.3 to 7.7; I2=0%) and diastolic ambulatory blood pressure by 4.5mmHg (95% CI 3.1 to 6.0; I2=56%).
Hydrochlorothiazide treatment of 12.5mg to 25mg was associated with less reduction in 24-hour ambulatory blood pressure compared with the 24-hour blood pressure reduction of angiotensin-converting enzyme inhibitors (mean blood pressure reduction 12.9/7.7mmHg), angiotensin-receptor blockers (mean blood pressure reduction 13.3/7.8mmHg), beta-blockers (mean blood pressure reduction 11.2/8.5mmHg), and calcium antagonists (mean blood pressure reduction 11.0/8.1mmHg).
Head-to-head comparisons
Hydrochlorothiazide (12.5mg to 25mg) was less efficacious in lowering systolic 24-hour ambulatory blood pressure compared with angiotensin-converting enzyme inhibitors (WMD 4.45, 95% CI 1.75 to 7.16; I2=0%; five RCTs), angiotensin-receptor blockers (WMD 5.13, 95% CI 1.73 to 8.54; I2=69%; seven RCTs), beta-blockers (WMD 6.19, 95% CI 5.07 to 7.32; I2=0%; three RCTs), calcium antagonists (WMD 4.47, 95% CI 0.85 to 8.08; I2=39%; five RCTs).
Hydrochlorothiazide (12.5mg to 25 mg) was less efficacious in lowering diastolic 24-hour ambulatory blood pressure compared with angiotensin-converting enzyme inhibitors (WMD 3.74, 95% CI 1.34 to 6.14; I2=26%; five RCTs), angiotensin-receptor blockers (WMD 2.89, 95% CI 1.10 to 4.68; I2=57%; seven RCTs), beta-blockers (WMD 6.71, 95% CI 5.96 to 7.45; I2=0%; three RCTs), calcium antagonists (WMD 4.16, 95% CI 2.06 to 6.26; I2=41%; five RCTs).
Dose response: There were no significant differences in ambulatory blood pressure reductions between hydrochlorothiazide 12.5mg and 25mg doses. Hydrochlorothiazide dose of 50mg was associated with significant reduction in 24-hour systolic ambulatory blood pressure (WMD 12.0mmHg, 95% CI 8.2 to 15.9; five RCTs) compared with 25mg dose (diastolic blood pressure WMD was 5.4mmHg, 95% CI 3.2 to 7.7).
Results remained unchanged in sensitivity analyses.
No evidence of publication bias was found.
The authors also reported on office blood pressure monitoring, which was not part of the review objective.