Thirty-three RCTs were included.
In the pooled analysis, there was statistically-significant heterogeneity for the pooled estimate of diastolic blood-pressure (chi-squared=60.42, d.f.=39, p=0.02) but not for the pooled estimate of systolic blood-pressure (chi-squared=47.54, d.f.=37, p=0.11).
The pooled analysis of all studies showed a statistically-significant reduction in systolic blood-pressure of -1.27 mmHg, (95% confidence interval, CI: -0.92, -0.29, p=0.01) but a non significant reduction in diastolic blood-pressure of -0.24 mmHg (95% CI: -0.92, 0.44, p=0.49).
In the 6 trials that classified participants as either normotensive or hypertensive, there were differences in treatment effects between the 2 groups. The pooled estimates showed reductions of -4.30 mmHg (95% CI: -6.47, -2.13, p<0.001) and -1.50 mm Hg (95% CI: -2.77, 0.23, p=0.02) in systolic and diastolic blood-pressure, respectively, in hypertensive patients; in normotensive patients the reductions were not statistically significant.
Regression analysis was used to explore the possible relationship between baseline blood-pressure and response to calcium supplementation. For both systolic and diastolic blood-pressure, the mean change in baseline blood-pressure revealed no relationship.