Meta-analysis or Systematic Review

M - Click here for explanation of classification scheme.

POSITIVE: See Quality Criteria Checklist below.

To summarize the evidence about the effects of oral calcium supplementation for adult patients with primary hypertension in order to inform decisions about recommendations for treatment.

• Randomized controlled trials (RCTs) comparing oral calcium supplementation with placebo, no treatment or usual care
• Crossover trials restricted to designs with two interventions and two treatment periods
• Treatment and follow-up eight weeks or more, and allowed an intention-to treat analysis
• Participants over 18 years old, with raised systolic blood pressure (SBP) 140mm Hg or higher or diastolic blood pressure (DBP) 85mm Hg or higher, without a known primary cause
• SBP and DBP reported at end of follow-up.

• Pregnancy
• Anti-hypertensive medication that changed during the study
• Calcium supplementation combined with other interventions.

Recruitment

• The following databases were searched for RCTs:
  • Cochrane Library (2004 Issue 1)
  • MEDLINE (2000 to January 9, 2004)
  • EMBASE (1999 to end of 2003)
  • Science Citation Index (1982 to end of 2003)
  • ISI Proceedings (1999 to end of 2003)
  • ClinicalTrials.gov
  • Current controlled trials
  • CAB abstracts (up to end July 2005).
• References lists of systematic reviews, meta-analyses and randomized controlled trials (RCTs) were checked. A general web search using the search engines Google, Zapmeta and Dogpile was done. The websites of the following organizations were also searched:
  • Blood Pressure Association
  • British Hypertension Society
  • American Society of Hypertension
  • Canadian Hypertension Society.
• There was no language restriction.

Design

Meta-analysis.

Intervention

• All titles and abstracts retrieved by electronic searching were screened independently by two reviewers; those studies that did not meet the inclusion criteria were excluded. Copies of the full text of potentially relevant references were obtained and their eligibility was assessed by one of two methods: one reviewer was primary assessor and decisions were checked by a second reviewer, or assessments were done independently by two reviewers. Differences between reviewers were resolved by discussion or by appeal to a third reviewer.
• Methodological quality of included trials was assessed independently by two reviewers based on the following criteria:
  • Blinding
  • Randomization
  • Allocation concealment
  • Loss to follow-up
  • Carryover effects (for crossover trials).
• Two reviewers independently abstracted endpoint data and data that might facilitate interpretation of any heterogeneity in the findings of trials (country in which the study was conducted, inclusion criteria, patient characteristics at baseline, trial quality and duration of follow-up using a pre-specified form). Differences were reconciled by discussion or by consultation with a third reviewer. Authors were contacted for missing endpoint data.

Statistical Analysis

Random effects meta-analyses, sub-group and sensitivity analyses were conducted.

• The findings of all included trials were aggregated in separate meta-analyses for parallel and crossover designs; these were combined in the same meta-analysis if they did not show heterogeneity
• The mean difference (and SDs) between final blood pressure for calcium and control interventions for both SBP and DBP was calculated. If SDs of final values were not available, change scores were used if their SDs were available.
• If trials had more than one treatment arm, a weighted mean of the outcome for all treatment arms was used. Mean differences were weighted according to the precision of each trial and combined in meta-analyses using a random effects model to estimate an overall pooled mean difference and its 95% confidence interval (CI).
• Heterogeneity between trials was assessed using the I² statistic
• Sub-group analyses: Trials were grouped by
  • Active treatment arm received higher (1.2g to 2g per day) and lower (less than 1.2g per day) doses of calcium
  • Participants' mean baseline SBP was higher (SBP higher than 145mm Hg) and lower (SBP of 145mm Hg or less).
• Post-hoc analyses to evaluate effect of calcium supplementation on serum calcium levels.

Dependent Variables

• Primary outcome measures:
  • Death from all causes
  • Coronary heart disease events
  • Cerebrovascular events
  • SBP at end of follow-up
  • DBP at end of follow-up.
• Secondary outcome measures:
  • Total withdrawal from treatment
  • Withdrawals due to adverse effects
  • Reported adverse effects
  • Serum calcium levels at end of follow-up.
• As no trials reported deaths or cardiovascular events, meta-analyses were done only for blood pressure, adverse events, overall withdrawals from treatment and serum calcium levels.

Independent Variables

Oral calcium supplementation, either as tablets or in powdered form, with a mean dose of 1.1g per day (range of 0.4g to 2g per day).

Control Variables

• Blinding
• Randomization
• Allocation concealment
• Loss to follow-up
• Carryover effects.

Initial N

• N=1,577 potentially relevant references identified
• N=48 studies identified for detailed evaluation.

Final N

• A total of 13 RCTs (N=485)
• Five studies were of parallel design (Grobbee 1986; Lyle 1992; Martinez 1989; Nowson 1989; Sanchez 1997) and eight studies were of crossover design (Galloe 1993; Kawano 1998; McCarron 1985; Strazzullo 1986; Takagi 1991; Tanji 1991; Weinberger 1993; Zoccali 1988)
• The number of participants in each trial ranged from nine to 90
• Between eight and 15 weeks follow-up.

Age

The mean age was 45 years (range from 16 years to 86 years).

Ethnicity

• Only two trials reported ethnicity, and in these 88% of the participants were Caucasian
• Seven of the other 11 trials were conducted in Europe or Australia; therefore, it is likely that most of their participants were Caucasian.

Other Relevant Demographics

A total of 65% of the participants were male.

Anthropometrics

• Participants' SBP ranged from 131mm Hg to 169mm Hg and their DBP from 78mm Hg to 99mm Hg. Overall mean blood pressure at baseline was 148/91mm Hg.
• All trials measured BP in the clinic; the trial of Takagi 1991 reported the mean over 24 hours of hourly readings taken while resting in a supine position, whereas other trials reported the mean of up to six readings.
• Only six of the 14 included studies reported the mean dietary calcium intake. This varied from 500mg per day to 1,411mg per day.

Location

• Four trials were conducted in the United States, two in Italy, two in Japan, two in Spain, one in Denmark, one in Australia and one in the Netherlands
• Review of the above studies were done by researchers in UK.

Key Findings

• The results of the individual trials were heterogeneous
• Meta-analysis of five parallel trials found calcium supplements were associated with a statistically significant reduction in SBP compared to control, and a non-significant reduction in DBP. There was no heterogeneity for SBP (I²=0%) and little heterogeneity for DBP (I²=23%).

Comparison 1. Calcium vs. Control (Parallel Trials Only)
 

Outcome	Number of Studies	Number of Participants	Statistical Method	Effect Size
SBP	5	250	Mean difference (IV, random, 95% CI)	-3.22 (-6.19, -0.24)
DBP	5	250	Mean difference (IV, random, 95%CI)	-2.39 (-4.79, 0.01)

 

• Meta-analysis of the eight crossover trials found calcium supplements were associated with a statistically non-significant reduction in SBP compared to control and no effect on DBP. There was considerable heterogeneity for SBP (I²=60%) and DBP (I²=51%).

Comparison 2. Calcium vs. Control (Crossover Trials Only)
 

Outcome	Number of Studies	Number of Participants	Statistical Method	Effect Size
SBP	8	428	Mean difference in SBP (random, 95% CI)	-2.30 (-4.88, 0.28)
DBP	8	428	Mean difference in DBP (random, 95% CI)	-0.28 (-1.66, 1.10)

 

• Combining all trials, participants receiving calcium supplementation as compared to control had a statistically significant reduction in SBP (mean difference of -2.5mm Hg; 95% CI: -4.5 to -0.6; I²=42%), but not DBP (mean difference of -0.8mm Hg; 95% CI: -2.1 to 0.4; I²=48%)
• Sub-group analyses indicated that heterogeneity between trials could not be explained by dose of calcium or baseline blood pressure. Overall, trials administering high (1.2g to 2g per day) and lower (less than 1.2g per day) doses of calcium showed similar changes in blood pressure; likewise for trials in populations with higher (SBP higher than 145mm Hg) and lower (SBP 145mm Hg or lower) mean baseline blood pressure.
• The one trial reporting adequate concealment of allocation (McCarron 1985) yielded results consistent with the primary meta-analysis. A significant reduction in SBP (mean difference of -3.8mm Hg; 95% CI: -6.5 to -1.1) but not DBP (mean difference of 0.0mm Hg; 95% CI: -1.9 to 1.9) was observed.
• Only one trial (Strazzullo 1986) reported blinding of participants, treatment providers and outcome assessors and yielded results consistent with the primary meta-analysis. This crossover trial enrolling 18 participants reported non-significant reductions in both SBP (mean difference of -3.0mm Hg; 95% CI: -7.1 to 1.1) and DBP (mean difference of  -1.0mm Hg; 95% CI: -3.4 to 1.4).
• Heterogeneity was reduced when poor quality trials were excluded.

Other Findings

• Loss to follow-up, reported in ten trials, was around 7% in both treatment and control groups
• Overall, the rate of withdrawal was similar in treatment and control groups
• Overall, participants receiving calcium supplements had higher serum calcium levels than those in the control groups (mean difference of 0.04mmol per L; 95%CI: 0.02 to 0.6)
• Funnel plots showed little evidence of publication bias.

• This meta-analysis demonstrated a statistically significant reduction in systolic blood pressure and no reduction in diastolic blood pressure with calcium supplementation as compared to control
• However, due to poor quality of included trials and heterogeneity between trials, the evidence in favor of causal association between calcium supplementation and blood pressure reduction is weak and is probably due to  bias. This is because poor quality studies generally tend to over-estimate the effects of treatment.
• Larger, longer duration and better quality double-blind placebo controlled trials are needed to assess the effect of calcium supplementation on blood pressure and cardiovascular outcomes.

Government:

National Institute for Clinical Excellence, UK