RESULTS

Results of the age sub-group analyses

The results for sub-groups of younger (< 65 years) and older (= 65 years) patients are reported in the BMJ special theme issues on Hypertension, May 2008 (www.bmj.com). The analyses are based on data from 31 trials with 190,906 patients. A summary of the main findings is provided below with links to supplementary tables and figures referred to in the BMJ paper.

Summary of results

The meta-analysis showed no clear difference between age-groups in the effects of blood pressure lowering or any difference between the effects of the drug classes on major cardiovascular events (all p =0.24) (Figure 1) Neither was there any significant interaction identified between age and treatment when age was fitted as a continuous variable (all p > 0.09))(Figure 2) . Meta-regression analyses were also conducted to explore the association between the difference in follow-up blood pressure levels between randomized groups and the log relative risk for cardiovascular events. There was no difference in the risk reduction achieved per unit blood pressure reduction between age-groups for the primary outcome of major cardiovascular events (Figure 3).

The results for the meta-analyses of secondary outcomes of stroke, coronary heart disease, heart failure, cardiovascular death and total mortality are shown in Table 1. Amongst the 35 comparisons between age-groups made for these secondary outcomes, there were two with a p value of 0.05 or less. These are likely to have arisen by chance.

The results for the continuous age analyses and the meta-regression analyses for these secondary outcomes are shown in Figure 4 and Figure 5 respectively.

Subsidiary analyses

Post hoc analyses were also done using the three way age split <60, 60-69 and =70 years in an effort to provide greater insight into possible interactions of age with treatment effects. These post-hoc analyses and the analyses of age as a continuous variable used only the 25 studies that provided individual participant data. As for the pre-specified 2-way age analyses (<65 and =65 years), there was no clear difference between the three age groups in the effects of blood pressure-lowering or between the effects of drug classes (Figure 6). The results for secondary outcomes are given in Table 2.

For the comparison of the three age groups (<60 versus 60-69 versus =70) there was, however, a significant difference apparent (p<0.001) with a decline in the risk reduction achieved per unit blood pressure reduction for older compared to younger individuals. Analyses of the cause specific outcomes showed attenuations of effects for stroke (p=0.04), coronary heart disease (p=0.02) and total mortality (p=0.03) (Figure 7). It is important that these subsidiary analyses be interpreted in light of their post-hoc nature, the multiple comparisons made and in the case of the meta-regressions, the non-randomised nature of the evaluations.

Results of the diabetes sub-group analyses

The results for sub-groups of patients with and without diabetes using data from the BPLTTC were reported in Archives of Internal Medicine 2005; 165:1410-19 (www.archinte.ama-assn.org). The analyses were based on data from 27 of the 29 trials included in the second main cycle (33,395 individuals with and 125,314 individuals without diabetes). A summary of the main findings is provided below with links to figures showing individual trial data. A slide set of the diabetes sub-group results can also be downloaded at www.thegeorgeinstitute.org/bplttc/resources.

Summary of results

Total major cardiovascular events were reduced to a comparable extent in individuals with and without diabetes by regimens based on ACE-I, calcium antagonists, angiotensin receptor blockers and diuretics/beta-blockers. There was limited evidence that lower blood pressure goals produced larger reductions in total events in patients with diabetes compared to those without (p homog = 0.03). (Figure)

For the cause-specific outcomes of stroke, coronary heart disease and heart failure, there was no difference between the subgroups for any treatment comparison (Figure) with the exception of angiotensin receptor blockers and "other" where ARBs provided significantly greater protection against heart failure to patients with diabetes (p homog = 0.002) but lesser protection to the same group for the outcome of stroke (p homog = 0.05). (Figure)

The following figures show the individual trial results for the diabetes sub-group analyses

  1. Stroke
  2. CHD
  3. Heart Failure
  4. Major Cardiovascualr events
  5. Cardiovascular mortality
  6. Total death

Results of the second round of analyses

The second pre-specified round of analyses was completed in July 2003 and included data from 29 trials and 162,341 participants (Table). These results were published in The Lancet in November 2003 and the complete article can be found at www.thelancet.com . A summary of the main results is provided below with links to tables and figures showing the individual trial data. A similar summary of the first cycle main findings is also available. 

Summary of results

The overall mean age of trial participants was 65 years, 52% were men and most were selected on the basis of having pre-existing cardiovascular disease or more than one cardiovascular risk factor at baseline. The characteristics of trial participants, adherence to randomised treatment and loss to follow-up in each trial are shown (Table). 

In placebo-controlled trials the relative risks of total major cardiovascular events were reduced by regimens based on ACE inhibitors (Figure 1) (22%; 95% CI 17-27) or calcium antagonists (Figure 1) (18%; 5-29). Greater risk reductions were produced by regimens that targeted lower blood pressure goals (Figure 1) (15% 5-24). ARB-based regimens (Figure 2) reduced the risks of total major cardiovascular events (10%; 4-17) compared with control regimens. 

There were no significant differences in total major cardiovascular events between regimens based on ACE inhibitors, calcium antagonists, or diuretics or beta-blockers, although ACE-inhibitor-based regimens reduced blood pressure less. There was evidence (Figure 3) of some differences between active regimens in their effects on cause-specific outcomes. 

For every outcome other than heart failure, the difference between randomised groups in achieved blood pressure reduction was directly related to the observed difference in risk (Figure 4).

Detailed results

The following figures show individual trial data for each of the regimen comparisons:
Stroke 
Coronary Heart Disease 
Heart Failure 
Major Cardiovascular Events 
Cardiovascular Mortality 
Total Mortality 


Results of the first round of analyses

The first pre-specified round of analyses was completed in 2000. The results were presented at the International Society of Hypertension meeting in Chicago in August 2000 and published in The Lancet in December 2000. Data from a total of 15 completed trials and 74,696 subjects were included. Six overviews were performed, each of which determined the effects of treatment on the 6 principal outcomes.

Placebo-controlled trials of ACE inhibitors
Four trials (HOPE, PART2, QUIET and SCAT) involving 12,124 subjects, mostly with coronary heart disease, were included in the overview (Figure). Compared with placebo, ACE inhibitor therapy resulted in reductions of 20-30% in the relative risk of stroke, coronary heart disease, major cardiovascular events and cardiovascular death, and a 16% reduction in risk of death from any cause.  

 

Placebo-controlled trials of calcium antagonists

Only 2 trials (SYST-EUR and PREVENT) involving 5520 mostly elderly subjects with systolic hypertension were included in the overview (Figure). Compared with placebo, calcium antagonist therapy resulted in reductions in the risk of major cardiovascular events and cardiovascular death of 28% and a 39% reduction in relative risk of stroke.

  

 

Trials comparing more intensive and less intensive blood pressure lowering strategies

Three trials (ABCD [hypertensive subgroup], HOT and UKPDS) involving 20,408 subjects with hypertension were included in the overview (Figure). The weighted mean difference in achieved blood pressure between the more intensive and less intensive groups overall was only 3/3 mmHg. Compared with less intensive strategies, more intensive blood pressure lowering resulted in reductions of around 15-20% in the relative risks of stroke, coronary heart disease and major cardiovascular events.

  


Trials comparing ACE inhibitor-based regimens with diuretic and/or beta-blocker- based regimens
Three trials (STOP2, UKPDS and CAPPP) involving 16,161 subjects with systolic hypertension were included in the overview (Figure). There were no differences observed between the regimens for any outcome, although modest differences in the effects on cause-specific outcomes could not be excluded. The findings were similar with or without the inclusion of CAPPP.  

 

Trials comparing calcium antagonist-based regimens with diuretic and/or beta-blocker- based regimens

Five trials (INSIGHT, NICS-EH, STOP2, NORDIL and VHAS) involving 23,454 subjects were included in the overview (Figure). Compared with those assigned diuretic and/or beta-blocker therapy, those assigned calcium antagonist therapy had a significant 13% reduction in risk of stroke. However, they also had a 12% greater risk of coronary heart disease. Both these results were of borderline significance. There were no differences in any other outcome. The effects of dihydropyridine and non-dihydropyridine agents appeared to be similar.

  


 

Trials comparing ACE inhibitor-based regimens with calcium antagonist- based regimens
Only 2 trials (ABCD [hypertensive subgroup] and STOP2) involving 4871 subjects were included in the overview (Figure). Compared with calcium antagonist therapy, ACE inhibitor therapy resulted in reductions in the risk of coronary heart disease and heart failure. However, the reliability of this result is uncertain as ABCD was stopped early due to an apparently extreme excess of coronary heart disease events in the calcium antagonist-treated group and this is may have introduced a bias.  

 








 

 

Background   |   Methods   |   Collaboration   |   Results   |   Resources  

Top | Home