OBJECTIVEWe used the individual patient data from clinical trials, pooled in the INDANA data set, to explore whether blood pressure reduction was related to the baseline individual characteristics, and quantify the potential associations.

METHODSWe used the data from 31 140 patients with essential hypertension recruited in four randomized placebo-controlled clinical trials, MRC35-64, MRC65-74, STEP and SYST-EU. Thiazide diuretics, β-blocker, and calcium channel blocker, three of six major BP lowering drugs were analyzed. Patients were all with the same first dosage of the drug in each trial. Age, body weight, height, level of total cholesterin (TC), systolic blood pressure (SBP) and diastolic blood pressure (DBP) when initialed and at first visit of follow-up, pharmacological treatment, gender, status of smoking, history of myocardium infarction were factors taken into model. Data were managed by software SAS(®). Statistical analyses were performed with SAS(®) and R. Model was developed to evaluate the relationship between decrease of SBP and characteristics of patients.

RESULTSInitial SBP is the only modifier of treatment effect on SBP response in the 3 BP lowering drug classes (β = 0.09, 0.37 and 0.18, respectively). Age and initial DBP were factors significantly correlated with SBP fall for diuretic (β = 0.17 and 0.14), and age was one of factors significantly correlated with SBP fall for β-blocker (β = -0.17). Smokers would receive less SBP fall compare to non-smokers in β-blocker active treated group (β = -2.07). There is converse effect of age between the diuretic and β-blocker; older people seem sensitive to diuretic, while young people are sensitive to β-blocker. As to calcium channel antagonist class, body weight is another modifier (β = 0.06) (All P value are 0.000 except 0.050 for body weight in calcium channel antagonist class).

CONCLUSIONWe identified 5 significant modifiers (baseline SBP and DBP, age, smoking status and body weight) for SBP response to treatment effect, while gender, TC and history of myocardial infarction are not modifiers for SBP response to treatment effect.