To the Editor.
—There has been considerable concern over the possible adverse effects of changes in serum lipid concentrations induced by certain antihypertensive drugs. The concern is based on two unproved assumptions, as follows: that lipid concentrations resulting from the effects of drugs have the same prognostic significance as those occurring naturally, and that the effects of drugs observed in short-term studies can be extrapolated to a lifetime of antihypertensive therapy.The recent review of this subject by Dr Weinberger in the Archives asserted that short- or long-term therapy with thiazide diuretics increases total cholesterol and triglyceride levels.1 Of the three investigations cited in support of this statement, one was a pilot study in which six patients were treated with bendroflumethiazide for only four weeks.2 The average follow-up period of the 39 patients in the second study was only six months.3 The study was unblinded, and the