Indeed. Calling HCTZ a "diuretic" is a misnomer, since the diuretic activity is minimal at these doses, and eventually disappears completely. And seems to have little to do with why it lowers blood pressure.
As to potassium problems, 12.5 mg is just about the lower limit of where they appear, but it's possible. Unfortunately, it's also the lower edge of antihypertensive effect, so that's a hard battle to win with monotherapy. And one reason why combinations of HCTZ at this dose, combined with some other potassium-sparing agent like an ACE-inhibitor, are popular and well-studied. The effects of these are synergistic on blood pressure, and antagonistic on potassium. I'm fond of captopril and HCTZ, an old combo, but still a good one. The newer ACE inhibitors have improved on half-life but not on mortality. And the even newer ATB drugs don't have the effect on mortality we expected, either.
Beta blockers have recently come under criticism as first line drugs after the giantic LANCET meta analysis
but a beta blocker and diuretic was where JNC VI recommended everybody start. That leaves us with the thiazides, but the caveats above apply. Thus, I'm for captopril/HCTZ first. Norvasc (amlodipine) in place of captopril if you cough.
Blood Press. 1994 Jul;3(4):231-5.
Relation between low dose of hydrochlorothiazide, antihypertensive effect and adverse effects.
Jounela AJ, Lilja M, Lumme J, Morlin C, Hoyem A, Wessel-Aas T, Borrild NJ.
Department of Hypertension Clinic, Deaconess Hospital, Oulu, Finland.
Thiazide diuretics are widely used in the drug treatment of hypertension buttheir dose-response curves for the antihypertensive and adverse metabolic effects differ. To characterize the lower end of the dose-response curve a double-blind, parallel group trial was performed as multicentre study in Scandinavia. One hundred and eleven patients with newly diagnosed or previously treated mild to moderate hypertension (untreated diastolic blood pressure of 95-115 mmHg after 4 weeks placebo) were randomly allocated to various doses of hydrochlorothiazide (3, 6, 12.5 or 25 mg) or placebo for 6 weeks. Blood pressure and biochemical variables (plasma renin activity, serum potassium, magnesium, urate, fasting glucose, total cholesterol, HDL-cholesterol, triglycerides and apolipoproteins A1 and B were measured. 12.5 mg hydrochlorothiazide had a borderline effect on blood pressure whilst 25 mg had a definite antihypertensive effect. Biochemical changes were seen in plasma renin activity, serum potassium and urate after the 12.5 and 25 mg dose. Three and 6 mg had no effect on blood pressure or metabolic parameters.
Randomized Controlled Trial