OBJECTIVE To evaluate possible derangement in sodium balance in patients with the simple virilizing (SV) form of congenital adrenal hyperplasia (CAH) which might have implications for therapeutic procedures. METHODS Patients were sodium loaded throughout the protocol and studied after interruption of cortisone therapy for 4 days, after treatment with dexamethasone 1 mg/m2/d for 3 days and after additional therapy with 9alpha-fluorocortisone (9alphaF) 0.1 mg/m2 for 3 days and 9alphaF 0.2 mg/m2 for 3 days. After each phase, basal and stimulated (2 h in an upright position), aldosterone and plasma renin concentrations (PRC) were evaluated. METHODS Nine children aged 5.0 to 12.8 years with the clinical classification of SV CAH were studied. Diagnosis was established at the age of 2.9 +/- 1.9 years (mean +/- SD) and the patients were treated with oral hydrocortisone at a mean dose of 22.5 mg/m2/d, given in two or three daily doses. Seven patients were heterozygous for the Ile172Asn point mutation in exon 4, and two for the Pro30Leu mutation in exon 1 of the CYP21 gene. All of them had a more severe mutation or deletion in the second allele. PRC was determined by RIA and expressed as Goldblatt units (GU). Aldosterone was determined by RIA. Genotyping for disease-causing deletions and mutations was performed by Southern blot analysis, PCR and direct sequencing of CYP21. RESULTS PRC was significantly higher in patients off hydrocortisone replacement therapy than in age matched control subjects (basal 3.3 +/- 0.5 vs 1.2 +/- 0.2 GU 10(-4)/ml [mean +/- SEM], p<0.001; stimulated 8.6 +/- 0.5 vs 2.4 +/- 0.4 GU 10(-4)/ml; p<0.05). Upon treatment with dexamethasone, patients with CAH demonstrated a decrease in basal (2.1 +/- 0.5 GU 10(-4)/ml) but not in stimulated PRC (8.8 +/- 2.6 GU 10(-4)/ml). When dexamethasone treatment was supplemented by 9alphaF, both supine (0.9 +/- 0.1 GU 10(-4)/ml) and stimulated (1.6 +/- 0.3 GU 10(-4)/ml) PRC were suppressed into the normal range. Aldosterone concentrations were elevated after interrupting hydrocortisone treatment only under basal conditions. Dexamethasone caused a decrease below the reference level and 9alphaF resulted in further suppression of aldosterone concentration. CONCLUSIONS All patients were hemizygous for a CYP21 mutation that is usually considered not to be associated with clinically relevant salt loss. However, we demonstrated an aldosterone secretion disturbance in patients with SV CAH which cannot be corrected by glucocorticoid treatment alone. Additional mineralocorticoid therapy should be considered in order to suppress PRC and reduce the glucocorticoid dose required for adequate control.