Thiazidic diuretics and hypomagnesemia: a link to calcium pyrophosphate arthritis (CPPA)


Atxotegi Saenz-De-Buruaga J, Fernandez-Gonzalez M, Barahona E, Vazquez-Puente C, Perez-Herrero N, Fernando Perez-Ruiz

Basque Health Department, Cruces University Hospital, Baracaldo - Spain


Background: CPPA is one of the most common inflammatory arthritis, and especially in the elder as chondrocalcinosis prevalence increases with age. There is some controversy in the literature whether there is a positive (1) or negative (2) association with loop diuretics, and positive association with hypomagnesemia, which is itself associated to thiazidic diuretics (3).

Objective: To evaluate the association of hypomagnesemia to CPPA and the contribution of diuretic use and diuretic dosing to hypomagnesemia as a predisposing factor to CPPA.

Methods: Transversal analysis of a case and control study with prospective recruitment. Cases and controls were paired for age and sex and included at first CPPA diagnosis. Cases should show CPP crystals in synovial fluid and chondrocalcinosis in at least one joint; controls were recruited among patients with joint effusion showing no CPP crystals and ax-ray with no chondrocalcinosis or suspicion of chronic inflammatory CPPA. Analysis included serum magnesium (SMG) levels, HypoMg < 1.7 mg/dl, use of diuretics (loop vs HCTZ), and dose of diuretics.

Results: 582 subjects were recruited: 295 cases and 287 controls, mean age 73 years (IGR 67-80). SMG levels were lower in CPPA cases than in controls: mean 2.00 ± 0.28 mg/dl vs. 2.08 ± 0.23 mg/dl respectively, p<0.01). The rate of HypoMg was 38/287 in cases (13.2%) vs. 19/280 in controls (6.8%), p=0.01. SMG and HypoMg was also lower in cases than in controls not taking diuretics but most of the HypoMg results in cases were associated with drug-induced (tacrolimus or cyclosporin) or primary renal loss of Mg (one gitleman and two gitleman-like cases genetically unidentified). Diuretic use was much more frequent in cases, 129/295 (43.7%) than in controls 74/287 (25.8%). HCTZ and loop diuretic use (40% and 60% respectively) was the same in both cases and controls, and therefore a pooled analysis of cases and controls was performed to evaluate the impact of diuretic use on SMG and HypoMg. SMG was lower in subjects on thiazidic diuretics that loop diuretics: 1.85mg/dl ± 0,31 vs. 2.06 ± 0.27ds, p<0.01). 27/81 (33.3%) subjects on HCTZ showed HypoMg vs. 16/121 (13.2%) of those on loop diuretics. 21/31 subjects (67.7%) on hctz > 12.5 mg/day and 6/54 (11.1%) on 12.5mg/day showed HypoMg. Despite SMG levels were similar in patients on loop diuretics was not different from that of subjects not on diuretics, the rate of HypoMg was lower for those not on diuretics.

Conclusions: 1.- patients with CPPA show lower SMG levels tan controls, and may be mostly attributed to drug-induced mg loss. 2.-both loop and thiazidic diuretics induce HypoMg, but this is only dose-dependent with thiazidic diuretics.

References: 1.- Rho HY, et al. Rheumatology 2012;51:2070-4. Https://doi:10.1093/rheumatology/kes204 2.- Balderrama K, et al. Arthritis Care Res (Hoboken)2017;69:1400-6. Https://doi: 10.1002/acr.23160 3.- Ruml LA, et al. Am j KIDNEY DIS 1999; 34: 107-13. HTTPS://DOI.ORG/10.1016/S0272-6386(99)70115-0