Pyrophosphate arthritis (PPA) has been associated to hypomagnesemia (hypoMg). Eventhough thiazide diuretics (TD) are considered as a risk factor for hypoMg; paradoxically, an absence of association has been described between PPA and TD, and present with loop diuretics (LD), but the authors lacked levels of serum magnesium (sMg) for statistical analysis and the diagnosis was only clinical; nonetheless, the authors advised considering the withdrawal of LD in patients with PPA.
Objectives: To evaluate sMg levels in patients with PPA and controls, as well as the impact of treatment with diuretics and administered doses on sMg. Methods: Prospective recruitment, transversal design, case-control study (patients with confirmed PPA) and controls (paired by age and gender, absence of chondrocalcinosis). sMg levels were obtained, as well as data regarding use of diuretics, type and doses. The rate of prescription of diuretics, type and doses (cut-off points ≤ 40mg qd equivalent of Furosemide and ≤ 12.5mg qd equivalent of hydrohlorothiazide) associated to sMg levels and presence of hypoMg (sMg<1.7 mg/dL) were also analyzed.
Results: 602 patients (53% male, 73±9 years), with 323 cases and 279 controls were analyzed. The prescription of any diuretic and TD was more frequent in cases than in controls: 40.2% vs 25.4% and 15.8% vs 9.7%, respectively (p<0.01). The levels of sMg and rate of hypoMg were different in both: 2.00±0.29 vs 2.08±0.23 mg/dL, p<0.01 and 7.2% vs 14.0%. Patients on diuretics had lower sMg levels than patients without them (1.99±0.32 vs 2.06 ± 0.24, p<0.01), but this effect was only observed in patients on TD (1.84±0.31 vs 2.08±0.29). The rate of hypoMg was higher in patients on diuretics (22.6% vs 5.0%; OR 5.5; 95%CI 3.1-9.8; p<0.01), and associated again with use of TD (36.0% vs 13.4%; OR 3.6; 95% CI 1.8-7.5; p<0.01).The analysis of diuretic doses showed that the rate of hypoMg for patients without diuretics was not statistically different than that of patients on LD with either low or high doses, or patients on low-dose TD (5, 20 and 13% respectively), but was different when compared to high-dose TD (68%, p<0.01).
Conclusions: sMg levels and hypoMg are associated to high-dose TD use, but not to lower doses or prescription of LD, what is concordant with pharmacodynamics and pharmacokinetics. Patients with PPA use TD more frequently than controls, showing lower sMg levels and higher rates of hypoMg when compared to controls and patients with LD.