Association between serum magnesium status and bronchial asthma severity: a gender-matched retrospective cohort study

Hypomagnesaemia has been implicated as a potentially modifiable correlate of poorer asthma control, but evidence from Indian outpatient populations remains limited. We evaluated the association between serum magnesium status (hypomagnesaemia versus normal or raised) and bronchial asthma severity, classified by National Asthma Education and Prevention Program (NAEPP) criteria, in a gender-matched retrospective cohort, adjusting for age and gender. We hypothesized a priori that hypomagnesaemia would be associated with a higher NAEPP severity grade. A retrospective cohort study was conducted at a tertiary care teaching hospital in Surat, Gujarat, India. Outpatient medical records were reviewed over a six-month period; 96 patients were included following 1:2 gender-based matching (32 hypomagnesaemic, 64 controls). Asthma severity was graded as mild (Grade 1), moderate (Grade 2), or severe (Grade 3) per NAEPP criteria. Association was assessed using chi-square analysis, unadjusted ordinal logistic regression, and a multivariable ordinal logistic regression adjusting for age and gender. Because the proportional-odds assumption could not be formally tested, two sensitivity analyses of severe versus non-severe asthma were performed. The hypomagnesaemia group had markedly higher asthma severity: 78.1% had severe asthma versus 7.8% in controls, and no hypomagnesaemic patient had mild asthma. Chi-square analysis confirmed a highly significant association (chi-square = 49.4, df = 2, p < 0.001). In the primary adjusted ordinal model, hypomagnesaemic patients had substantially greater odds of being in a higher severity category (OR = 50.53; 95% CI 15.13 to 210.1; p < 0.001). Sensitivity analyses of severe versus non-severe asthma gave concordant but more conservative estimates (exact conditional OR = 39.3, 95% CI 10.75 to 178.6; Firth penalized OR = 36.8, 95% CI 11.0 to 123.0). Neither age (p = 0.096) nor gender (p = 0.935) was independently associated with severity. Hypomagnesaemia was strongly associated with greater asthma severity in this cohort, and the study hypothesis was supported in this sample. Because exposure and outcome were ascertained contemporaneously from retrospective records, these data cannot establish whether hypomagnesaemia contributes to severe asthma, arises as a consequence of severe disease or its treatment, or reflects a shared cause. The wide confidence intervals reflect sparse data in the extreme severity categories. These findings are hypothesis-generating and do not support a change in current practice; prospective studies with serial magnesium measurement and adjustment for treatment, renal, and nutritional confounders are required.