Chronic kidney disease (CKD) is a common and debilitating disease of older cats that has been discussed extensively on this blog, both in the context of diagnosis and therapy. Perhaps equally relevant is the question of prognostic indications, values that may allow for veterinarians and owners to better estimate survival and quality of life factors.
Several prognostic markers have been found that are associated with survival in cats with CKD. These include (but are not limited to) serum creatinine, proteinuria, serum potassium and serum phosphorus levels. Some of these (such as creatinine) are independent predictors of survival, while others (such as phosphorus) are linked to other factors.
Fibroblast Growth Factor 23 (FGF23) is a hormone that decreases serum potassium levels and has been shown in recent years to be a strong marker (and potentially a causational factor) for complications of CKD such as metabolic bone disease, renal fibrosis, tissue mineralization, and overall mortality.
Magnesium is an electrolyte that plays an important role in many bodily processes. Among other essential roles it inhibits vascular mineralization, decreases the release of profibrotic cytokines, and in human medicine is a risk factor for kidney-associated mortality. It may also have an inverse relationship with FGF23 in human medicine.
The purpose of this study was to determine the relationship between plasma magnesium levels with FGF23, mortality, and disease progression in cats with azotemic CKD. The study was designed as a retrospective cohort study, recruiting cats from two practices in England.
A group of 120 client owned cats greater than 9 years were recruited as a control group. Cats were considered healthy if physical exam, history, bloodwork and urinalysis showed no abnormalities. The group of cats considered to have CKD were also greater than 9 years and had a serum creatinine greater than 2mg/dL in conjunction with a urine specific gravity less than 1.035. Cats diagnosed with hyperthyroidism or diabetes mellitus or receiving treatment with corticosteroids were excluded. 174 cats were recruited into the CKD group, 88 were female and 86 male.
Plasma magnesium levels were measured after enrollment in the study to allow blinding of the case selection process.
Extensive data was collected from medical records, including age, breed, sex, body weight, body condition score, muscle score, systolic blood pressure, PCV, routine plasma biochemical variables, ionized calcium concentration, venous blood gases and pH values, plasma calcidiol, calcitriol, FGF23, PTH, total T4, USG, urine culture, and urine protein‐to‐creatinine ratio. Date of death and if azotemia progressed were also recorded.
Data was analyzed using several statistical methods to allow for determination of correlations and trends. As data on muscle mass, ionized Ca, vitamin D metabolites, and venous blood gas were available for <50% of cats they were excluded from analysis. A normal reference range for plasma magnesium in healthy cat was also established.
Of the 174 cats enrolled in the CKD group, 114 cats were in IRIS stage 2, 50 cats were stage 3, and 10 cats stage 4. 21% were hypertensive, and the average age was 14.4 years.
The median plasma magnesium was 2.07mg/dL, with an established reference interval of 1.73-2.57mg/dL. 11.5% of cats were hypomagnesemic (lower than normal serum magnesium concentration) and 6% of cats were hypermagnesemic (higher than normal serum magnesium concentration). IRIS stage 4 cats were more likely to be hypermagnesemic, though this was not statistically demonstrated.
Elevated FGF23 and systemic hypertension were both independently associated with hypomagnesemia. Cats with systemic hypertension had a median magnesium of 1.97mg/dL, compared to 2.09 in the normotensive group. Treatment with amlodipine did not influence magnesium concentration.
While no direct relation was seen between plasma magnesium and FGF23, when controlling for creatinine and phosphate a significant inverse correlation was seen with strongest effects in IRIS 4. FGF23 was shown to be a strong predictor of hypomagnesemia.
In univariate analysis both hypomagnesaemia and hypermagnesaemia were associated with a shorter survival, however the link between death and hypomagnesemia was not present in multivariate analysis. Hypomagnesemia remained an independent predictor of mortality. Degree of hypomagnesemia was correlated with risk of death. Hypomagnesemia was not a predictor of disease progression in a multivariate model.
While hyperphosphatemia was a predictor of mortality, this was only true in cats who were also hypomagnesemic. This is interesting, and suggests that hyperphosphatemia may be of greater concern in cats with low magnesium levels.
This study concluded that decreased plasma magnesium is associated with an increased risk of mortality and an increase in FGF23 levels. It also suggests that hyperphosphatemia may be to some extent ameliorated by normal magnesium levels (or that it is exacerbated by hypomagnesemia). Further, it draws a correlation between magnesium level and systemic hypertension.
This is valuable information that suggests the routine monitoring of magnesium levels may be of some significance in the management of CKD. While commonly available, magnesium analysis is not performed as often as other values, such as phosphorus and potassium, in the screening of cats with CKD. An obvious follow-up question is whether this association is causational or merely correlational; is the low magnesium the cause of increase morbidity and mortality, or simply another marker. Further investigation as to whether correction of magnesium levels in these cats reduces risk of death or FGF23 may provide valuable insight into this question. (MRK)