Vikas Pandey, MD

Luo Y, Wang X, Matsushita K, Wang C, Zhao X, Hu B, et al. Associations Between Estimated Glomerular Filtration Rate and Stroke Outcomes in Diabetic Versus Nondiabetic Patients. Stroke. 2014.

In the United States, a commonly recited phrase among neurologists in regards to stroke epidemiology is that stroke is the fourth leading cause of death in the U.S. and the leading cause of disability. In China however (1.3 billion, most populous country in the world), stroke is the leading cause of both death and disability. For this reason, the authors were motivated to identify stroke-related risk factors in their home country, in this case the link between kidney function (based on estimated glomerular filtration rate [eGFR]) and stroke risk in diabetic and non-diabetic patients. Estimated GFR has been linked previously with poor clinical outcomes in stroke patients but whether this link could be modified by accounting for particular disease (in this case, diabetes) is still unclear.

The group used a prospective cohort of 17280 patients of which 26.8% were diabetic from 2 large national databases and followed them after one year for all-cause mortality, stroke recurrence, and stroke disability. They did not differentiate between the types of stroke (ischemic, subarachnoid hemorrhage, or hemorrhagic infarct) in regards to the primary outcomes. Diabetes was defined as fasting glucose > 126 mg/dl, non-fasting glucose > 200 mg/dl, previous symptoms of hyperglycemic crisis, use of glucose lowering drugs or self-reported history of diabetes. EGFR was calculated using CKD-EPI equation with an adjusted coefficient of 1.1 for the Asian population. The eGFR was split into categories of <45, 45-59, 60-89, 90-119 and >120 ml/min/1.73m^2. The 90-119 group served as the reference group for analysis due to the idea that hyperfiltration (>120 group) may have been due to muscle loss related to malnutrition or diabetes-related kidney hyperfiltration or other confounders. All of the endpoints were found to have significantly increased Odds ratio (OR) in those with a reduced eGFR (<45) in non-diabetics however in the diabetic group with eGFR <45, the stroke disability measure reached non-significance with the other two measures of all-cause mortality and stroke recurrence being significant. On the higher end (eGFR >120), the group found that non-diabetic patients had only increased risk of all-cause mortality, but not with stroke recurrence or stroke disability. Among diabetic patients with eGFR >120, all of the endpoints were significantly increased.

Diabetes is an undeniable risk factor for both kidney dysfunction and stroke and thus the question asked by the authors of whether there is truly a link between eGFR and poor stroke outcome is a valid one. The data reaffirms that a low eGFR (<45) is an indicator of poor stroke outcome, regardless of if the patient is diabetic or not. A topic that is more obsolete and with more conflicting previous data, however, is the idea that an increased eGFR also has a potential role in predicting poor stroke outcome.  The group found that the risk gradient for increased eGFR predicting mortality, stroke recurrence and stroke disability is steeper in the diabetic group versus the non-diabetic group. The group admits that the conditions that lead to increased eGFR such as lower muscle mass, inflammation, differences in body composition and undiagnosed malignancy may confound some of the results as these lead to increased mortality however aside from these causes is the idea of kidney hyperperfusion in diabetics which may be a result of diabetic pathophysiology and may signify hyperperfusion elsewhere (i.e. intracerebral). Thus dividing ischemic and hemorrhagic stroke subtypes and looking at the same endpoint and testing whether ischemic stroke may do better with higher eGFR and hemorrhagic strokes better with low eGFR may be a future area of study for the group. The study has many limitations including lack of generalizability (one ethnicity studied), selection bias (eGFR largely based off initial values) and the confounders previously mentioned but the study does perform the function of having clinicians worldwide pay closer attention to kidney dysfunction and its role in stroke.