Obesity, Cardiometabolic Disease, and Postoperative Acute Kidney Injury: A Retrospective Cohort Study

The burden of obesity and cardiometabolic conditions increases in prevalence with age. Thus, more complex medical care is needed to manage these conditions and their possible resulting effects, including surgical procedures and management of potential adverse postoperative outcomes. This dissertation examines the association between obesity, cardiometabolic disease, and postoperative acute kidney injury (AKI). To address the aims, we utilize data from a) a multicenter cohort of patients with total knee and hip replacements, and b) a single-center cohort of surgical patients enhanced with genetic data. In Chapter 2 (Aim 1), we evaluated latent classes of cardiometabolic conditions among patients in a multicenter cohort of total joint arthroplasty cases and found three robust groups: (1) a class with moderate probability of hypertension and low probability of other factors representing 45.2% (n = 37,032) of the population; (2) a class with high probability of hypertension and high cholesterol and moderate probability of diabetes representing 45.1% (n=36,889) of the population; and (3) a class with high probability of hypertension, high cholesterol, cardiac arrhythmias, coronary artery disease (CAD), and congestive heart failure (CHF), and moderate probability of diabetes and peripheral vascular disease representing 9.7% (n=7,950) of the population. Obesity and cardiovascular disease were associated with increased risk of AKI within 7 days following surgery. Compared to those in class 1 without obesity, those in class 3 with obesity had 3.6 times the odds of AKI (95%CI:3.1,4.3) while those in class 3 without obesity had 2.5 times the odds of AKI (95%CI:2.0,3.0) compared to the same group. Individuals with more significant cardiometabolic disease had higher odds of AKI regardless of the presence of obesity, though obesity conferred additional risk. In Chapter 3 (Aim 2), we examined whether the associations between cardiometabolic class, obesity, and AKI were mediated by modifiable intraoperative factors of general anesthesia use and minutes of intraoperative hypotension. There was little evidence of a clinically significant mediating effect of general anesthesia use or minutes of hypotension. Thus, while these factors are independent risk factors for acute kidney injury, controlling these factors intraoperatively appears to not meaningfully reduce the risk of AKI conferred by cardiometabolic comorbidity. In Chapter 4 (Aim 3), we utilized validated and published polygenic risk scores (PRS) for type 2 diabetes, coronary artery disease, and BMI in a single-center cohort of surgical patients to assess the association between these PRS and postoperative AKI. There was little evidence of an association between PRS and AKI both overall and after adjustment. Notably, however, we found that the PRS performed similarly to the disease phenotype when assessing overall model discrimination, indicating the potential utility of these scores in datasets where robust electronic health record data for these complex conditions are not available, or where the phenotype may not yet have occurred. This dissertation evaluates the effect of phenotypic and polygenic risk of cardiometabolic conditions and obesity on the development of postoperative acute kidney injury. The results suggest that the differential risk conferred by patterning of cardiometabolic disease requires additional clinician consideration when determining the best course of postoperative care, specifically around monitoring for AKI. More research is needed to determine what mechanistic factors, such as systemic inflammation, may be driving the observed associations to more fully inform who is most at risk for AKI following surgeries.PHDEpidemiological ScienceUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/174363/1/aledat_1.pd