Customize your JAMA Network experience by selecting one or more topics from the list below.
Razinia T, Saver JL, Liebeskind DS, Ali LK, Buck B, Ovbiagele B. Body Mass Index and Hospital Discharge Outcomes After Ischemic Stroke. Arch Neurol. 2007;64(3):388–391. doi:10.1001/archneur.64.3.388
Obesity is a risk factor for vascular disease and has been associated with poorer outcomes in hospitalized patients.
To evaluate the relationship between body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) and discharge outcomes among persons hospitalized for ischemic stroke.
The relationship between BMI and discharge outcomes was analyzed in 805 consecutive patients with ischemic stroke admitted to a university hospital stroke service. Patients were categorized into 4 BMI categories representing lean, overweight, and class I and class II obesity. Outcome measures analyzed included discharge functional activity, direct discharge to home, and length of hospital stay. The independent effect of BMI on outcome was evaluated with regression analysis, adjusting for other variables known to predict outcome after ischemic stroke.
Four hundred fifty-one individuals (56.0%) met study criteria. Mean age was 65 years, and 28.0% were men. In multivariate analysis, the highest BMI category (compared with lowest BMI category) was associated with the least likelihood of being discharged directly home from the hospital stroke service (26% vs 45%, P = .04), and the upper 2 BMI categories (compared with lowest BMI category) were associated with a trend toward extended length of hospital stay (6.3 vs 5.2 days, P = .08). No significant differences in the functional activity outcome were noted across BMI categories.
Elevated BMI is associated with a lower likelihood of being discharged home and a trend toward extended hospital stay among patients hospitalized for ischemic stroke. Body mass index at hospital admission had no relation to discharge functional activity after stroke.
Excess weight is a documented independent risk factor for stroke.1-3 Furthermore, obesity has been associated with a proinflammatory state,4 a prothrombotic state,5 and increased risk for in-hospital complications.6 As such, obesity may not just play a role in stroke occurrence but might independently influence stroke outcome. Delineating the impact of obesity on short-term outcome could identify a useful prognostic factor and underscore the need to include patient weight status into routine stroke assessment. In this study, we evaluated the impact of body mass index (BMI) on hospital discharge outcomes following ischemic stroke.
Data were collected prospectively from September 1, 2003, to July 2, 2006, on consecutive patients with ischemic stroke aged 18 years and older who were admitted to a university hospital stroke service. Ischemic stroke was defined as clinical signs of focal disturbance of cerebral function, lasting more than 24 hours, of presumed ischemic origin.7 Neuroimaging consisting of magnetic resonance imaging (MRI) or computed tomography (CT) of the brain was used to rule out events mimicking stroke. Presenting stroke severity was assessed with the National Institutes of Health Stroke Scale (NIHSS), in which stroke severity was graded on a 42-point scale by investigators who were certified in the application of the NIHSS.8,9 All patients were assessed at discharge using the modified Rankin Scale, which measures functional status on a 7-rank scale.10
Weight and height were measured by the unit nurse on admission; when this was not feasible, they were obtained verbally from the patient or caregiver. Body mass index was calculated as weight in kilograms divided by height in meters squared. Study subjects were divided into 4 groups based on BMI categories specified by the US Preventive Services Task Force.11 Patients with a BMI of 35 or greater represented class II obesity; those with a BMI of 30 to 34, class I obesity; those with a BMI of 25 to 29, overweight; and those with a BMI of less than 25, lean.
Prespecified outcome measures to be evaluated in relation to BMI category included the following:
Favorable discharge functional outcome, defined as modified Rankin Scale score of 0 or 1.
Favorable discharge destination, defined as patient sent home directly from hospital stroke service (compared with transfer to inpatient rehabilitation setting, general medicine service, or skilled nursing facility).
Reduced length of hospital stay, defined as median number of days from date of admission to date of discharge or transfer from stroke service.
To evaluate the role of possible confounding factors, other potential determinants of incident stroke outcome were also analyzed. These potential determinants were prespecified based on prior reports in the literature and the clinical judgment of the authors12,13 and were as follows: age at time of admission, history of hypertension, history of atrial fibrillation, history of diabetes mellitus, systolic blood pressure upon admision, stroke severity (NIHSS) upon admission, premorbid statin use, and stroke subtype per modified TOAST (Trial of ORG 10172 in Acute Stroke Treatment) classification.14 The study was approved by the local institutional review board.
Statistical analyses were carried out using the statistical software STATA version 9 (StatCorp, College Station, Tex). A bivariate assessment of the relationship between BMI category and modified Rankin Scale score and length of hospital stay was made using Spearman rank correlation coffiecient. The relationship between BMI and the binary outcome of discharge to home (yes/no) was computed by the χ2 test for trend. The ordinal variable covariates (age and systolic blood pressure) were compared across BMI categories using Wilcoxon rank sum test (Kruskal-Wallis test), and the remaining covariate (binary variable) comparisons were computed using exact χ2 method. Least absolute deviation robust regression was used for the continuous or quasicontinuous outcomes of modified Rankin Scale score and length of hospital stay. Logistic regression was used for the binary outcome of discharge to home. For all regression models, dummy-coded variables for the 3 BMI categories of overweight, class I obesity, and class II obesity were forced in the model controlling for the remaining covariates. The lowest BMI category was the referent group for purposes of comparison.
Of 805 consecutive patients with ischemic events, 451 (56.0%) met study criteria. Body mass index values could not be obtained in 213 patients with ischemic stroke because either they were not recorded by nursing staff or the patient or caregiver could not provide this information; 141 individuals had a final diagnosis of transient ischemic attack.
Mean age of study patients was 65 years old (range, 83-101 years). Forty-eight patients received recanalization treatment, and there were no differences in the rate of recanalization therapies among BMI categories (P = .39). Table 1 represents the baseline demographic and clinical characteristics of the participants according to the 4 BMI categories. Patients in the lowest BMI category were significantly older and more likely to have a history of atrial fibrillation and coronary artery disease than those in the highest BMI category. History of diabetes mellitus, hyperlipidemia, and smoking were significantly more likely in patients in the highest BMI category compared with those in the lowest.
Results of the multivariate analyses can be found in Table 2. Patients in the highest BMI category were least likely to have favorable functional activity at the time of hospital discharge compared with those in lower BMI categories, but this difference did not reach statistical significance. However, compared with the lowest BMI category, those in the highest BMI category were significantly less likely to be discharged directly home from the stroke service. In addition, patients in the highest BMI category showed a trend toward a longer duration of stay in the stroke service compared with those in the lowest BMI group. Eighteen of the 451 patients died and in-hospital mortality did not differ among the BMI groups (4.3% vs 3.1% vs 6.6% vs 0%, P = .52).
Our study suggests that higher BMI at the time of hospital admission for ischemic stroke is associated with a lower likelihood of being discharged directly home and longer stay in the hospital stroke service. Our findings are in accord with studies that have examined the impact of BMI on hospital outcomes among patients with general medical conditions,15 those with recent trauma,16,17 and those requiring intensive care.18,19 Extended length of hospital stay has been consistently shown to be associated with obesity,15,20,21 and even though one study found this difference between the highest and lower categories of BMI to be only 1 day, such a delay may have financial and psychological implications, and expose patients to the additional risk of acquiring nosocomial conditions.15
The BMI groups did not differ in presenting stroke severity; therefore, the differences in short-term outcome must be due to other factors such as a more frequent occurrence of in-hospital complications or longer recovery time.18,19 We did not collect any data that could confirm or deny these factors in our study. Alternatively, other unmeasured deleterious factors associated with obesity may have been present in the patients with higher BMI, making them susceptible to poorer outcomes. The fact that obesity did not exert a greater effect on short-term outcomes may reflect an attenuating effect of better nutritional status. A prior study did not find BMI to be an independent predictor of disability.22
This study has limitations. For some patients, BMI was derived using reports of the patient or family members. However, the reliability of self-reports of height and weight has been validated in other studies.23 Because of the small number of fatal ischemic strokes in the stroke service, associations between BMI categories and fatal outcomes could not be interrogated. As our measure of obesity, we used BMI, an index of total body fat, rather than waist circumference, an index of abdominal fat. Although both indices of obesity carry untoward vascular risk, abdominal fat is thought to play a more direct role than total body fat in the development of various metabolic abnormalities.24 However, large epidemiologic studies have shown a high correlation between BMI and waist circumference.25 Finally, missing BMI data could have influenced the results, and issues such as cuff size, day and night time variability, or pseudohypertension may have influenced our blood pressure data.
In summary, this study suggests that hospitalized obese individuals may have poorer discharge clinical outcomes than their leaner counterparts. Future studies are required to confirm these results and to identify the pathophysiologic mechanisms at play. In the interim, our study provides further impetus for identifying and appropriately treating obese individuals at risk for first and recurrent stroke.
Correspondence: Tannaz Razinia, Department of Neurology, UCLA Stroke Center, UCLA Medical Center, 710 Westwood Plaza, Los Angeles, CA 90095 (TannazRazinia@gmail.com).
Accepted for Publication: September 1, 2006.
Author Contributions:Study concept and design: Razinia and Ovbiagele. Acquisition of data: Razinia, Saver, Liebeskind, Ali, Buck, and Ovbiagele. Analysis and interpretation of data: Razinia, Saver, and Ovbiagele. Drafting of the manuscript: Razinia, Saver, and Ovbiagele. Critical revision of the manuscript for important intellectual content: Razinia, Saver, Liebeskind, Ali, Buck, and Ovbiagele. Statistical analysis: Saver. Obtained funding: Saver and Ovbiagele. Administrative, technical, and material support: Razinia, Saver, and Ovbiagele. Study supervision: Saver and Ovbiagele.
Financial Disclosure: None reported.
Funding/Support: This study was supported in part by Award P50 NS044378 from the National Institutes of Health, part of the National Institute of Neurological Disorders and Stroke.
Acknowledgment: We thank Jeffrey Gornbein, PhD, for his statistical consultation.
Create a personal account or sign in to: