The Obesity Paradox in Stroke: Impact on Mortality and Short-term Readmission Raquel Barba, MD, PhD,* Javier Marco, MD, PhD,† Justo Ruiz, MD,‡ s Canora, MD, PhD,‡ Juan Hinojosa, MD, PhD,‡ Susana Plaza, MD, PhD,x Jesu and Antonio Zapatero-Gaviria, MD, PhD‡
Background: The aim of the present study was to assess the association of obesity with the mortality of hospitalized patients with acute stroke and the risk of readmission in less than 30 days. Methods: A retrospective chart review of a cohort of consecutive patients admitted with stroke as the primary reason for discharge in Spain between January 1, 2005, and December 31, 2011, was performed. Patients with a diagnosis of obesity were identified. The mortality and readmittance indexes of obese patients were compared against the subpopulation without theses diagnosis. Results: A total of 201,272 stroke admittances were analyzed, and 14,047 (7.0%) diagnosis of obesity were identified. In-hospital global mortality reached 14.9%, and readmittance risk was 5.9%. Obese patients showed a lower in-hospital mortality risk (odds ratio [OR], .71; 95% confidence interval [CI], .67-.76) and early readmittance risk (OR, .89; 95% CI, .82-.96) than the nonobese even after adjusting for possible confounding factors. Conclusions: Obesity in those hospitalized for stroke is associated with reduced in-hospital mortality risk and early readmittance. Key Words: Stroke—obesity—mortality—readmission. Ó 2015 by National Stroke Association
Introduction Obesity is related to increased risk for cardiovascular disease being recognized as an important risk factor for primary stroke in the general population.1,2 This association has been documented in both genders and different ethnic populations3; based on these findings the American Heart Association and the American Stroke
From the *Servicio de Medicina Interna, Hospital Rey Juan Carlos, M ostoles; †Servicio de Medicina Interna, Hospital Clınico Universitario San Carlos, Madrid; ‡Servicio de Medicina Interna, Hospital Universitario Fuenlabrada, Madrid; and xServicio de Medicina Interna, Hospital Severo Ochoa, Leganes, Spain. Received July 22, 2014; revision received October 22, 2014; accepted November 5, 2014. Address correspondence to Antonio Zapatero-Gaviria, MD, PhD, Servicio de Medicina Interna, Hospital Universitario de Fuenlabrada, Madrid, Spain. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.11.002
766
Association recommend the treatment of obesity for both primary and secondary stroke prevention.4,5 However, in patients with established cardiovascular diseases, obese patients tend to have a more favorable prognosis, which is called the obesity paradox.6-8 The relation between obesity and survival in patients who have experienced an acute stroke has been sparsely investigated; some studies9 report that mortality in obese patients is lower than in patient with normal weight; thus, it is unclear if recommendations derived from primary prevention should be extended to secondary prevention of cardiovascular disease. Recent studies now report that the stroke obesity paradox may even be extended to include risk of stroke recurrence; compared with normal-weight patients, the risk of readmission for recurrent stroke was also lower in obese stroke patients.10,11 The present study aims to investigate the association between obesity and mortality in patients hospitalized with acute stroke and the risk of readmission for recurrence of stroke in the next 30 days after discharge.
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 4 (April), 2015: pp 766-770
IMPACT OF OBESITY IN STROKE EVOLUTION
Material and Methods We identified every patient discharged from an Internal Medicine Department from hospitals of the Spanish Public Health Service between January 1, 2005, and December 31, 2011. Hospitals discharge data were obtained from the Basic Minimum Data Set. Basic Minimum Data Set contains sociodemographic and clinical data for each documented hospital admission including the following: gender and age, primary and secondary diagnoses (according to the International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code), primary and secondary procedures, admission and discharge status, length of stay, and hospital characteristics (group 1, ,150 beds; group 2, 150-200 beds; group 3, 200500 beds; group 4, 500-1000 beds; group 5, .1000 beds). For every patient, a diagnosis-related group was identified. Diagnosis-related groups are a method of classifying patient hospitalizations by diagnosis and procedures on the assumption that similar costs are expended on patients by using similar resources. The Basic Minimum Data Set registry is compulsory for every patient admitted to a hospital of the Spanish National Health Service, a system that cares for more than 90% of the country’s population. Cases were selected if they were discharged with the diagnosis of cerebrovascular disease. The diagnosis of cerebrovascular disease was identified using ICD-9-CM codes 430.00-438.99 in the diagnosis field (except 432.10). The standardized definition of the variable readmission in the Spanish Basic Minimum Data Set has been defined as a new hospitalization in the following month with the same Major Diagnostic Category in the main diagnosis. Patients who had a secondary diagnosis of obesity (ICD-9-CM, 278.00-278.02) were analyzed. The diagnosis of obesity is introduced by the internist responsible for the patient with stroke during admission, and is the same, who makes the hospital discharge report. The age adjusted Charlson Comorbidity Index (CCI) was computed for each patient. This index reflects the number and importance of comorbid diseases, relies on ICD-9-CM categories, and was used to adjust for severity of illness.12,13 The following risk factors were identified using ICD-9CM codes in any primary or secondary diagnosis field: anemia: ICD-9-CM, 280.00-285.99; tobacco: ICD-9-CM, 305.10; atrial fibrillation: ICD-9-CM, 427.3-427.32; hypercholesterolemia: ICD-9-CM, 272.0, 272.2, 272.4; diabetes: ICD-9-CM, 250.00-250.99; hypertension: ICD-9-CM, 401.0, 401.1, 401.9; acute myocardial infarct: ICD-9-CM, 410.xx, 412; and heart failure: ICD-9-CM, 398.91, 404*, 402.11, 402.91,428-428.9.
Data Analysis A descriptive analysis of these patients was carried out, and the demographic variables among the patients diag-
767
nosed with or without obesity were compared. We used the chi-square test for categorical variables with the Yates correction, the Fisher exact test for dichotomous variables when the expected value of a cell was less than 5, and the Student t test or analysis of variance for quantitative variables. The odds ratios (ORs) and 95% confidence intervals (CIs) were estimated from the regression coefficients. As this is an administrative database, the control of the confounding variables is basic. For this reason, a multivariate logistic regression analysis was carried out with the aim of determining the excess of mortality attributable to obesity, after the correction of possible confounding variables such as the age of the patient (in years, as a continuous variable), Charlson index (in points, as a continuous variable), sex, and all variables that had demonstrated a statistically significant relation in the univariate analysis with mortality and are not included in the Charlson index. A logistic regression analysis with backward stepwise procedure and P more than .10 as the criterion for exclusion were used to find the best predictive models. All statistical analyses were carried out with the use of an SPSS Software, version 16 (SPSS Inc, Chicago, IL).
Results We identified 201,272 discharges with acute stroke during the study period. Median age of patients was 77.09 years (standard deviation, 11.64); 49.2% of the patients were men. Median stay was 10.283 days (standard deviation, 14.55). A CCI of 2 or more was present in 11.9% of the cases. The subtypes of stroke were 91.6% ischemic stroke, 7.8% intracerebral hemorrhage, and .6% subarachnoid hemorrhage. A total of 14,047 (7.0%) subjects were obese. The main characteristics in obese and nonobese patients of our series are listed in Table 1. Compared with nonobese, obese patients were more frequently women (60.1% versus 50.2%; P , .001), younger (72.2 versus 77.4; P , .001), and more frequently smokers (12.2% versus 9.0%; P , .001). Comorbid conditions were common and are listed in Table 1. Multivariable logistic regression analysis was performed. The main result was that obese patients had a 29% lower risk of mortality than nonobese patients, after adjusting for potential confounding factors (OR, .71; 95% CI, .67-.76; Table 2). In addition, the risk of readmission was linked to increasing age, CCI, and gender, and obese patients were significantly less likely to be readmitted for stroke (OR, .89; 95% CI, .82-.96; Table 2).
Discussion The present study shows that obese stroke patients had a significantly lower mortality during hospitalization and
R. BARBA ET AL.
768
Table 1. Risk factors and baseline characteristics of 201,272 acute stroke patients according to obesity or nonobesity Characteristics
Nonobese (187,225)
P value
72.2 (11.6) 8445 (60.1) 9310 (66.3) 6780 (48.3) 1710 (12.2) 5187 (36.9) 602 (4.3) 960 (6.8) 3018 (21.5) 9.9 (12.8) 1277 (9.1) 721 (5.9) 1707 (12.2)
77.4 (11.2) 93,955 (50.2) 103,295 (55.2) 57,720 (30.8) 17,083 (9.1) 42,423 (22.7) 8655 (4.6) 10,562 (5.6) 48,409 (25.9) 10.3 (14.4) 28,656 (15.3) 10,752 (6.6) 22,060 (11.8)
,.001 ,.001 ,.001 ,.001 ,.001 ,.001 .01 ,.001 ,.001 .001 ,.001 ,.001 .193
12,863 (91.6) 1095 (7.8) 89 (.6)
165,875 (88.8) 19,068 (10.2) 2282 (1.2)
,.001
Obese (14,047)
Age, mean (SD), yr Gender (female) Hypertension Diabetes Smoking Hypercholesterolemia Acute myocardial infarction Heart failure Atrial fibrillation Length of stay, mean (SD), days Mortality Readmission (,30 days) Charlson .2 Subtypes of stroke Ischemic Intracerebral hemorrhage Subarachnoid hemorrhage Abbreviation: SD, standard deviation. Data are n (%) unless otherwise mentioned.
a lower risk of readmission for recurrent stroke in the next 30 days after discharge. The risk of mortality is lower in obese patients by 29% and of readmission at 30 days by 11%, respectively. Our study includes a great number of patients (201,272): all subjects admitted to internal medicine departments in Spain during 7 years with a diagnosis of stroke. Among this population we identified 14,047 (7%) with a diagnosis of obesity. Our findings are similar to other studies. Olsen et al14 in 2008 with a registry of 21,884 patients in Denmark (Danish Stroke Register) found that compared with normal-weight patients, mortality was lower in overTable 2. Multivariate analysis evaluating the association between obesity and clinical outcomes Characteristics Inpatient mortality Age (10 yr) Gender (female) Atrial fibrillation Charlson .2* Obesity Readmission Age (10 yr) Gender (female) Atrial fibrillation Charlson .2 Obesity
OR
95% CI
P value
1.69 .99 1.40 1.45 .71
1.67-1.72 .96-1.01 1.36-1.44 1.40-1.51 .67-.76
,.001 .461 ,.001 ,.001 ,.001
1.01 1.06 1.50 2.34 .89
.99-1.03 1.01-1.1 1.44-1.56 2.23-2.45 .82-.96
.092 .04 ,.001 ,.001 .004
Abbreviations: CI, confidence interval; OR, odds ratio. *Less or equal to 2 points compared with more than 2 points.
weight, obese, and severely obese stroke patients. Towfighi and Ovbiagele15 in 2009 assessed the independent association between body mass index (BMI) and mortality among stroke survivors, with overweight and obesity having a protective effect with increasing age. Vemmos et al9 over a period of 16 years prospectively investigated a total of 2785 patients after acute stroke. Based on BMI, obese and overweight stroke patients had significantly better early and long-term survival rates compared with those with normal BMI. However, other studies do not reach the same conclusions. A Korean study with a total of 1592 consecutive patients showed that an independent association between BMI and long-term mortality after ischemic stroke was found only in underweight patients.16 Kim et al17 investigated the impact of obesity on the initial neurologic severity, considered the most important prognostic factor in stroke, and concluded that patients with higher BMI levels were more likely to have milder strokes at admission, and they suggest that the initial neurologic index might be a more powerful factor than BMI levels. The same researches of the Danish Stroke Register recently studied only the deaths caused by the index stroke in relation to BMI and found no evidence of an obesity paradox in patients with stroke, and it appears to be the result of selection bias because of lack of control for the severity of the diseases leading to death after a stroke.18 Concerning stroke recurrence, in an analysis of more than 20,000 patients of The Prevention Regimen for Effectively Avoiding Second Stroke with recent ischemic stroke, the presence of baseline obesity was not an independent predictor of recurrent stroke over a 2.5-year follow-up period.19
IMPACT OF OBESITY IN STROKE EVOLUTION
The reason for the ‘‘obesity paradox’’ in stroke is unclear. Our obese patients show a higher prevalence in the association of hypertension, diabetes, and hypercholesterolemia and smoking habit than the normal-weight population. These circumstances could make them visit their physician earlier and be appropriately medicated for these diseases.19 Observational studies have shown that previous use of angiotensin-converting enzyme inhibitors is associated with better stroke outcomes.20,21 Another possible explanation is that compared with normal-weight patients, overweight/obese patients have a better metabolic reservoir, an aspect that allows them to deal better with the system catabolic imbalance and impaired metabolic efficiency induced by the stroke.19 An aspect we would like to emphasize is that most studies assessing obesity and mortality have relied on BMI to define obesity. However, the relationship between BMI and cardiovascular mortality follows a U-shaped or J-shaped curve, with the lowest mortality in overweight (BMI, 25.0-29.9 kg/m2) and mildly obese (BMI, 30.0-34.9 kg/m2) individuals.22,23 The question is how accurate is the BMI as an anthropometric measure to diagnose obesity. The diagnostic performance of BMI and the correlation with body fat percentage, with cutoff values of more than 25% in men and more than 35% in women was assessed, and it was shown that a BMI measurement of 30.0 kg/m2 or more had a high specificity (97%) but a poor sensitivity (42%) for the detection of excess body fat.24 So it seems that to establish a clear-cut relationship between obesity and cardiovascular-related mortality, other measures of obesity, especially those accounting for central obesity, such as waist circumference, waist-to-hip ratio, and waist-to-height ratio, which might be more accurate for cardiovascular risk stratification,23 should be used. Some potential limitations of the present study deserve comment, particularly using administrative data. There has only been limited evaluation of how frequently obesity is actually captured in administrative databases or how accurately is captured. Avery recent study by Martin et al25 to assess the validity of obesity coding in administrative database confirmed that obesity was poorly coded; however, when coded, it was coded accurately. Perhaps administrative databases are not an optimal data source for obesity prevalence but could be used to define obese cohorts for the follow-up. Another limitation is that we do not have information about other services that care for patients with stroke, as neurology and geriatrics. An important limitation is that this data set analysis could not be controlled or adjusted for stroke severity, and stroke severity is an important predictor of outcomes. The strength of our study is the sample size, including more than 200,000 stroke patients, all subjects admitted to internal medicine departments in Spain during 7 years with a diagnosis of stroke with all its statistical power.
769
In conclusion, on the basis of the present study of patients with stroke, using a unique database from an integrated health-care system, we conclude that obesity appears to be associated with an important reduction in mortality during admissions for stroke and a reduction in the risk of readmission for the same reason. The obesity paradox is a complex phenomenon that requires additional investigation and future studies should consider weight change when evaluating the longitudinal association among health, overweight/obesity, and outcomes. It is possible that other measures of obesity besides BMI should be needed.
References 1. Bazzano LA, Gu D, Whelton MR, et al. Body mass index and risk of stroke among Chinese men and women. Ann Neurol 2010;67:11-20. 2. Strazzullo P, D’Elia L, Cairella G, et al. Excess body weight and incidence of stroke: meta-analysis of prospective studies with 2 million participants. Stroke 2010;41: e418-e426. 3. Katsiki N, Ntaios G, Vemmos K. Stroke, obesity and gender: a review of the literature. Maturitas 2011;69: 239-243. 4. Furie KL, Kasner SE, Adams RJ, et al. Guidelines for the prevention of stroke in patients with stroke or transient ischemic attack: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011;42:227-276. 5. Goldstein LB, Bushnell CD, Adams RJ, et al. Guidelines for the primary prevention of stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011;42: 517-584. 6. Golledge J, Cronin O, Iyer V, et al. Body mass index is inversely associated with mortality in patients with peripheral vascular disease. Atherosclerosis 2013;229: 549-555. 7. Lavie CJ, Milani RV, Ventura HO. Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss. J Am Coll Cardiol 2009;53:1925-1932. 8. Zapatero A, Barba R, Gonzalez N, et al. Influence of obesity and malnutrition on acute heart failure. Rev Esp Cardiol (Engl Ed) 2012;65:421-426. 9. Vemmos K, Ntaios G, Spengos K, et al. Association between obesity and mortality after acute first-ever stroke: the obesity-stroke paradox. Stroke 2011;42: 30-36. 10. Andersen KK, Olsen TS. The obesity paradox in stroke: lower mortality and lower risk of readmission for recurrent stroke in obese stroke patients. Int J Stroke 2013, Mar 12; http://dx.doi.org/10.1111/ijs.12016. 11. Andersen KK, Olsen TS. Body mass index and stroke: overweight and obesity less often associated with stroke recurrence. J Stroke Cerebrovasc Dis 2013;22: e576-e581. 12. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373-383. 13. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613-619.
770 14. Olsen TS, Dehlendorff C, Petersen HG, et al. Body mass index and poststroke mortality. Neuroepidemiology 2008;30:93-100. 15. Towfighi A, Ovbiagele B. The impact of body mass index on mortality after stroke. Stroke 2009;40: 2704-2708. 16. Ryu WS, Lee SH, Kim CK, et al. Body mass index, initial neurological severity and long-term mortality in ischemic stroke. Cerebrovasc Dis 2011;32:170-176. 17. Kim Y, Kim CK, Jung S, et al. Obesity-stroke paradox and initial neurological severity. J Neurol Neurosurg Psychiatry 2014;1-5. http://dx.doi.org/10.1136/jnnp-2014308664. 18. Dehlendorff C, Andersen KK, Olsen TS. Body mass index and death by stroke: no obesity paradox. JAMA Neurol 2014;71:978-984. 19. Ovbiagele B, Bath PM, Cotton D, et al. Obesity and recurrent vascular risk after a recent ischemic stroke. Stroke 2011;42:3397-3402.
R. BARBA ET AL. 20. Chitravas N, Dewey HM, Nicol MB, et al. Is prestroke use of angiotensin-converting enzyme inhibitors associated with better outcome? Neurology 2007;68:1687-1693. 21. Ovbiagele B, Saver JL. The smoking-thrombolysis paradox and acute ischemic stroke. Neurology 2005; 65:293-295. 22. Adams KF, Schatzkin A, Harris TB, et al. Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. N Engl J Med 2006;355:763-778. 23. Cepeda-Valery B, Pressman GS, Figueredo VM, et al. Impact of obesity on total and cardiovascular mortality– fat or fiction? Nat Rev Cardiol 2011;8:233-237. 24. Romero-Corral A, Somers VK, Sierra-Johnson J, et al. Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes (Lond) 2008;32:959-966. 25. Martin BJ, Chen G, Graham M, et al. Coding of obesity in administrative hospital discharge abstract data: accuracy and impact for future research studies. BMC Health Serv Res 2014;14:70.
Background: The aim of the present study was to assess the association of obesity with the mortality of hospitalized patients with acute stroke and the risk of readmission in less than 30 days. Methods: A retrospective chart review of a cohort of consecutive patients admitted with stroke as the primary reason for discharge in Spain between January 1, 2005, and December 31, 2011, was performed. Patients with a diagnosis of obesity were identified. The mortality and readmittance indexes of obese patients were compared against the subpopulation without theses diagnosis. Results: A total of 201,272 stroke admittances were analyzed, and 14,047 (7.0%) diagnosis of obesity were identified. In-hospital global mortality reached 14.9%, and readmittance risk was 5.9%. Obese patients showed a lower in-hospital mortality risk (odds ratio [OR], .71; 95% confidence interval [CI], .67-.76) and early readmittance risk (OR, .89; 95% CI, .82-.96) than the nonobese even after adjusting for possible confounding factors. Conclusions: Obesity in those hospitalized for stroke is associated with reduced in-hospital mortality risk and early readmittance. Key Words: Stroke—obesity—mortality—readmission. Ó 2015 by National Stroke Association
Introduction Obesity is related to increased risk for cardiovascular disease being recognized as an important risk factor for primary stroke in the general population.1,2 This association has been documented in both genders and different ethnic populations3; based on these findings the American Heart Association and the American Stroke
From the *Servicio de Medicina Interna, Hospital Rey Juan Carlos, M ostoles; †Servicio de Medicina Interna, Hospital Clınico Universitario San Carlos, Madrid; ‡Servicio de Medicina Interna, Hospital Universitario Fuenlabrada, Madrid; and xServicio de Medicina Interna, Hospital Severo Ochoa, Leganes, Spain. Received July 22, 2014; revision received October 22, 2014; accepted November 5, 2014. Address correspondence to Antonio Zapatero-Gaviria, MD, PhD, Servicio de Medicina Interna, Hospital Universitario de Fuenlabrada, Madrid, Spain. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.11.002
766
Association recommend the treatment of obesity for both primary and secondary stroke prevention.4,5 However, in patients with established cardiovascular diseases, obese patients tend to have a more favorable prognosis, which is called the obesity paradox.6-8 The relation between obesity and survival in patients who have experienced an acute stroke has been sparsely investigated; some studies9 report that mortality in obese patients is lower than in patient with normal weight; thus, it is unclear if recommendations derived from primary prevention should be extended to secondary prevention of cardiovascular disease. Recent studies now report that the stroke obesity paradox may even be extended to include risk of stroke recurrence; compared with normal-weight patients, the risk of readmission for recurrent stroke was also lower in obese stroke patients.10,11 The present study aims to investigate the association between obesity and mortality in patients hospitalized with acute stroke and the risk of readmission for recurrence of stroke in the next 30 days after discharge.
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 4 (April), 2015: pp 766-770
IMPACT OF OBESITY IN STROKE EVOLUTION
Material and Methods We identified every patient discharged from an Internal Medicine Department from hospitals of the Spanish Public Health Service between January 1, 2005, and December 31, 2011. Hospitals discharge data were obtained from the Basic Minimum Data Set. Basic Minimum Data Set contains sociodemographic and clinical data for each documented hospital admission including the following: gender and age, primary and secondary diagnoses (according to the International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code), primary and secondary procedures, admission and discharge status, length of stay, and hospital characteristics (group 1, ,150 beds; group 2, 150-200 beds; group 3, 200500 beds; group 4, 500-1000 beds; group 5, .1000 beds). For every patient, a diagnosis-related group was identified. Diagnosis-related groups are a method of classifying patient hospitalizations by diagnosis and procedures on the assumption that similar costs are expended on patients by using similar resources. The Basic Minimum Data Set registry is compulsory for every patient admitted to a hospital of the Spanish National Health Service, a system that cares for more than 90% of the country’s population. Cases were selected if they were discharged with the diagnosis of cerebrovascular disease. The diagnosis of cerebrovascular disease was identified using ICD-9-CM codes 430.00-438.99 in the diagnosis field (except 432.10). The standardized definition of the variable readmission in the Spanish Basic Minimum Data Set has been defined as a new hospitalization in the following month with the same Major Diagnostic Category in the main diagnosis. Patients who had a secondary diagnosis of obesity (ICD-9-CM, 278.00-278.02) were analyzed. The diagnosis of obesity is introduced by the internist responsible for the patient with stroke during admission, and is the same, who makes the hospital discharge report. The age adjusted Charlson Comorbidity Index (CCI) was computed for each patient. This index reflects the number and importance of comorbid diseases, relies on ICD-9-CM categories, and was used to adjust for severity of illness.12,13 The following risk factors were identified using ICD-9CM codes in any primary or secondary diagnosis field: anemia: ICD-9-CM, 280.00-285.99; tobacco: ICD-9-CM, 305.10; atrial fibrillation: ICD-9-CM, 427.3-427.32; hypercholesterolemia: ICD-9-CM, 272.0, 272.2, 272.4; diabetes: ICD-9-CM, 250.00-250.99; hypertension: ICD-9-CM, 401.0, 401.1, 401.9; acute myocardial infarct: ICD-9-CM, 410.xx, 412; and heart failure: ICD-9-CM, 398.91, 404*, 402.11, 402.91,428-428.9.
Data Analysis A descriptive analysis of these patients was carried out, and the demographic variables among the patients diag-
767
nosed with or without obesity were compared. We used the chi-square test for categorical variables with the Yates correction, the Fisher exact test for dichotomous variables when the expected value of a cell was less than 5, and the Student t test or analysis of variance for quantitative variables. The odds ratios (ORs) and 95% confidence intervals (CIs) were estimated from the regression coefficients. As this is an administrative database, the control of the confounding variables is basic. For this reason, a multivariate logistic regression analysis was carried out with the aim of determining the excess of mortality attributable to obesity, after the correction of possible confounding variables such as the age of the patient (in years, as a continuous variable), Charlson index (in points, as a continuous variable), sex, and all variables that had demonstrated a statistically significant relation in the univariate analysis with mortality and are not included in the Charlson index. A logistic regression analysis with backward stepwise procedure and P more than .10 as the criterion for exclusion were used to find the best predictive models. All statistical analyses were carried out with the use of an SPSS Software, version 16 (SPSS Inc, Chicago, IL).
Results We identified 201,272 discharges with acute stroke during the study period. Median age of patients was 77.09 years (standard deviation, 11.64); 49.2% of the patients were men. Median stay was 10.283 days (standard deviation, 14.55). A CCI of 2 or more was present in 11.9% of the cases. The subtypes of stroke were 91.6% ischemic stroke, 7.8% intracerebral hemorrhage, and .6% subarachnoid hemorrhage. A total of 14,047 (7.0%) subjects were obese. The main characteristics in obese and nonobese patients of our series are listed in Table 1. Compared with nonobese, obese patients were more frequently women (60.1% versus 50.2%; P , .001), younger (72.2 versus 77.4; P , .001), and more frequently smokers (12.2% versus 9.0%; P , .001). Comorbid conditions were common and are listed in Table 1. Multivariable logistic regression analysis was performed. The main result was that obese patients had a 29% lower risk of mortality than nonobese patients, after adjusting for potential confounding factors (OR, .71; 95% CI, .67-.76; Table 2). In addition, the risk of readmission was linked to increasing age, CCI, and gender, and obese patients were significantly less likely to be readmitted for stroke (OR, .89; 95% CI, .82-.96; Table 2).
Discussion The present study shows that obese stroke patients had a significantly lower mortality during hospitalization and
R. BARBA ET AL.
768
Table 1. Risk factors and baseline characteristics of 201,272 acute stroke patients according to obesity or nonobesity Characteristics
Nonobese (187,225)
P value
72.2 (11.6) 8445 (60.1) 9310 (66.3) 6780 (48.3) 1710 (12.2) 5187 (36.9) 602 (4.3) 960 (6.8) 3018 (21.5) 9.9 (12.8) 1277 (9.1) 721 (5.9) 1707 (12.2)
77.4 (11.2) 93,955 (50.2) 103,295 (55.2) 57,720 (30.8) 17,083 (9.1) 42,423 (22.7) 8655 (4.6) 10,562 (5.6) 48,409 (25.9) 10.3 (14.4) 28,656 (15.3) 10,752 (6.6) 22,060 (11.8)
,.001 ,.001 ,.001 ,.001 ,.001 ,.001 .01 ,.001 ,.001 .001 ,.001 ,.001 .193
12,863 (91.6) 1095 (7.8) 89 (.6)
165,875 (88.8) 19,068 (10.2) 2282 (1.2)
,.001
Obese (14,047)
Age, mean (SD), yr Gender (female) Hypertension Diabetes Smoking Hypercholesterolemia Acute myocardial infarction Heart failure Atrial fibrillation Length of stay, mean (SD), days Mortality Readmission (,30 days) Charlson .2 Subtypes of stroke Ischemic Intracerebral hemorrhage Subarachnoid hemorrhage Abbreviation: SD, standard deviation. Data are n (%) unless otherwise mentioned.
a lower risk of readmission for recurrent stroke in the next 30 days after discharge. The risk of mortality is lower in obese patients by 29% and of readmission at 30 days by 11%, respectively. Our study includes a great number of patients (201,272): all subjects admitted to internal medicine departments in Spain during 7 years with a diagnosis of stroke. Among this population we identified 14,047 (7%) with a diagnosis of obesity. Our findings are similar to other studies. Olsen et al14 in 2008 with a registry of 21,884 patients in Denmark (Danish Stroke Register) found that compared with normal-weight patients, mortality was lower in overTable 2. Multivariate analysis evaluating the association between obesity and clinical outcomes Characteristics Inpatient mortality Age (10 yr) Gender (female) Atrial fibrillation Charlson .2* Obesity Readmission Age (10 yr) Gender (female) Atrial fibrillation Charlson .2 Obesity
OR
95% CI
P value
1.69 .99 1.40 1.45 .71
1.67-1.72 .96-1.01 1.36-1.44 1.40-1.51 .67-.76
,.001 .461 ,.001 ,.001 ,.001
1.01 1.06 1.50 2.34 .89
.99-1.03 1.01-1.1 1.44-1.56 2.23-2.45 .82-.96
.092 .04 ,.001 ,.001 .004
Abbreviations: CI, confidence interval; OR, odds ratio. *Less or equal to 2 points compared with more than 2 points.
weight, obese, and severely obese stroke patients. Towfighi and Ovbiagele15 in 2009 assessed the independent association between body mass index (BMI) and mortality among stroke survivors, with overweight and obesity having a protective effect with increasing age. Vemmos et al9 over a period of 16 years prospectively investigated a total of 2785 patients after acute stroke. Based on BMI, obese and overweight stroke patients had significantly better early and long-term survival rates compared with those with normal BMI. However, other studies do not reach the same conclusions. A Korean study with a total of 1592 consecutive patients showed that an independent association between BMI and long-term mortality after ischemic stroke was found only in underweight patients.16 Kim et al17 investigated the impact of obesity on the initial neurologic severity, considered the most important prognostic factor in stroke, and concluded that patients with higher BMI levels were more likely to have milder strokes at admission, and they suggest that the initial neurologic index might be a more powerful factor than BMI levels. The same researches of the Danish Stroke Register recently studied only the deaths caused by the index stroke in relation to BMI and found no evidence of an obesity paradox in patients with stroke, and it appears to be the result of selection bias because of lack of control for the severity of the diseases leading to death after a stroke.18 Concerning stroke recurrence, in an analysis of more than 20,000 patients of The Prevention Regimen for Effectively Avoiding Second Stroke with recent ischemic stroke, the presence of baseline obesity was not an independent predictor of recurrent stroke over a 2.5-year follow-up period.19
IMPACT OF OBESITY IN STROKE EVOLUTION
The reason for the ‘‘obesity paradox’’ in stroke is unclear. Our obese patients show a higher prevalence in the association of hypertension, diabetes, and hypercholesterolemia and smoking habit than the normal-weight population. These circumstances could make them visit their physician earlier and be appropriately medicated for these diseases.19 Observational studies have shown that previous use of angiotensin-converting enzyme inhibitors is associated with better stroke outcomes.20,21 Another possible explanation is that compared with normal-weight patients, overweight/obese patients have a better metabolic reservoir, an aspect that allows them to deal better with the system catabolic imbalance and impaired metabolic efficiency induced by the stroke.19 An aspect we would like to emphasize is that most studies assessing obesity and mortality have relied on BMI to define obesity. However, the relationship between BMI and cardiovascular mortality follows a U-shaped or J-shaped curve, with the lowest mortality in overweight (BMI, 25.0-29.9 kg/m2) and mildly obese (BMI, 30.0-34.9 kg/m2) individuals.22,23 The question is how accurate is the BMI as an anthropometric measure to diagnose obesity. The diagnostic performance of BMI and the correlation with body fat percentage, with cutoff values of more than 25% in men and more than 35% in women was assessed, and it was shown that a BMI measurement of 30.0 kg/m2 or more had a high specificity (97%) but a poor sensitivity (42%) for the detection of excess body fat.24 So it seems that to establish a clear-cut relationship between obesity and cardiovascular-related mortality, other measures of obesity, especially those accounting for central obesity, such as waist circumference, waist-to-hip ratio, and waist-to-height ratio, which might be more accurate for cardiovascular risk stratification,23 should be used. Some potential limitations of the present study deserve comment, particularly using administrative data. There has only been limited evaluation of how frequently obesity is actually captured in administrative databases or how accurately is captured. Avery recent study by Martin et al25 to assess the validity of obesity coding in administrative database confirmed that obesity was poorly coded; however, when coded, it was coded accurately. Perhaps administrative databases are not an optimal data source for obesity prevalence but could be used to define obese cohorts for the follow-up. Another limitation is that we do not have information about other services that care for patients with stroke, as neurology and geriatrics. An important limitation is that this data set analysis could not be controlled or adjusted for stroke severity, and stroke severity is an important predictor of outcomes. The strength of our study is the sample size, including more than 200,000 stroke patients, all subjects admitted to internal medicine departments in Spain during 7 years with a diagnosis of stroke with all its statistical power.
769
In conclusion, on the basis of the present study of patients with stroke, using a unique database from an integrated health-care system, we conclude that obesity appears to be associated with an important reduction in mortality during admissions for stroke and a reduction in the risk of readmission for the same reason. The obesity paradox is a complex phenomenon that requires additional investigation and future studies should consider weight change when evaluating the longitudinal association among health, overweight/obesity, and outcomes. It is possible that other measures of obesity besides BMI should be needed.
References 1. Bazzano LA, Gu D, Whelton MR, et al. Body mass index and risk of stroke among Chinese men and women. Ann Neurol 2010;67:11-20. 2. Strazzullo P, D’Elia L, Cairella G, et al. Excess body weight and incidence of stroke: meta-analysis of prospective studies with 2 million participants. Stroke 2010;41: e418-e426. 3. Katsiki N, Ntaios G, Vemmos K. Stroke, obesity and gender: a review of the literature. Maturitas 2011;69: 239-243. 4. Furie KL, Kasner SE, Adams RJ, et al. Guidelines for the prevention of stroke in patients with stroke or transient ischemic attack: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011;42:227-276. 5. Goldstein LB, Bushnell CD, Adams RJ, et al. Guidelines for the primary prevention of stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011;42: 517-584. 6. Golledge J, Cronin O, Iyer V, et al. Body mass index is inversely associated with mortality in patients with peripheral vascular disease. Atherosclerosis 2013;229: 549-555. 7. Lavie CJ, Milani RV, Ventura HO. Obesity and cardiovascular disease: risk factor, paradox, and impact of weight loss. J Am Coll Cardiol 2009;53:1925-1932. 8. Zapatero A, Barba R, Gonzalez N, et al. Influence of obesity and malnutrition on acute heart failure. Rev Esp Cardiol (Engl Ed) 2012;65:421-426. 9. Vemmos K, Ntaios G, Spengos K, et al. Association between obesity and mortality after acute first-ever stroke: the obesity-stroke paradox. Stroke 2011;42: 30-36. 10. Andersen KK, Olsen TS. The obesity paradox in stroke: lower mortality and lower risk of readmission for recurrent stroke in obese stroke patients. Int J Stroke 2013, Mar 12; http://dx.doi.org/10.1111/ijs.12016. 11. Andersen KK, Olsen TS. Body mass index and stroke: overweight and obesity less often associated with stroke recurrence. J Stroke Cerebrovasc Dis 2013;22: e576-e581. 12. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373-383. 13. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613-619.
770 14. Olsen TS, Dehlendorff C, Petersen HG, et al. Body mass index and poststroke mortality. Neuroepidemiology 2008;30:93-100. 15. Towfighi A, Ovbiagele B. The impact of body mass index on mortality after stroke. Stroke 2009;40: 2704-2708. 16. Ryu WS, Lee SH, Kim CK, et al. Body mass index, initial neurological severity and long-term mortality in ischemic stroke. Cerebrovasc Dis 2011;32:170-176. 17. Kim Y, Kim CK, Jung S, et al. Obesity-stroke paradox and initial neurological severity. J Neurol Neurosurg Psychiatry 2014;1-5. http://dx.doi.org/10.1136/jnnp-2014308664. 18. Dehlendorff C, Andersen KK, Olsen TS. Body mass index and death by stroke: no obesity paradox. JAMA Neurol 2014;71:978-984. 19. Ovbiagele B, Bath PM, Cotton D, et al. Obesity and recurrent vascular risk after a recent ischemic stroke. Stroke 2011;42:3397-3402.
R. BARBA ET AL. 20. Chitravas N, Dewey HM, Nicol MB, et al. Is prestroke use of angiotensin-converting enzyme inhibitors associated with better outcome? Neurology 2007;68:1687-1693. 21. Ovbiagele B, Saver JL. The smoking-thrombolysis paradox and acute ischemic stroke. Neurology 2005; 65:293-295. 22. Adams KF, Schatzkin A, Harris TB, et al. Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. N Engl J Med 2006;355:763-778. 23. Cepeda-Valery B, Pressman GS, Figueredo VM, et al. Impact of obesity on total and cardiovascular mortality– fat or fiction? Nat Rev Cardiol 2011;8:233-237. 24. Romero-Corral A, Somers VK, Sierra-Johnson J, et al. Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes (Lond) 2008;32:959-966. 25. Martin BJ, Chen G, Graham M, et al. Coding of obesity in administrative hospital discharge abstract data: accuracy and impact for future research studies. BMC Health Serv Res 2014;14:70.
