Role of Body Weight in Therapy Participation and Rehabilitation Outcomes Among Individuals With Traumatic Spinal Cord Injury

Transcription

1 Archives of hysical Medicine and Rehabilitation journal homepage: Archives of hysical Medicine and Rehabilitation 2013;94(4 Suppl 2):S ORIGINAL ARTICLE Role of Body Weight in Therapy articipation and Rehabilitation Outcomes Among Individuals With Traumatic Spinal Cord Injury Wenqiang Tian, MD, hd, a Ching-Hui Hsieh, hd, OT, a Gerben DeJong, hd, a Deborah Backus, T, hd, b Suzanne Groah, MD, MSH, a amela H. Ballard, MD a From a MedStar National Rehabilitation Hospital, Washington, DC; and b Shepherd Center, Atlanta, GA. Abstract Objective: To examine the association between body weight, therapy participation, and functional outcomes among people with spinal cord injury (SCI). Design: Multisite prospective observational cohort study. Setting: Six acute rehabilitation facilities. articipants: atients (NZ1017) aged 12 years admitted for their initial rehabilitation after SCI. Interventions: Not applicable. Main Outcome Measure: Motor FIM at inpatient rehabilitation discharge and 1 year postinjury. Results: and overweight/obese patients consisted of 2 different clusters of SCI patients. patients were more likely to be younger, black, less educated, single, have Medicaid as a primary payer, and more likely to have had a cervical level injury because of violence and vehicular-related events than their overweight and obese counterparts. We found few significant differences in hours of therapy during inpatient rehabilitation across weight groups. Among patients with C5-8 ASIA Impairment Scale (AIS) grades A, B, and C injuries, underweight patients received fewer hours of physical therapy per week than patients with a healthy weight (Z.028). patients with paraplegia AIS grades A, B, and C received more hours of occupational therapy during their rehabilitation stay (<.001) than other weight groups. A higher percentage of underweight patients had pressure ulcers during inpatient rehabilitation in C5-8 AIS grades A, B, and C and paraplegia AIS grades A, B, and C groups. Only in the paraplegia AIS grades A, B, and C group did we find a significant association between weight groups and discharge motor FIM score. Regression models showed that among C1-4 AIS grades A, B, and C patients, the overweight group had better 1-year follow-up motor FIM scores than other weight groups. Conclusions: atients who had an unhealthy body weight, that is, being underweight or obese, often have therapy participation and profiles different from those deemed healthy, or just overweight. For patients with paraplegia AIS grades A, B, and C, being overweight or obese was associated with diminished motor FIM outcomes at discharge from rehabilitation. The relation between body weight status, therapy participation, and outcomes are not consistent among study group participants. Archives of hysical Medicine and Rehabilitation 2013;94(4 Suppl 2):S ª 2013 by the American Congress of Rehabilitation Medicine Over the past 2 decades, there has been a dramatic increase in the prevalence of obesity in the U.S. As of 2010, 35.5% of U.S. adultsdmore than 72 milliondare obese, defined as a body mass Supported by the National Institute on Disability and Rehabilitation Research, Office of Rehabilitative Services, U.S. Department of Education (grant no. H133A060103, H133N060028, and H133N060009). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. index (BMI) of 30kg/m 2. 1,2 With the rising number of people with obesity, it is not surprising to see more obese patients with spinal cord injury (SCI) undergoing rehabilitation care. revious studies report the prevalence of obesity among people with SCI varies from 18% to 25%. 3-8 Being underweight is also a significant but overlooked issue among SCI patients on the other tail of unhealthy weight. A higher prevalence of underweight patients (BMI<18.5) was reported among people with SCI (3.6%e7.6%) than in the general population (2.4%). 3,4,6,8, /13/$36 - see front matter ª 2013 by the American Congress of Rehabilitation Medicine

2 S126 Limited research has examined the impact of body weight on rehabilitation outcomes, such as functional recovery and discharge destination, among patients with traumatic SCI. Stenson et al 5 found that mean changes in self-care and mobility FIM were significantly lower for obese patients with paraplegia than those with a healthy weight. Chen et al 3 found that the odds of rehospitalization were 1.53 times higher for obese persons than healthy weight and overweight persons. In general, these studies report a negative association between obesity and rehabilitation outcomes. Being underweight may increase a person s risks of acquiring medical complications after SCI, such as pressure ulcers. Very little evidence exists regarding the possible effect of being underweight on therapy and outcomes. Being underweight is said to be associated with pain, unplanned hospitalizations, and lower community participation. 3 revious research examining the association between body weight and rehabilitation outcomes overlooks 2 important issues: how does an unhealthy body weight affect patient participation in rehabilitation therapy? And, how can current rehabilitation practice be customized to overcome these weight-related conditions and improve patient outcomes? The purpose of this study is thus to examine how body weight (1) is associated with patient participation in rehabilitation therapy by examining both the type and intensity of therapeutic activities across disciplines among people with SCI, and (2) is associated with functional and social outcomes, controlling for therapy activities. In short, we need to better understand how body weight relates to participation in rehabilitation and how participation in therapy is associated with outcomes. Given the lack of previous relevant research, no hypotheses are made in this study. Methods This study used data from the multisite prospective cohort observational research study, SCIRehab The SCIRehab study examined the role of rehabilitation treatment to outcomes at rehabilitation discharge and at 1 year postinjury, taking into account both individual and injury characteristics. 13 Settings and participants Six inpatient rehabilitation facilities participated in the study: the Rocky Mountain Regional Spinal Injury System at Craig Hospital, Denver, CO; Carolinas Rehabilitation, Charlotte, NC; Mount Sinai Medical Center, New York, NY; MedStar National Rehabilitation Hospital, Washington, DC; Rehabilitation Institute of Chicago, Chicago, IL; and Shepherd Center, Atlanta, GA. Facility selection List of abbreviations: AIS ASIA Impairment Scale BMI body mass index CMI case-mix index CSI Comprehensive Severity Index ED emergency department OT occupational therapy T physical therapy SCI spinal cord injury SL speech language pathology SW social work TR therapeutic recreation UTI urinary tract infection was based on willingness to participate, geographic diversity, and expertise in SCI rehabilitation. Each facility submitted and obtained their individual institutional review board approval for study enrollment and data collection. Selection criteria for the study included: (1) 12 years of age, (2) admitted to the designated SCI unit of the participating facility for their initial inpatient rehabilitation after SCI, and (3) provided informed consent. atients were excluded if they spent (1) >2 weeks in another rehabilitation center prior to admission to the SCIRehab facility, or (2) >1 week of their rehabilitation stay on a non-sci rehabilitation unit in the SCIRehab facility (because staff of the non-sci units were not trained in the data collection methods and complete data could not be ensured). Study enrollment began in August 2007 and concluded in December A follow-up interview was conducted at 1 year postinjury to capture information on postdischarge treatments, health status, health care use, and outcomes. Measurement and data collection W. Tian et al Data were mainly collected from the following sources: (1) review of patients medical records by trained chart abstractors, (2) discipline-specific process of care data entered by clinicians at each therapy session using hand-held devices with built-in electronic data collection protocols, 10,14,15 (3) standardized data collection forms (ie, forms I and II) used by the National Institute on Disability and Rehabilitation Research SCI Model Systems to obtain data at discharge and follow-up, and (4) a supplemental follow-up survey administered at 1 year postinjury either inperson or by telephone interview. 10,14,15 Body weight groups Using a patient s height and weight data captured at rehabilitation admission via chart review, the patient was assigned into 1 of 4 weight categoriesdunderweight, healthy weight, overweight, and obesedwhich are defined as BMI<18.5, 18.5BMI24.9, 25.0BMI29.9, and BMI30.0, 16 respectively. It was argued that standard BMI cutoff is not sensitive to determine the obesity among SCI patients and lower BMI cutoffs could better identify obese individuals with SCI. 17 However, a consensus on the cutoffs for the population with SCI has not been reached yet. Outcome measures The FIM measures a patient s functional independence. The FIM consists of 2 subscales: (1) a 13-item motor subscale and (2) a 5-item cognitive subscale. Each item uses a 7-point scale, with higher numbers indicating more independence. 18,19 The motor FIM was the study s main outcome variable. We used the admission FIM as a risk adjuster, and the discharge and 1-year postinjury motor FIMs as outcome measures. In this article, and the others in the supplement, we used Rasch-transformed 13,20 FIM scores (range, 0e100) to mitigate problems resulting from the use of ordinal scores. Other measures of inpatient rehabilitation outcomes included inpatient rehabilitation length of stay, interrupted stay (yes or no), discharge location, and medical complications, such as pressure ulcer and urinary tract infection (UTI). One-year postinjury outcomes were measured by a series of variables including (1) place of residence at 1 year postinjury, (2) self-reported health in general (scaled from 1, much better than 1y ago, to 5, much worse than 1y ago), (3) societal participation measured by the Craig Handicap Assessment and Reporting Technique subscales, and (4) emergency department (ED) visits and rehospitalization.

3 Body weight and rehabilitation for people with SCI rocess of care/treatment data During inpatient rehabilitation, clinicians from occupational therapy (OT), physical therapy (T), therapeutic recreation (TR), speech language pathology (SL), psychology, social work (SW), and nursing (education and care management data only) used discipline-specific protocols embedded in hand-held devices to record types of activities as well as time associated with each therapeutic activity during each therapy session. 10,15,21-27 T and OT activities were further classified and analyzed as preparatory (activities typically including impairment-based treatment activities, eg, strengthening, balance, and range of motion/stretching, to prepare for functional training), functional (skills, eg, transfers, bed mobility, wheelchair mobility, gait training in T, activities of daily living, and home management in OT), and other (activities that did not fit specifically in 1 of these 2 groups). Intensity of care for each discipline, measured by total hours of therapy activities per week, was used to measure the process of care offered to patients during inpatient rehabilitation stay. ostdischarge therapy data were obtained via the 1-year follow-up survey based on patients self-report. ostdischarge therapy activities were measured as hours of T, OT, TR, other (sum of hours of SL, psychology, SW, and nursing), and all disciplines combined. atient characteristic and injury data atient characteristics included demographic and socioeconomic information as characterized by age, sex, race, education, employment status at injury, primary payer, and primary language. atients were classified into 4 impairment groups: (1) C1-4 ASIA Impairment Scale (AIS) grades A, B, and C; (2) C5-8 AIS grades A, B, and C; (3) paraplegia AIS grades A, B, and C; and (4) AIS all grade Ds, based on both the International Standards for Neurological Classification of SCI 28,29 motor level and AIS. We formed these 4 groups in an effort to divide the larger sample into subgroups containing patients with similar functional status, while keeping subgroup size sufficiently large for intergroup comparisons. The Comprehensive Severity Index (CSI), 30 a disease-specific measure of medical acuity, was used to quantify how severely ill the patient was at admission, discharge, and throughout their rehabilitation stay. The case-mix index (CMI) in the Medicare program was also used to measure patient s severity based on patient motor function, age, and comorbidities (ie, tiers). 31 Data analysis Descriptive analyses were first conducted to examine differences in patient characteristics of the overall sample. Therapy activities and outcomes across different weight groups were then examined within each of the 4 impairment groups separately. Chi-square tests for categorical variables and analysis of variance for continuous variables were used to examine differences across weight groups. In addition, post hoc analysis was done to examine the differences between each pair of weight groups. A set of multivariate regression models was conducted to examine the association of weight groups with Rasch-transformed motor FIM at discharge and at 1 year postinjury, controlling for therapy activities, for each impairment group. In addition to weight groups, other predictors included: (1) patient characteristics including age, sex, race, education, employment status at injury, payer, and primary language; (2) health status measured at admission, days from injury to rehabilitation admission, etiology of injury, admission CSI, CMI, Rasch-adjusted admission motor FIM, and cognitive FIM; (3) medical complications during inpatient rehabilitation stay: pressure ulcer and UTI; (4) therapy activities during inpatient stay, T hours per week, OT hours per week, TR hours per week, SL therapy hours per week, psychological hours per week, SW hours per week, and RN hours per week; and (5) therapy activities postdischarge (only applied in models predicting postdischarge motor FIM): T hours, OT hours, TR hours, and others. A series of dummy variables representing each participating facility was also allowed to enter all regression models to control for possible site effects that are not reflected in the study s therapy taxonomies and other variables. In each instance, we used stepwise regression (entry,.10; exit,.05) to select those variables with statistical significance. Results atient characteristics and health status by weight groups Among the 1032 patients, 15 were excluded because of missing BMI data. Of the remaining 1017 patients, 879 responded to the follow-up interviews, yielding a response rate of 85.2%. There were no significant differences between follow-up interview respondents and nonrespondents in patient characteristics and admission status. The patients were 37.7 years old on average, and 80.9% were men. Only 37.6% were married, and the majority (70.9%) were white. Two thirds of patients were employed prior to injury, and 63.9% used private insurances. The mean body weight SD was kgs, with a mean BMI of ; 190 (18.7%) patients were obese, 270 (27.4%) were overweight, and 85 (8.4%) were underweight. The percentage of obese patients in the C5-8 AIS grades A, B, and C group was significantly lower than in other impairment categories (C1-4 AIS grades A, B, and C: 18.9%; C5-8 AIS grades A, B, and C: 12.3%; paraplegia AIS grades A, B, and C: 20.7%; AIS all grade Ds: 21.6%; Z.034). Table 1 presents patient characteristics across 4 weight groups. In general, obese and overweight patients tended to be older, married, and better educated, and were less likely to be students than those who were underweight or a healthy weight. The underweight group had a higher percentage of Medicaid beneficiaries and a longer period from injury to rehabilitation admission than the other 3 weight groups. patients were more likely to have sustained a violence-related injury, and obese patients had higher percentages of medical/surgical/other etiology of injury than other groups. There were no significant differences in medical severity across weight groups. Therapy patterns S127 Figure 1 shows the hours of therapy per week across disciplines during inpatient rehabilitation stay. Among those with high tetraplegia (C1-4 AIS grades AeC group), the healthy weight group received the fewest hours of nursing education per week (Z.039). In the low tetraplegia group (C5-8 AIS grades AeC group), there was a significant difference in hours of T per week between weight groups (Z.028). When assessed further by evaluating the time spent in the different types of T activities (preparatory or impairment-based treatment activities, mobility training, and other activities), no differences were found between body weight and impairment groups. In the paraplegia AIS grades

4 S128 W. Tian et al Table 1 Characteristic Demographic characteristics and injury information and health and functional status by weight groups (nz85) Healthy Weight (nz463) (nz279) (nz190) All (NZ1017) Demographic characteristics Age at injury (y) Male Race White Black Other Married Education <High school High school graduate >High school Vocational status prior to injury Employed full- or part-time Student Retired Unemployed and others Health plan at rehabilitation admission rivate insurance/payer Medicare Medicaid Worker s compensation English as primary language rivate residence at injury Injury information, health, and functional status Days from injury to rehabilitation admission Neurologic status C1-4 AIS grades AeC (%) C5-8 AIS grades AeC (%) araplegia AIS grades AeC (%) AIS all grade Ds (%) Injury: work related % yes Etiology of injury Vehicular Fall Sports Violence Medical/surgical/other Severity/acuity Admission CSI Maximum CSI Case-mix index Functional status Admission motor FIM Admission cognitive FIM NOTE. Values are mean SD or %. A, B, and C group, the obese group spent significantly more hours per week engaged in OT than the underweight and healthy weight groups (<.001), and the difference was primarily driven by the time that the obese group spent in OT functional activities (ie, activities of daily living and mobility training). In the AIS all grade Ds group, the obese group spent more hours per week in OT functional activities; the healthy weight group spent more time participating in TR per week, whereas the underweight group spent less time in TR (<.01). Table 2 presents the rehabilitation services received after inpatient rehabilitation discharge and the combined amount of therapy hours in both inpatient and postdischarge periods. T and OT accounted for the vast majority of postdischarge therapy in all impairment groups. In general, patients in different BMI groups received a similar amount of therapy in all disciplines after they were discharged from inpatient rehabilitation. The only 2 differences were: (1) underweight patients in C5-8 AIS grades A, B, and C groups received more other therapy hours than the other 3 BMI

5 Body weight and rehabilitation for people with SCI S129 Fig 1 Total hours per week of therapy by various disciplines during inpatient rehabilitation, by weight category and impairment group. Abbreviations: ara, paraplegia; sy, psychology; RN Edu, Registered Nursing, Education. groups (<.01); and (2) underweight, overweight, and obese patients in the all AIS grade Ds group received fewer total hours of postdischarge therapy than healthy weight patients (<.05). When combining inpatient and postdischarge rehabilitation services, underweight patients in the high tetraplegia group received fewer T, OT, and total hours than the other 3 BMI groups (<.05). In the low tetraplegia group, obese patients received fewer total therapy hours than those with a healthy weight or those who were overweight. Although underweight patients had many fewer hours of total therapy than healthy weight patients (139.2 vs 275.5h), the difference is not statistically significant given the large SD of therapy hours among underweight patients. Rehabilitation outcomes Figure 2 shows the motor FIM scores on inpatient rehabilitation admission, at discharge, and at 1 year postinjury of patients in different weight groups. In the high tetraplegia group, overweight patients made more motor FIM gain from inpatient admission to 1-year follow-up interview and achieved a higher 1-year postinjury motor FIM score than healthy weight patients (<.05). No significant differences were found in either discharge or 1-year postinjury motor FIM scores across BMI groups among patients in the low tetraplegia group. Within the paraplegia AIS grades A, B, and C group, obese patients achieved lower motor FIM gains during inpatient rehabilitation and had lower motor FIM scores on inpatient discharge compared with other weight groups (<.001). In the AIS all grade D group, obese patients achieved lower motor FIM scores on inpatient rehabilitation discharge than healthy weight patients ( vs , <.01). Table 3 reports inpatient rehabilitation and 1-year follow-up outcomes in other measures. In both the low tetraplegia and paraplegia AIS grades A, B, and C groups, a higher percentage of underweight patients had pressure ulcers than those in the other 3 BMI groups. patients in the paraplegia AIS grades A, B, and C group had a higher rate of rehospitalization than other patients. patients generally had a higher percentage of ED visits, although this was not statistically significant. Association between weight and outcomes Table 4 presents the results of regression models predicting motor FIM scores at inpatient rehabilitation discharge and at 1-year follow-up interview in each impairment category. Controlling for process of care and other covariates, a significant association between weight group and motor FIM score at discharge was only seen in the paraplegia AIS grades A, B, and C group. On inpatient discharge, obese and overweight patients achieved lower motor FIM scores than those with a healthy weight by 3.1 points (<.001) and 1.2 points (<.05), respectively. Among patients with high tetraplegia, being overweight was associated with higher

6 Table 2 Variable Hours of therapy received in postdischarge follow-up period and inpatient and outpatient combined (nz21) Healthy Weight (nz110) C1-4 AIS Grades AeC C5-8 AIS Grades AeC (nz63) (nz45) (nz20) Healthy Weight (nz91) (nz47) ostdischarge therapy hours T total OT total TR total Other total Combined outpatient Inpatient therapy and postdischarge therapy hours T inpatient and outpatient OT inpatient and outpatient TR inpatient and outpatient All combined inpatient and outpatient araplegia AIS Grades AeC AIS All Grade Ds Variable (nz28) Healthy Weight (nz144) (nz78) (nz63) (nz8) Healthy Weight (nz49) (nz47) ostdischarge therapy hours T total OT total TR total Other total All combined outpatient total Inpatient therapy and postdischarge therapy hours T inpatient and outpatient total OT inpatient and outpatient total TR inpatient and outpatient total All combined inpatient and outpatient NOTE: Values are mean SD or as otherwise indicated. (nz22) (nz30) S130 W. Tian et al

7 Body weight and rehabilitation for people with SCI S131 Fig 2 Motor FIM at rehabilitation admission, discharge, and 1-year follow-up. Abbreviations: DC, discharge; ara, paraplegia. 1-year follow-up motor FIM scores. There were no significant associations between BMI group and follow-up motor functional outcomes in the other impairment categories. Intensity of T activities was a positive predictor of both shortand long-term motor functional outcomes among patients with tetraplegia, that is, the C1-4 AIS grades A, B, and C and C5-8 AIS grades A, B, and C groups. Among paraplegia AIS grades A, B, and C patients, the more intensive T the patients received, the higher the discharge motor FIM score they achieved. However, intensities of OT and speech language therapy were negatively associated with discharge motor FIM scores in paraplegia AIS grades A, B, and C and C1-4 AIS grades A, B, and C groups separately. There was no significant association found between postdischarge therapy and 1-year motor FIM scores. Discussion This study examined the association between body weight, therapy participation, and functional outcomes among SCI patients. Our findings suggested that unhealthy weight, either being obese or underweight, is related to a rehabilitation therapy patient s received and functional outcomes. In this study, 18.7% and 27.4% of the study group was obese and overweight, respectively, on rehabilitation admission. The National SCI Model System Data reported a prevalence of 22.1% for obesity and of 31.4% for being overweight. 3,5 In contrast, 35.7% and 33.1% of U.S. adults were obese and overweight in 2010, respectively. 1,32 This study also reported a higher percentage of being underweight (8.4%) than previous findings that ranged from 3.6% to 7.5%. 3,6,8,9 This prevalence is almost 3.7 times the national level. 33 These comparisons with the general population, however, are not age-adjusted, because newly injured patients tend, on average, to be younger than the general population and thus should exhibit lower body weight. Moreover, the lower percentage of being overweight and obese (compared with national data) may also be because of patients initially losing a significant amount of weight during the early weeks after injury because of poor appetite and hypercatabolism, because the body is under great stress immediately after the initial trauma. 17,34 In general, body weight, as measured by BMI group status, is associated with the process of care and outcomes of rehabilitation, and the effects vary across impairment subgroups both during inpatient rehabilitation and after discharge. Significant associations between being overweight or obese and functional outcomes were only seen in the paraplegia AIS grades A, B, and C group. It is plausible that manual wheelchairs are often the primary means of locomotion among paraplegia AIS grades A, B, and C patients, and these patients are capable of self-transfers and self-propelling. 35 Some functional activities are therefore more challenging among this group compared with the AIS all grade Ds group who are more ambulatory and tetraplegia patients who might use a power wheelchair or have a caregiver. 5 In the short-term, the effects of weight group status were generally seen in the inpatient rehabilitation stay among those in the paraplegia AIS grades A, B, and C group, but, in the long-term, the effects of body weight were more often observed after discharge among those in the high tetraplegia group and among all AIS grade Ds. Rehabilitation therapy Our findings indicate that patients who had an unhealthy body weight, that is, being underweight or obese, often have therapy

8 Table 3 Health Outcomes Health outcomes by weight group within impairment groups (nz26) Healthy (nz132) C1-4 AIS Grades AeC C5-8 AIS Grades AeC (nz77) (nz55) (nz21) Healthy (nz101) (nz55) Inpatient rehabilitation Rehabilitation length of stay Interrupted stay (%) Complications (%) UTI ressure ulcers Discharge location (%) rivate residence Hospital Nursing home Other year follow-up interview Health rate in general CHART hysical independence Mobility Occupation Social integration rivate residence Hospitalization (% yes) Rehospitalizations ED visits (% yes) araplegia AIS Grades AeC AIS All Grade Ds Health Outcomes (nz29) Healthy (nz173) (nz90) (nz76 (nz9) Healthy (nz57) (nz57) Inpatient rehabilitation Rehabilitation length of stay Interrupted stay (%) Complications UTI ressure ulcers Discharge location rivate residence Hospital Nursing home Other year follow-up interview Health rate in general (continued on next page) (nz25) (nz34) S132 W. Tian et al

9 Body weight and rehabilitation for people with SCI S133 Table 3 (continued) araplegia AIS Grades AeC AIS All Grade Ds (nz34) (nz57) Healthy (nz57) (nz9) (nz76 (nz90) Healthy (nz173) (nz29) Health Outcomes CHART hysical independence total Mobility total Occupation total Social integration rivate residence Hospitalization (%) No. of rehospitalizations ED visits (% yes) NOTE. Values are mean SD or as otherwise indicated. Abbreviation: CHART, Craig Handicap Assessment and Reporting Technique. participation and profiles that are different from those who are deemed healthy, or just overweight. atients in the obese category received more hours per week of OT. This may be because these patients may have more difficulty with activities of daily living, for example, dressing, bathing, and toileting, because of their high weight The finding that those in the overweight high tetraplegia group received more total hours of OT, as well as more total hours of all services combined, is harder to explain, but may be because of the fact that these patients had more potential for change than those who were obese, so they required more time to meet this potential. The reason that patients in the underweight group received fewer total OT hours, as well as fewer total hours of combined inpatient and postacute services across the first year, may be because these patients could perform most activities of daily living rather easily given their low weight. 36 Some of the differences between weight groups can be explained better by assessing the differences within each of the 4 AIS groups. For instance, patients in the C1-4 AIS grades A, B, and C group who were also classified as underweight, overweight, or obese received more hours of nursing services per week. This may be because of the fact that these patients needed more education related to nutrition and diet, or more nursing education related to bed mobility, hygiene, and related activities because of their weight status. 39 Furthermore, patients classified as obese in the high tetraplegia group also received more total nursing hours postacutely, but fewer total hours for all inpatient and postdischarge rehabilitation services combined during the first year compared with the healthy and overweight groups. Those in the underweight groups also received fewer T total inpatient and postacute hours as well, potentially because of health issues because of their weight status. An alternative explanation is that these patients achieved their goals faster, which may have been facilitated by their underweight status. There were no significant differences in hours of therapy in other disciplines including speech activities, psychological, SW, and nursing education services across weight groups, because these therapy activities are not likely impacted by body weight issues. Functional outcomes Only in the paraplegia AIS grades A, B, and C group did patients who were overweight or obese achieve less physical functional improvement than those in other BMI groups. A study using the National SCI Model Systems data reported similar findings, namely, that obesity was a significant predictor of motor FIM scores (selfcare and mobility) among those with paraplegia, but not among patients with tetraplegia. 5 The more limited functional capacity of patients with tetraplegia limits the potential impact of excess body weight on their functional outcomes. The inherently greater functional capacity of patients in the paraplegia AIS grades A, B, and C group, increases the potential role of body weight on functional gains and outcomes. In short, the potential impact of body weight on SCI outcomes is a function of functional capacity, that is, greater capacity, as in paraplegia versus tetraplegia, which means that body weight can have a greater potential impact on outcomes. However, patients in the C1-4 AIS grades A, B, and C group who were overweight achieved better long-term motor FIM scores than patients of other weight groupsdan interesting finding that is difficult to interpret. This may be because of the overweight patients in the high tetraplegia group consisting of a relatively higher percentage of patients with AIS grade C (motor

10 S134 W. Tian et al Table 4 Results of regression models predicting motor FIM Rasch-Adjusted Motor FIM on Discharge Explanatory Variables C1-4 AIS Grades AeC C5-8 AIS Grades AeC araplegia AIS Grades AeC AIS All Grade Ds C1-4 AIS Grades AeC 1-Y ostinjury Rasch-Adjusted Motor FIM C5-8 AIS Grades AeC araplegia AIS Grades AeC AIS All Grade Ds Weight group 1.240* 6.180* y Healthy weight (reference) rocess of care T per week y z y z 1.880* OT per week y SL per week z SW per week 4.240* Length of stay y z 0.15 y y atient characteristics Age y y z Female y z Married y Education <12y 2.360* Retired 2.750* White race z Unemployed * rimary language not English z Health status Admission motor FIM y y y y y Admission cognitive FIM 0.030* Onset days 0.030* z z z y CMI 1.700* y ressure ulcer 1.300* Facility Location y y Location y y y y Location z Location z z Location y y Location 6 (reference) Adjusted R * <.05. y <.001. z <.01. incomplete) injuries, and who have better motor preservation than those with AIS grades A or B, than other weight groups. No differences were found in long-term outcomes between weight groups in other AIS grades. Two possible explanations are (1) BMI is one of the factors that determine patients motor functional outcome, but not the leading one; and (2) BMI was measured according to patient weight and height records when they were admitted to the rehabilitation facility. atients may lose weight after trauma and gain weight later, and thus BMI is likely a dynamic parameter that varies over the course of an initial hospitalization and rehabilitation stay. 17,34,35 Study limitations BMI was calculated using weight and height measured when patients were admitted to inpatient rehabilitation, which cannot capture the changes in BMI over time. atients with traumatic SCI often experience dramatic weight changes, losing weight in initial weeks and then gaining more weight afterward because of less activity or inactivity. A 5-year trend study conducted in the Netherlands found a significant increase in BMI in the first year after SCI, and a continuous increase in a 5-year period after discharge from inpatient rehabilitation. 40 atients BMI changes over time, and a patient categorized as healthy weight may become overweight or obese after 1 year. A static admission BMI may not be an adequate indicator in establishing an association between body weight and therapy participation and rehabilitation outcomes. In this study, we used BMI cutoff points for the general, ablebodied population to define obesity, overweight, healthy weight, and underweight. Some argue, however, that people with SCI have greater body fat mass and less fat-free mass per BMI unit because of inactivity. 41,42 Many people with SCI do not appear to be obese but carry large amounts of fat tissue. Therefore, BMI values used in the able-bodied population often underestimate the body fat of people with SCI. Some investigators and health care providers

11 Body weight and rehabilitation for people with SCI suggest that >22kg/m 2 instead of >30kg/m 2 should be considered as being obese among people with SCI. The definition of BMI groups used in this study may be not sensitive enough to capture all patients who were obese. Conclusions We found no consistent relation between body weight status, therapy participation, and outcomes among study group participants. Being overweight or obese is associated with diminished functional outcomes at discharge for patients with paraplegia AIS grades A, B, and C. Higher intensity of physical therapy during inpatient rehabilitation was generally associated with better motor FIM scores both at discharge and follow-up across impairment and body weight groups. Further research is needed to examine how body weight changes over the course of an individual s postinjury life and how these changes affect rehabilitation participation and outcomes both in the short-term and long-term. Keywords Body mass index; Rehabilitation; Spinal cord injuries; Treatment outcomes Corresponding author Wenqiang Tian, MD, hd, Center for ost-acute Innovation and Research, MedStar National Rehabilitation Hospital, 102 Irving St, Washington, DC address: Wenqiang.Tian@ medstar.net. References 1. Flegal KM, Carroll MD, Kit BK, Ogden CL. revalence of obesity and trends in the distribution of body mass index among US adults, JAMA 2012;307: National Heart, Lung, and Blood Institute. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. Bethesda: National Institute for Health; Chen Y, Cao Y, Allen V, Richards JS. Weight matters: physical and psychosocial well being of persons with spinal cord injury in relation to body mass index. Arch hys Med Rehabil 2011;92: Gupta N, White KT, Sandford R. Body mass index in spinal cord injury e a retrospective study. Spinal Cord 2006;44: Stenson KW, Deutsch A, Heinemann AW, Chen D. Obesity and inpatient rehabilitation outcomes for patients with a traumatic spinal cord injury. Arch hys Med Rehabil 2011;92: Tomey KM, Chen DM, Wang X, Braunschweig CL. Dietary intake and nutritional status of urban community-dwelling men with paraplegia. Arch hys Med Rehabil 2005;86: Weaver FM, Collins EG, Kurichi J, et al. revalence of obesity and high blood pressure in veterans with spinal cord injuries and disorders: a retrospective review. Am J hys Med Rehabil 2007;86: Alschuler KN, Gibbons LE, Rosenberg DE, et al. Body mass index and waist circumference in persons aging with muscular dystrophy, multiple sclerosis, post-polio syndrome, and spinal cord injury. Disabil Health J 2012;5: Johnston MV, Diab ME, Chu BC, Kirshblum S. reventive services and health behaviors among people with spinal cord injury. J Spinal Cord Med 2005;28: Gassaway J, Whiteneck G, Dijkers M. Clinical taxonomy development and application in spinal cord injury research: the SCIRehab roject. J Spinal Cord Med 2009;32: S Whiteneck G, Gassaway J. SCIRehab: a model for rehabilitation research using comprehensive person, process and outcome data. Disabil Rehabil 2010;32: Whiteneck G, Gassaway J. SCIRehab: a test of practice-based evidence methodology. J Spinal Cord Med 2011;34: Whiteneck GG, Gassaway J. SCIRehab uses practice-based evidence methodology to associate patient and treatment characteristics with outcomes. Arch hys Med Rehabil 2013;94(4 Suppl 2):S Whiteneck G, Dijkers M, Gassaway J, Lammertse D. The SCIRehab roject: classification and quantification of spinal cord injury rehabilitation treatments. reface. J Spinal Cord Med 2009; 32: Whiteneck G, Gassaway J, Dijkers M, Jha A. New approach to study the contents and outcomes of spinal cord injury rehabilitation: the SCIRehab roject. J Spinal Cord Med 2009;32: Centers for Disease Control and revention. Defining overweight and obesity. Available at: Accessed February 5, Cox SA, Weiss SM, osuniak EA, Worthington, rioleau M, Heffley G. Energy expenditure after spinal cord injury: an evaluation of stable rehabilitating patients. J Trauma 1985;25: Fiedler R, Granger CV. Functional Independence Measure: a measurement of disability and medical rehabilitation. Tokyo: Springer-Verlag; Fiedler RC, Granger CV, Russell CF. UDS(MR)SM: follow-up data on patients discharged in Uniform Data System for Medical Rehabilitation. Am J hys Med Rehabil 2000;79: Mallinson T. Rasch analysis of repeated measures. Rasch Measurement Transactions 2011;25: Abeyta N, Freeman ES, rimack D, et al. SCIRehab roject series: the social work/case management taxonomy. J Spinal Cord Med 2009;32: Cahow C, Skolnick S, Joyce J, Jug J, Dragon C, Gassaway J. SCI- Rehab roject series: the therapeutic recreation taxonomy. J Spinal Cord Med 2009;32: Gordan W, Spivak-David D, Adornato V, et al. SCIRehab roject series: the speech language pathology taxonomy. J Spinal Cord Med 2009;32: Johnson K, Bailey J, Rundquist J, et al. SCIRehab roject series: the supplemental nursing taxonomy. J Spinal Cord Med 2009;32: Natale A, Taylor S, LaBarbera J, et al. SCIRehab roject series: the physical therapy taxonomy. J Spinal Cord Med 2009;32: Ozelie R, Sipple C, Foy T, et al. SCIRehab roject series: the occupational therapy taxonomy. J Spinal Cord Med 2009;32: Wilson C, Huston T, Koval J, Gordon SA, Schwebel A, Gassaway J. SCIRehab roject series: the psychology taxonomy. J Spinal Cord Med 2009;32: Marino R. Reference manual for the international standards for neurological classification of spinal cord injury. Chicago: American Spinal Injury Association; Marino RJ, Barros T, Biering-Sorensen F, et al. International standards for neurological classification of spinal cord injury. J Spinal Cord Med 2003;26(Suppl 1):S Horn SD, Smout RJ, DeJong G, et al. Association of various comorbidity measures with spinal cord injury rehabilitation outcomes. Arch hys Med Rehabil 2013;94(4 Suppl 2):S Center for Medicare and Medicaid Services, HHS. Medicare program; Inpatient Rehabilitation Facility rospective ayment System for Federal Fiscal Year of 2012; Changes in Size and Square Footage of Inpatient Rehabilitation Units and Inpatient sychiatric Units; Correction. Federal Register/Vol. 76. No. 186/Monday, September 26, 2011/Rules and Regulations: Ogden CL, Carroll MD, Kit BK, Flegal KM. revalence of obesity in the United States, Hyattsville: National Center for Health Statistics; Fryar CD, Ogden CL. revalence of underweight among adults aged 20 years and over: United States, Hyattsville: National Center for Health Statistics; 2012.

12 S Laven GT, Huang CT, DeVivo MJ, Stover SL, Kuhlemeier KV, Fine R. Nutritional status during the acute stage of spinal cord injury. Arch hys Med Rehabil 1989;70: Dearwater SR, Laorte RE, Cauley JA, Brenes G. Assessment of physical activity in inactive populations. Med Sci Sports Exerc 1985;17: Alley DE, Chang VW. The changing relationship of obesity and disability, JAMA 2007;298: Fontaine KR, Barofsky I. Obesity and health-related quality of life. Obes Rev 2001;2: Reynolds SL, Saito Y, Crimmins EM. The impact of obesity on active life expectancy in older American men and women. Gerontologist 2005;45: W. Tian et al 39. okorny ME, Scott E, Rose MA, et al. Challenges in caring for morbidly obese patients. Home Healthc Nurse 2009;27: de Groot S, ost MW, ostma K, Sluis TA, van der Woude LH. rospective analysis of body mass index during and up to 5 years after discharge from inpatient spinal cord injury rehabilitation. J Rehabil Med 2010;42: Jones LM, Legge M, Goulding A. Healthy body mass index values often underestimate body fat in men with spinal cord injury. Arch hys Med Rehabil 2003;84: Laughton GE, Buchholz AC, Martin Ginis KA, Goy RE. Lowering body mass index cutoffs better identifies obese persons with spinal cord injury. Spinal Cord 2009;47: