Factors related to activity limitations in a group of Cuban Americans before and after hip fracture.

Health-related quality of life represents the total effect of individual and environmental factors on impairment, functional limitation, disability, and health status. (1) Individual factors may include biological factors, demographic factors, comorbidity, health habits, personal behaviors, lifestyles, psychological traits, and social interactions and relationships. Environmental factors may include available medical or rehabilitative care, medications and other interventions, physical and social environment, and ethnic or cultural norms. The Guide to Physical Therapist Practice (1) describes an expanded disablement model that shows interactions among individual and environmental factors and disability (Figure). This model was proposed by Guccione to describe the process of disablement in people with arthritis (2); however, little work has been done to test the relationships among different factors in this model of disability in the context of physical therapist practice.

[FIGURE OMITTED]

In order to provide the most effective interventions, I contend that physical therapists should seek a better understanding of the interactions among psychosocial, environmental, and medical factors related to acute injury and the process of disablement. Impairments related to an acute injury such as hip fracture result in activity limitations and disability; but social support and psychological status may mediate the impact of an acute illness or injury on disability. Researchers have shown that social support is related to recovery of function both 2 months (3,4) and 6 months (5) after hip fracture in a general population with hip fractures (4,5) and in community-dwelling older women with hip fractures. (3) In several studies, the data demonstrated that depressive symptoms also were related to recovery from hip fractures both 2 months after fracture (3) and 12 months after hip fracture (6-8) when controlling for other prognostically important factors such as physical function, cognitive function, age, and comorbidity. (3,6-8) Most of these studies involved large cohorts of individuals, with little representation of minority ethnic groups. (3-8)

The results of all of these studies suggest that psychosocial variables may be important mediators in the recovery process. Social support in the form of physical or emotional assistance may be related to better premorbid physical abilities and may encourage better recovery of physical abilities after hip fracture. Similarly, elderly people with symptoms of depression such as feelings of sadness and low energy levels show more functional limitations prior to hip fracture and may have more difficulty recovering functional abilities. In addition, those with fewer comorbid conditions, or those who are able to bear weight more quickly following a fracture, may show faster and better recovery of abilities.

Researchers (9-12) examining health status in community-dwelling Hispanic people have found higher levels of activity limitation and disability in this population than in nonminority populations. A few researchers have examined the association between physical, cognitive, and emotional factors and health status in Hispanic populations. After studying living arrangements and multiple measures of physical, cognitive, and emotional function in a nationally representative sample of white, black, and Hispanic adults, Waite and Hughes (10) concluded that, in general, older Hispanic adults tended to have more physical, cognitive, and emotional impairments than older non-Hispanic adults. This disadvantage, however, seems to vary according to the living arrangement of the individual. When compared with non-Hispanic adults, married Hispanic adults living in households with others (eg, children, parents) show higher levels of self-rated physical and emotional health, less difficulty performing functional tasks, and less depression. (10) Single Hispanic men and women 51 to 61 years of age who are living alone showed more depression and poorer cognitive function than non-Hispanic adults. (10) Thus, there seems to be some interaction between mental health status and social support factors among Hispanic adults. Cultural differences in the value of family relationships and different cultural preferences may lead to different expectations in the availability and presence of social support in time of need. (13) Social support systems and psychological status differ in Hispanic and non-Hispanic populations, and these differences have rarely been accounted for in previous studies of disability and health status. (3-8)

The purpose of my study was to determine factors related to activity limitations in a group of Cuban Americans recovering from hip fracture. Activity limitation was examined at 3 points in time: prior to hip fracture, at the time of discharge from a rehabilitation facility, and 2 months after discharge from the rehabilitation facility. The theoretical basis for my study is the expanded disablement model described by Guccione (2) and in the Guide to Physical Therapist Practice. (1) This model suggests that a number of individual, environmental, and health-related factors might interact to mediate the disablement process. In my study, I examined the role of several of these factors, including social support, psychological status, and comorbidity, as determinants of activity limitations before and after hip fracture. I also examined the role of factors such as weight-bearing status, age, and sex in recovery of physical abilities after a hip fracture.

Method

Subjects

Subjects were recruited from an inpatient rehabilitation unit within a 300-bed private hospital in an urban community of south Florida. Subjects entered the cohort at the time of discharge from the rehabilitation facility. After they were identified by a case manager, I approached them to elicit voluntary participation in this study and to obtain informed consent. Inclusion criteria included being men and women who identified themselves as Cuban American, 50 years of age or older with a hip fracture that required surgical repair, and admitted from home to a rehabilitation hospital at the time of the hip fracture and then discharged to the home following a stay at the rehabilitation hospital. Subjects who were unable to complete the instruments due to cognitive deficits (as determined by the case manager's assessment) or language barriers (speaking a language other than English or Spanish) were excluded from the study. Subjects with a known previous comorbidity identified in the medical record that severely affected quality of life (eg, cancer, major depression, psychosis, multiple sclerosis) and subjects undergoing elective total hip arthroplasty also were excluded from the study.

Measurement

Activity limitations prior to hip fracture and activity limitations 2 months after discharge from the rehabilitation hospital were measured by use of the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) physical function subscale (PF-10). The SF-36 is designed to test general health status in 8 domains. (14) For the purpose of this study, an English-Spanish translation of the SF-36 was used. This translation is unpublished, but was obtained from the International Quality of Life Assessment Project (B Gandek; personal communication; April 26, 2001). Reliability and construct validity of measurements obtained with this instrument have been demonstrated in groups of Cuban-American men with varying severity of benign prostatic hypertrophy. (15,16) Validity testing and normative testing of all subscales of the SF-36 have been performed with a large nationally representative sample of Americans. (14) Subgroups with chronic diseases such as osteoarthritis, diabetes, and hypertension were analyzed in this testing, but the researchers did not examine subgroups of patients with acute diseases and disorders. (14)

The PF-10 is used to measure degree of limitation in performing vigorous and moderate activities of daily living, walking, climbing stairs, and self-care. The PF-10 contains 10 items rated on a 3-point scale so that high scores indicate less activity, limitation and low scores indicate more activity limitation. (14) Raw scores were transformed to a scale ranging from 0 to 100, as described in the SF-36 user's manual as standard use of the instrument. (14)

Activity limitations at the time of discharge from the rehabilitation facility were measured by use of the mobility subscale of the Functional Independence Measure (FIM). The FIM is part of the Uniform Data Set (UDS), (17) which was developed to document the severity of physical disability and measure the outcomes of rehabilitation for people following a stroke. This instrument has been shown to yield reliable and valid measurements of rehabilitation outcomes in a sample of 11,102 patients receiving inpatient rehabilitation in the northwestern United States. (18) The FIM is designed to measure the degree of assistance an individual needs to perform basic mobility, locomotion, and self-care activities, and it examines 11 items on a 7-point scale, with high scores indicating less need for assistance in performing activities.

Mental health status was measured using the mental health and role-emotional subscales of the SF-36. (14) The mental health subscale measures the frequency at which an individual feels depressed, nervous, calm, or happy. This subscale contains 5 items that are measured on a 6-point scale ranging from "all of the time" to "none of the time." Higher scores indicate that the individual frequently feels calm and happy and infrequently feels depressed and nervous. (14) The role-emotional subscale measures the extent to which emotional problems interfere with the kind or amount of work that an individual does. This subscale consists of 3 items measured on a 4-point scale, with higher scores indicating no problems with work or other daily activities as a result of emotional problems and lower scores indicating problems. (14) Raw scores were transformed to a scale ranging from 0 to 100 according to SF-36 criteria, and scores for the 2 subscales were summated.

Social support was measured by a survey instrument derived from and utilized in the MacArthur Studies of Successful Aging. (19) This survey instrument was used to measure the amount of social support received from 4 different sources: spouse, children, relatives, and friends. The amount of social support received from each source in several domains, including instrumental support (physical or tangible support), emotional support, and negative support, also was measured. (19) The emotional support subscale measures the frequency at which each source of support makes the individual feel loved and cared for and the extent to which people listen to the individual's problems. The instrumental support subscale measures the frequency at which the individual receives physical or financial assistance from each source. The negative support subscale measures the frequency at which the individual feels that each support source is demanding or critical. Each subscale produces ordinal measurements and has a possible range of scores from 6 to 24. (19) For instrumental support and emotional support, higher scores indicate more support. Higher scores for negative support indicate more feelings that the source is demanding or critical. The scores obtained with these subscales have been found to be reliable and valid in describing the quantity and quality of social support in older community-dwelling adults. (19) The reliability and validity of data obtained with this social support scale have not been examined in Cuban-American or other Hispanic populations.

Comorbidity was measured by the number of other coexisting diagnosed medical conditions identified by the subjects utilizing a checklist of diagnoses and the UDS. (17) Weight-bearing status, according to documentation in the medical record, was noted at the time of discharge from the rehabilitation facility.

Procedure

Data were collected at 4 points in time: before fracture (data were collected retrospectively by interview at the time of discharge from the rehabilitation hospital), at the time of admission to the rehabilitation hospital (data were extracted from quality assurance records), at the time of discharge from the rehabilitation hospital (data were collected by interview and extracted from quality assurance records), and 2 months after discharge from the rehabilitation hospital (data were collected by interview). Table 1 summarizes the data for each time point. Each subject was interviewed 2 times: 1 or 2 days before discharge from the rehabilitation facility and 2 months after discharge from the rehabilitation facility. All interviews were conducted in Spanish, and all survey instruments were in Spanish. During the first interview, subjects were instructed to complete the SF-36 with reference to the 4 weeks preceding their injuries. Subjects also completed the social support interview and demographics survey at this time. The second interview was conducted in the subject's home 2 months after discharge from the rehabilitation facility. At this time, subjects completed the SF-36 a second time with reference to the previous 4 weeks. All other data were obtained from quality assurance records.

Data Analysis

Multiple linear regression analyses using the backward-deletion process were used to determine the predictors of activity limitation at 3 time points. (20(p81)) Independent variables were those that have been identified as mediators in the disablement process, including psychosocial variables (mental health status and social support), demographic variables (age and sex), and medical variables (number of comorbidities and weight-bearing status). Activity limitations before the fracture were measured retrospectively using the SF-36 physical function subscale to reflect activity limitations. Independent variables that were measured included mental health status, 3 measures of social support, age, sex, and number of comorbid conditions. Activity limitations at the time of discharge from the rehabilitation facility were measured using the motor subscale of the FIM. The same independent variables were used to identity the predictors of activity limitation at the time of discharge from the rehabilitation hospital. The FIM motor subscale scores at the time of admission to the rehabilitation facility and weight-bearing status were included in this model to control for severity of activity limitations at admission. Activity limitations 2 months after discharge from the rehabilitation facility were measured using the SF-36 physical function subscale, and activity limitations before the fracture were included in the model to control for previous differences. Independent variables included concurrent mental health, 3 measures of social support, age, sex, number of comorbidities, and weight-bearing status.

For each point in time, a full model was developed using all of the variables, and variables that did not contribute to the model were deleted one at a time to develop the final model in which all of the variables were significant at P<.10. In this way, independent variables that were related to each of the dependent variables were identified, and variables that were not related were deleted from the model. For each of the models, collinearity diagnostics were run, including calculation of eigenvalues, variance inflation factors, and condition indices. (21) Data were analyzed using SAS statistical software. *

Thirty-seven subjects participated in the first interview, and 28 subjects participated in the second interview. A total of 9 subjects were lost to follow-up. Six subjects were lost to follow-up due to their refusal to participate in the interview process or due to inability to contact them, and 3 subjects were hospitalized and unable to participate in the interview process. Therefore, data for 28 subjects were used in the final data analysis. The subjects were 82.1% female and had a mean age of 78.7 years (SD=9.4). They had a median of 3 comorbid conditions. All of the subjects were of Hispanic origin (Tab. 2). The mean score for premorbid activity limitations was 70.7 (SD=34.0) (n=37) on a scale of 0 to 100, and the mean activity limitation score 2 months after discharge from the rehabilitation facility was 33.6 (SD=20.9) (n=28). The mean FIM mobility subscale score at the time of admission to the rehabilitation facility was 34.6 (SD=5.4) (out of a possible 77), and the mean FIM mobility subscale score at the time of discharge from the rehabilitation facility was 51.8 (SD=7.0).

I performed t-test analyses to compare prefracture scores of the subjects who were lost to follow-up with the scores of subjects whose data were included in the final analysis (Tab. 2). The 9 subjects who were lost to follow-up did not differ from the subjects who participated in both interviews except in number of comorbid conditions. The subjects who were lost to follow-up had fewer comorbid conditions than did the final study group.

Results

The results of multilinear regression using the backward-deletion process are presented in Tables 3 through 5. Final models were developed that accounted for 31% of the variance in prefracture activity limitations, 53% of the variance in activity limitations at the time of discharge from the rehabilitation hospital, and 43% of the variance in activity limitations 2 months after discharge from the rehabilitation hospital. The partial correlations indicate the relative contribution of each of the variables to the model; variables with larger partial correlations explain more of the variance in the model than do variables with smaller partial correlations. An assumption of multilinear regression is that the predictors or independent variables are correlated to the dependent variable, but are not correlated among themselves. When there are intercorrelations among the predictors, the importance of a single factor is confounded by the other factors with which it is correlated. Thus, it is important to determine whether multicollinearity exists in the model. (20(p76)) Results of collinearity diagnostics indicated that there was no evidence of collinearity in any of the 3 models. All of the variance inflation factors were less than 10, none of the eigenvalues approached 0, and none of the condition indices approached 30.

Most of the variance in prefracture activity limitations was attributed to age followed by mental health status and sex. More prefracture activity limitations were related to older age, poorer mental health, and male sex. None of the social support variables contributed to the model. Comorbidity also did not contribute to the model.

Most of the variance in activity limitations at the time of discharge from the rehabilitation facility was related to activity limitations at the time of admission, sex, and age, but weight-bearing status and mental health status also contributed to the model. More activity limitations at the time of discharge were associated with more activity limitations at the time of admission, status for weight bearing as tolerated, older age, and male sex. Variables that did not contribute to the model included all of the social support variables, prefracture activity limitations, and number of comorbid conditions.

Most of the variance in activity limitations 2 months after discharge from the rehabilitation facility was accounted for by 3 social support variables: instrumental (physical) support, emotional support, and negative support. Partial weight-bearing status also contributed to the model. More activity limitations 9 months after discharge were related to more instrumental support, less emotional support, and less negative support; status for weight bearing as tolerated or non-weight bearing (versus partial weight bearing); and older age. Social support accounted for 43% of the variance in activity limitations at 2 months after discharge, such that those subjects with more emotional support and more negative support were less limited and those subjects with more instrumental support were more limited.

Discussion

The expanded disablement model, as proposed by Guccione, (2) recognizes social interactions and relationships, and psychological status or mental health status, as contributors to the process of disablement. The results of my study provide evidence to support the relationships between both mental health status and social support and activity limitations before and after hip fracture, but these relationships seem to change throughout the recovery process.

My results indicate that mental health status, defined in terms of the frequency of an individual's report of feelings of depression or anxiety and the extent to which emotional health limited activities, was related to activity limitations both before the fracture and at the time of discharge from the rehabilitation hospital. Mental health status, however, was not related to activity limitations 2 months after discharge from the rehabilitation facility.

The findings related to mental health status are consistent with those of several other studies that have investigated depressive symptoms and functional decline. (22-24) Bruce et al (22) showed that depressive symptoms represented a risk factor for functional limitations even in the absence of acute medical extents. Penninx and colleagues (23,24) replicated this finding and found that chronicity of depression had an impact on decline in physical function over time. Treatment and remission of depression was found to be protective against future decline. These researchers' findings support my findings that mental health status and activity limitations are associated and that subjects with more depressive symptoms tend to have more activity limitations prior to hip fracture. Because Hispanic Americans (25) and Cuban Americans (26) have higher levels of depression than non-Hispanic people, these are important culturally and ethnically dependent findings.

My results suggest that mental health status also is related to activity limitations at the time of discharge from a rehabilitation facility. Other researchers (3,6-8) have shown that depressive symptoms are related to recovery front hip fracture. In 3 of these studies, the influence of depression 12 months after fracture was examined, and it was found that individuals reporting fewer depressive symptoms were more likely to achieve independence in walking and other functional tasks (6) and to have better recovery in both functional status (7,8) and psychosocial status. (7) Mutran et al (3) investigated functional outcomes 2 months after hip fracture surgery and found that inadequacy of social support and depression resulted in less improvement in walking ability than occurred with more social support and less depression, but the researchers did not examine other types of activity limitations. Mental health status appears to mediate the process of recovery after hip fracture in the general population, and my study suggests that it also might be particularly important in the Cuban-American population because of higher prevalence of depression in this population. (10)

None of the social support variables were related to prefracture activity limitations or activity limitations at the time of discharge from the rehabilitation facility, but social support was related to activity limitations 2 months after discharge. Previously, researchers (22-29) have shown that social support is related to activity limitations in the community-dwelling elderly population, whereas other researchers (30) have not been able to demonstrate this association.

Research by Berkman et al (30) failed to support the relationship between social support and functional level. In that study, subjects were grouped into 3 cohorts (high, medium, and low function) based on their performance on a series of cognitive and physical tests. Neither levels of emotional support, instrumental support, and negative support nor support size were different in these 3 groups. (30) Mendes de Leon et al (31) found that being "embedded" in a network of social relationships provides long-term protection against disability (activity limitations), either by reducing the risk of developing disability or by promoting recovery from disability. Further study of the effects of quality and quantity of social support on activity limitations in the absence of a medical event is needed. Because of variations in social norms and customs in different cultural groups, it is important to examine these relationships in culturally diverse populations.

Instrumental support, such as having someone to help with activities of daily living, was associated with more activity limitations, and emotional support, such as feeling cared for and having someone to listen to you when you have problems, was associated with less activity limitations 2 months after discharge from the rehabilitation facility. Cummings et al (5) found social support to be important in recovery of function 6 months after hip fracture, such that individuals with larger social support networks were less limited. Cummings et al, however, did not differentiate among types of social support. Mutran et al (3) found that social support influenced the recovery process primarily 2 months following injury and was less important 6 months following injury. Similarly, Wilcox et al (4) found that emotional support was more important early in the process of recovery from hip fracture (6 weeks) and that instrumental support was more important later in the recovery process (6 months). In my study, high levels of instrumental support might have been the cause of more activity limitations and physical dependence because other people in the support network are available to provide them with the help that they need. Based on my data and the findings of previous studies, appropriate type and timing of social support appears to be highly important in facilitating recovery front hip fracture.

I found that negative support, such as a feeling of demand or criticism front the social network, was inversely related to hip fracture recovery 2 months after discharge from the rehabilitation facility, such that those with more negative support were less limited. Both criticism and feelings of being needed may facilitate the process of recovery by encouraging the individual to resume prefracture activities independently as soon as possible. No previous studies have investigated the influence of negative aspects of social support on recovery from hip fracture.

Examination of the scores for activity limitations at the 3 time points by weight-bearing status reveals that weight-bearing status was related to activity limitations 2 months after discharge from the rehabilitation facility. Those subjects who had full weight bearing at the time of discharge from the rehabilitation facility had the most activity limitations 2 months after discharge, and those who had partial weight bearing at discharge had the least activity limitations (Tab. 6). Examination of prefracture activity limitations shows that the trend in activity limitations 2 months after discharge from the rehabilitation facility was the same as the trend in prefracture activity limitations: subjects who were weight bearing as tolerated at the time of discharge were more limited prior to hip fracture. An individual's initial status for weight bearing as tolerated appears to offer an advantage during the acute stage of recovery, but that advantage disappears in the later stages. The importance of weight-bearing stares in predicting activity limitations 2 months after discharge from the rehabilitation facility may be a reflection of the prefracture differences among the 3 groups.

The expanded disablement model also suggests that comorbidity has an effect on the process of disablement, and many previous researchers (3,7,32,33) have suggested that comorbidity is related to recovery from hip fracture. In my study, comorbidity was not associated with activity limitations at any of the 3 time points. Magaziner et al (34) also were unable to find an association between comorbid conditions and activity limitations, but they looked at 12-month rather than 2-month recovery. Most researchers who have found a relationship between comorbidity and recovery have looked at long-term recovery at 6 to 12 months. Perhaps comorbidity is a predictor of recovery at the end of the recovery process, but it is less important in the middle of the recovery process. The effect of comorbidity may be due to the fact that previous researchers used large population-based cohorts that were followed prospectively, whereas I used a cohort that was admitted for intensive rehabilitation. Presumably, my cohort would have had people with fewer comorbid conditions than the general population of people who fracture their hips because people with many comorbid conditions would likely be placed in skilled nursing facilities or other institutions rather than in intensive rehabilitation programs.

Prefracture activity limitations were not a predictor of activity limitations at the time of discharge from the rehabilitation facility or 2 months after discharge. This relationship is contrary to the findings of other researchers (3,6,8,35) who found that prefracture activity limitations were related to activity limitations after hip fracture. That my study did not confirm the relationship between prefracture activity limitations and activity limitations 2 months after discharge from the rehabilitation facility might be explained by the fact that this effect is more powerful later in the recovery process (6-12 months). It also could be that this effect is population-specific, and it is impossible to determine if this is the case with the design of this study. Most of the recovery in functional abilities occurs in the first 6 months after hip fracture. (33) A severe traumatic injury, such as hip fracture, may be an equalizer between those who were quite active and those who were quite limited at the time of fracture, so that in the period immediately following hip fracture, the acute effect of the fracture on activity limitations masks the effect of prefracture differences in activity limitations. Some of the effects for prefracture activity limitations and comorbidities may have been less apparent in my sample because of the homogeneity of this group (ie, Cuban-Americans), as they were all living at home, were cognitively functional, qualified for admission to the rehabilitation program, and were discharged to their homes.

The expanded disablement model also proposes that a person's sex has an effect on the process of disablement. The results of my study showed that sex was a predictive factor in prefracture activity limitations and activity limitations at the time of discharge from the rehabilitation facility. Although sex was a predictor of activity limitations at the time of discharge, activity limitation scores for men and women were the same (FIM mobility subscale scores of 52.0 and 51.5, respectively). The fact that women tended to have much higher levels of depressive symptoms than did men probably accounts for the importance of sex in the model, as all of the other variables in the model (length of stay, hours of therapy, and activity limitations at admission) were similar in men and women. Although all predictive models included sex as a control, there was a potential sex bias in my study because only 16% of the sample was male.

There are several limitations to my study. Prefracture activity limitations and activity limitations 2 months after discharge from the rehabilitation facility were measured using the SF-36, and activity limitations at discharge were measured using the FIM mobility scale. These scales have different measurement properties and constructions that lead to some difficulty in interpreting the findings in a clinical context. For example, the FIM measures the assistance that a subject needs to perform a particular task, whereas the SF-36 measures how much assistance a subject feels that he or she needs to complete a particular functional task. In addition, the FIM is an instrument with which measurements are made by direct observation, whereas the SF-36 relies on self-report. Although these measurement scales are different from each other, the fact that age, sex, and weight-bearing status were predictors across 2 time frames lends some credibility to the use of different measurement scales. Another limitation to my study was that the prefracture health information was collected retrospectively, rather than prior to admission to the rehabilitation facility. This may have led to some bias due to subjects' poor recollection or interpretation of their prior level of function. Finally, the small number of subjects enrolled in the study leads to some limitation in generalization. The validity of the statistical inferences made in this article may be affected by the small sample size because there may be instability in the statistical models. The small number of subjects also may lead to some difficulty in generalizing the findings of this study to larger groups.

Conclusion

I found that psychosocial variables were mediators of activity limitations in my sample of Cuban-American men and women recovering from hip fracture. Mental health status was related to activity limitations before hip fracture, but not after hip fracture. Social support was related to activity limitations 2 months after hip fracture, but not before hip fracture or early in the recovery process. The effects of emotional versus instrumental support were different. Instrumental support was related to more activity limitations, and emotional support was related to less activity limitations after hip fracture. Older age and male sex also were related to more activity limitations at varying times in the recovery process.

High levels of social support are thought to be helpful in recovery from hip fracture. (3-5) In my sample of Cuban-American adults, however, high levels of instrumental support were detrimental in recovery of activity levels during the later stages of rehabilitation. In contrast, high levels of emotional support and the presence of some negative aspects of support, such as criticism and demand from the social network, seemed to offer an advantage in the later phases of recovery. Although several researchers demonstrated that mental health status was related to recovery of activity levels after hip fracture, this relationship was important only in the early stages of recovery. Perhaps there are cultural differences in the meaning of disability and disease that mediate the effect of poor mental health status on recovery in this group.

The expanded disablement model discussed here suggests that individual factors such as mental health status and social processes and relationships may have an effect on the process of disablement. The results of my study provide evidence supporting the expanded disablement model. The data describe the relationships among factors in the model in a distinct cultural group, a group of Cuban-American adults. Perhaps cultural variations in mental health, social support, and activity limitations lead to some differences between Cuban Americans and other groups regarding the importance of these factors on recovery from hip fracture. Future studies of this nature should include measures of physical and cognitive impairment and health habits and lifestyle and should include a comparison group of nonminority adults. In addition, the effects of these individual and environmental factors on disability, as defined by the Guide to Physical Therapist Practice, (1) should be investigated.

Table 1.
Data Collected at Each Point in Time During Recovery From
Hip Fracture (a)

                                                       2 Months After
Data Reflecting   Admission to     Discharge From      Discharge From
Prefracture       Rehabilitation   Rehabilitation      Rehabilitation
Status            Facility         Facility            Facility

SF-36 physical    Comorbidities    Length of stay      SF-36 physical
  function                                               function
  subscale                                               subscale
SF-36 mental      FIM mobility     Hours of physical   SF-36 mental
  health and        subscale at      therapy and         health and
  vitality          time of          occupational        vitality
  subscales         admission        therapy             subscales
Social support                     FIM mobility
  interview                          subscale at
Demographics                         time of
                                     discharge

(a) SF-36=Medical Outcomes Study 36-Item Short-Form Health Survey,
FIM=Functional Independence Measure.

Table 2.
Comparison Between Subjects Who Dropped Out (n=9) and Subjects Who
Completed the Study (n=28)

                     Dropped Out      Completed
                     (n=9)            Study (n=28)

                                                       P Value
                                                       for t Test
Variable (a)         [bar.X]   SD     [bar.X]   SD     (df=35)

Age (y)              83.7       5.3   78.7       9.4   .13
Comorbidity (n)       1.8       1.8    3.2       1.8   .07
Social support
  Instrumental        9.1       3.5    8.8       4.1   .81
  Emotional          11.0       3.0   11.7       3.2   .56
  Negative            1.4       3.3    1.7       2.5   .80
Prefracture
    health status
  Mental health      75.9      24.5   77.5      15.8   .82
    status (SF-36)
  Activity           68.8      34.0   70.7      30.4   .87
    Limitations
    (PF-10)

(a) SF-36=Medical Outcomes Study 3-Item Short-Forms Health Survey,
PF-10=physical function subscale of the SF-36.

Table 3.
Predictors of Prefracture Activity Limitations

Full Model, n=28, Adjusted [r.sup.2]=.320, P for F
Statistic=.02 (df=27)

                                                 P Value for
                       Parameter   Partial       t Statistic
Variable               Estimate    Correlation   (df=1)

Negative support        -0.28      .12           .92
Emotional support       -2.21      .03           .42
Instrumental support     2.97      .04           .14
Mental health status     0.17      .10           .24
Male sex               -37.62      .03           .05
Age                     -1.83      .03           .03
Comorbidities           -4.15      .08           .15

Final Model, n=28, Adjusted [r.sup.2]=.308, P=.008

                                                 P Value for
                       Parameter   Partial       t Statistic
Variable               Estimate    Correlation   (df=1)

Mental health status     0.47      .10           .07
Male sex               -34.64      .02           .03
Age                     -1.95      .30           .003

Table 4.
Predictors of Activity Limitations at Time of Discharge From
Rehabilitation Facility

Full Model, n=28, Adjusted [r.sup.2]=.462, P=.04, (df=27)

                                                    P Value for
                          Parameter   Partial       t Statistic
Variable                  Estimate    Correlation   (df=1)

Activity limitations at    0.40       .31           .16
  admission
Prefracture activity       0.04       .10           .46
  limitations
Prefracture mental         0.05       .01           .12
  health status
Instrumental support      -0.62       .11           .17
Emotional support          0.73       .02           .23
Negative support          -0.29       .02           .67
Partial weight bearing     2.45       .02           .50
Weight bearing as          7.32       .19           .06
  tolerated
Comorbidities              0.001      .02           .99
Age                       -0.46       .09           .06
Male sex                  -7.32       .12           .12

Final Model, n=28, Adjusted [r.sup.2]=.526, P=.0007

                                                    P Value for
                          Parameter   Partial       t Statistic
Variable                  Estimate    Correlation   (df=1)

Activity limitations       0.55       .31           .06
  at admission
Prefracture mental         0.15       .06           .03
  health status
Weight bearing as          4.82       .08           .02
  tolerated
Age                       -0.42       .12           .003
Male sex                  -8.63       .27           .02

Table 5.
Predictors of Activity Limitations 2 Months After Discharge From
Rehabilitation

Full Model, n=28, Adjusted [r.sup.2]=.434, P=.02 (df=27)

                                                        P Value for
                              Parameter   Partial       t Statistic
Variable                      Estimate    Correlation   (df=1)

Prefracture activity            0.05      .01           .69
  limitations
Mental health status at         0.08      .00           .34
  2 months
Instrumental support           -3.43      .15           .02
Emotional support               4.77      .00           .02
Negative support                4.97      .30           .01
Comorbidity                    -0.43      .00           .83
Male sex                      -15.80      .08           .23
Partial weight bearing         14.68      .21           .12
Weight bearing as tolerated     9.80      .06           .35
Age                             0.54      .13           .43

Final Model, n=28, Adjusted [r.sup.2]=.440, P=.0014 (df=27)

                                                        P Value for
                              Parameter   Partial       t Statistic
Variable                      Estimate    Correlation   (df=1)

Instrumental support           -2.99      .15           .02
Emotional support               3.45      .00           .03
Negative support                4.11      .28           .003
Partial weight-bearing         17.04      .22           .02

Table 6.
Activity Limitations at 3 Time Points by Weight-Bearing
Status (n=28) (a)

                                        2 Months
                                        After
       Prefracture      Discharge       Discharge
       (SF-36)          (FIM)           (SF-36)

       [bar.X]   SD     [bar.X]   SD    [bar.X]   SD

NWB    76.9      37.1   48.1      2.5   31.2       9.9
PWB    81.9      18.3   55.2      5.3   43.8      28.6
WBAT   59.2      30.2   52.2      8.8   28.3      19.3

* SAS Institute Inc, PO Box 8000, Cary, NC 27511.

References

(1) Guide to Physical Therapist Practice. 2nd ed. Phys Ther. 2001;81: 25-33.

(2) Guccione AA. Arthritis and the process of disablement. Phys Ther. 1994;74:408-414.

(3) Mutran EJ, Reitzes DC, Mossey J, Fernandez ME. Social support, depression, and recovery of walking ability following hip fracture surgery. J Gerontol B Psychol Sci Soc Sci. 1995;50:S354-S361.

(4) Wilcox VL, Kasl SV, Berkman LF. Social support and physical disability in older people after hospitalization: a prospective study. Health Psychol. 1994;13:170-179.

(5) Cummings SR, Phillips SL, Wheat ME, et al. Recovery of function after hip fracture, the role of social supports. J Am Geriatr Soc. 1988;36:801-806.

(6) Mossey JM, Mutran EJ, Craik R. The effect of persistent depressive symptoms on hip fracture recovery. J Gerontol. 1990;45:M163-M168.

(7) Mossey JM, Mutran EJ, Knott K, Craik R. Determinants of recovery 12 months after hip fracture: the importance of psychosocial factors. Am J Public Health. 1989;79:279-286.

(8) Marotolli RA, Berkman LF, Cooney LM. Decline in physical function following hip fracture. J Am Geriatr Soc. 1992;40:861-866.

(9) Tucker KL, Falcon LM, Bianchi LA, et al. Self-reported prevalence and health correlates of functional limitation among Massachusetts elderly Puerto Ricans, Dominicans, and a non-Hispanic white neighborhood comparison group. J Gerontol A Biol Sci Med Sci. 2000;55: M90-M97.

(10) Waite LJ, Hughes ME. At risk on the cusp of old age: living arrangements and functional status among black, white, and Hispanic adults. J Gerontol B Psychol Sci Soc Sci. 1999;54:S136-S144.

(11) Stump TE, Clark DO, Johnson RJ, Wolinski FD. The structure of health status among Hispanic, African-American, and white older adults. J Gerontol B Psychol Sci Soc Sci. 1997;52:S49-S60.

(12) Jette AM, Crawford SL, Tennstedt. SL. Toward understanding ethnic differences in late-life disability. Res Aging. 1996;18:292-309.

(13) Himes CL, Hogan DP, Eggebeen DJ. Living arrangements of minority elders. J Gerontol B Psychol Sci Soc Sci. 1996;51:S42-S48.

(14) Ware JE, Kosinski M, Gandek B. SF-36 Health Survey: Manual and Interpretation Guide. Lincoln, RI: QualityMetric Inc; 2000.

(15) Arocho R, McMillan CA, Sutton-Wallace P. Construct validation of the USA-Spanish version of the SF-36 Health Survey in a Cuban-American population with benign prostatic hyperplasia. Qual Life Res. 1998;7:121-126.

(16) Arocho R, McMillan CA. Discriminant and criterion validation of the US-Spanish version of the SF-36 Health Survey in a Cuban-American population with benign prostatic hyperplasia. Med Care. 1998;36:766-772.

(17) Guide for the Uniform Data Set for Medical Rehabilitation. Version 5.0. Buffalo, NY: State University of New York at Buffalo; 1996.

(18) Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the Functional Independence Measure and its performance among rehabilitation inpatients. Arch Phys Med Rehabil. 1993;74:531-535.

(19) Seeman TE, Berkman LF, Blazer D, Rowe JW. Social ties and support and neuroendocrine function: the MacArthur Studies of Successful Aging. Ann Behav Med. 1994;16:95-106.

(20) Stevens J. Applied Multivariate Statistics for the Social Sciences. 3rd ed. Mahwah, NJ: Lawrence Erlbaum Associates; 1996:76, 81.

(21) SAS/STAT User's Guide, Version 6. 4th ed. Vol 2. Cary, NC: SAS Institute Inc; 1989:1417.

(22) Bruce ML, Seeman TE, Merrill SS, Blazer DG. The impact of depressive symptomatology on physical disability: MacArthur Studies of Successful Aging. Am J Public Health. 1994;84:1796-1799.

(23) Penninx BW, Guralnik JM, Ferrucci L, et al. Depressive symptoms and physical decline in community-dwelling older persons. J Am Med Assoc. 1998;279:1720-1726.

(24) Penninx BW, Deeg DJ, van Eijk JT, et al. Changes in depression and physical decline in older adults: a longitudinal perspective. J Affect Disord. 2000;61:1-12.

(25) Mills TL, Henretta JC. Racial, ethnic, and sociodemographic differences in the level of psychosocial distress among older Americans. Res Aging. 2001;23:131-152.

(26) Llorente MD, Eisdorfer C, Loewenstein DA, Zarate YA. Suicide among Hispanic elders: Cuban Americans in Dade County, Florida 1990-1993. J Ment Health Aging. 1996;2:79-87.

(27) Seeman TE, Berkman LF, Charpentier PA, et al. Behavioral and psychosocial predictors of physical performance: MacArthur Studies of Successful Aging. J Gerontol A Biol Sci Med Sci. 1995;50:M177-M183.

(28) Seeman TE, Bruce ML, McAvay GJ. Social network characteristics and onset of ADL disability: MacArthur Studies of Successful Aging. J Gerontol B Psychol Sci Soc Sci. 1996;51:S191-S200.

(29) Unger JB, McAvay G, Bruce ML, et al. Variation in the impact of social network characteristics on physical functioning in elderly persons: MacArthur Studies of Successful Aging. J Gerontol B Psychol Sci Soc Sci. 1999;54:S245-S251.

(30) Berkman LF, Seeman TE, Albert M, et al. High, usual, and impaired functioning in community-dwelling older men and women: findings from the MacArthur Foundation Research Network on Successful Aging. J Clin Epidemiol. 1993;46:1129-1140.

(31) Mendes de Leon CF, Glass TA, Beckett LA, et al. Social networks and disability transitions across eight intervals of yearly data in the New Haven EPESE. J Gerontol B Psychol Sci Soc Sci. 1999;54:S162-S172.

(32) Kirk-Sanchez NJ, Roach KE. Relationship between duration of therapy services in a comprehensive rehabilitation program and mobility at discharge in patients with orthopedic problems. Phys Ther. 2001;81:888-895.

(33) Craik RL. Disability following hip fracture. Phys Ther. 1994;74: 387-398.

(34) Magaziner J, Simonsick EM, Kashner TM, et al. Survival experience of aged hip fracture patients. Am J Public Health. 1989;79:274-278.

(35) Young Y, Brant L, German P, et al. A longitudinal examination of functional recovery among older people with subcapital hip fractures. J Am Geriatr Soc. 1997;45:288-294.

NJ Kirk-Sanchez, PT, PhD, is Assistant Professor, Department of Physical Therapy, School of Health, College of Health and Urban Affairs, Florid International University, University Park, Miami, FL 33199 (USA) ([e-mail]).

This work was completed as part of Dr Kirk-Sanchez's doctoral dissertation at the University of Miami, Coral Gables, Fla.

This study was supported by a PODS-II Scholarship from the Foundation for Physical Therapy.

This study was approved by the University of Miami Medical Sciences Subcommittee for the Protection of Human Subjects in Research and the institutional review board of the hospital from which the cohort was drawn.

This article was received May 24, 2002, and was accepted October 29, 2003.