| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
a Department of Health Care Studies, Section of Medical Sociology, Maastricht University, The Netherlands
b Northern Center for Healthcare Research, University of Groningen, The Netherlands
c Department of Social Psychiatry, University of Groningen, The Netherlands
Gertrudis I. J. M. Kempen, Department of Health Care Studies, Section of Medical Sociology, Maastricht University, P.O. Box 616, 6200 MD Maastricht, the Netherlands E-mail: G.Kempen{at}zw.unimaas.nl.
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Patients (N = 171) who had sustained fall-related injuries (hip fractures, other fractures, or sprains and dislocations) participated in the study. Disability scores were collected at baseline (before the injury) and 8 weeks, 5 months, and 12 months after the injury. The authors used analysis of variance to assess possible differences between 3 levels of education and social support with respect to changes in disability scores from baseline to the 3 follow-up measurements while adjusting for covariates.
Results. Preinjury assessed educational level or social support did not play a role in short-term changes in disability. In the long term (5 and 12 months after the injury), recovery was significantly associated with social support: Those with higher levels of support had a better recovery. Although patients with high levels of education most closely approached their pre-event level of disability as well, differences did not reach statistical significance. Short-term changes in disability appeared to be determined by the severity of the injury. Social support began to influence recovery only when the impact of severity expired.
Discussion. Patients recovering from fall-related injuries who had reported high levels of social support before their injury had recovered better at 5 and 12 months. Encouragement and special attention given by health professionals to maintain social support may be beneficial for rehabilitation after fall-related injuries in older persons.
FALLS by elderly persons are a common problem. Community-based studies report that 23.1–39.2% of those living independently and 20–44.9% of institutionalized elderly persons fall at least once a year (
Ryynanen, Kivela, Honkanen, Laippala, and Soini 1991
;
Stalenhoef, Crebolder, Knottnerus, and Van der Horst 1997). About 5% of falls result in fractures (1% in hip fractures) and 5% in other injuries (
Stalenhoef et al. 1997). The short- and long-term consequences of nonfatal fall injuries for mobility and independence in activities of daily living (ADLs) and instrumental activities of daily living (IADLs) are substantial. Only half of patients with a hip fracture, for example, regain pre-event levels of mobility, and 33–74% regain pre-event ADL and IADL independence (
Cummings et al. 1988;
Jette, Harris, Clearly, and Campion 1987;
Kitamura et al. 1998;
Koval, Skovron, Aharonoff, and Zuckerman 1998;
Kreutzfeldt, Haim, and Bach 1984;
Magaziner, Simonsick, Kashner, Hebel, and Kenzora 1990;
Tinetti et al. 1999). In addition, the use of medical and informal care increases, as does the chance of becoming hospitalized or institutionalized (
Madhock and Green 1993;
Nankhonya, Turnbull, and Newton 1991).
Research on falls by elderly persons has focused mainly on the prevalence of and risk factors for falls. Studies on the determinants of recovery after a fall are scarce. Empirical research indicates that there is considerable variation in the extent to which older patients regain their premorbid levels of ADL and IADL independence. Obviously, medical and biological factors contribute substantially to this variation. Several studies have shown that a patient's general health condition and age influence the recovery process after a hip fracture and that general and technical complications related to surgery influence a patient's chance of recovery (
Broos, Haaften, Stappaerts, and Gruwez 1989;
Broos, Stappaerts, Luiten, and Gruwez 1988;
Ceder, Thorngren, and Wallden 1980;
El Banna, Raynal, and Gerebtzof 1984;
Fox et al. 1998;
Koval et al. 1998;
Mossey, Knott, and Craik 1990;
Mossey, Mutran, Knott, and Craik 1989). However, even when the abovementioned biomedical factors are similar, variation in recovery remains (
Ceder et al. 1980;
Jette et al. 1987). A few studies have suggested that psychological factors may play a role in the recovery process of elderly persons after hip fractures. For instance, depressive symptoms, cognitive dysfunctioning, and postsurgical disorientation appear to have a negative impact on recovery after hip fracture (
Billig, Ahmed, and Kenmore 1988;
Magaziner et al. 1990;
Mossey et al. 1989,
Mossey et al. 1990;
Mutran, Reitzes, Mossey, and Fernandez 1995;
Young, Brant, German, Kenzora, and Magaziner 1997).
To date, little attention has been paid to the role of social factors such as socioeconomic status and social support in the process of recovery after fall-related injuries. The associations of social inequality with health status and functioning in elderly persons have been well documented in the international literature over the years (for a review, see e.g.,
Parker, Thorslund, and Lundberg 1994). Previous studies showed that undesirable events and adverse experiences have stronger negative (emotional) consequences for persons holding lower socioeconomic status positions than for their higher status counterparts (e.g.,
Bebbington, Hurry, Tennant, and Der 1986;
McLeod and Kessler 1990). These findings suggest a different vulnerability for low- and high-status persons. This different vulnerability may be explained by the fact that socioeconomic status shapes the conditions of how individuals cope with stressful situations (
De Ridder 1995).
Rim 1990, for example, found substantial social class differences in coping styles, and
Ranchor, Bouma, and Sanderman 1996 reported that socioeconomic status, in particular educational level, was significantly associated with a range of psychosocial characteristics including several aspects of personality (negative self-esteem, social desirability, hostility). These latter studies showed lower class participants in many respects to be at a disadvantage compared with higher class participants. Socioeconomic status may therefore be an important determinant of outcomes of health events.
Another factor that may play a role in the health of the patient concerns social support. There is some evidence that social support might affect recovery after hip fracture, although the reported results are equivocal.
Magaziner and colleagues 1990 found that the number of contacts with members of the social network during the recovery process had a positive effect. No effect was found with respect to the presence of a spouse or the size of the social network.
Cummings and colleagues 1988, however, did find an effect for the size of a patient's social network. The underlying mechanism is not clear. One may argue that patients with more social contacts have a greater motivation to remain physically active and maintain their mobility and therefore put more effort into rehabilitation (
Cummings et al. 1988). Conversely, patients with less social support may be forced to greater efforts to regain independence. With regard to other possible fall-related injuries than hip fractures, there is a complete lack of research.
In the present study we examined the role of two social factors in recovering ADLs and IADLs after injuries to the extremities related to a fall in older persons: socioeconomic status and social support. More specifically, we studied whether high levels of education and social support speed up the process of regaining preinjury levels of ADL and IADL functioning after a fall. We used a prospective design with a preinjury baseline wave including the assessment of ADL and IADL functioning, social support, and education and three postinjury waves assessing ADL and IADL functioning. We furthermore adjusted for four possible confounders measured at baseline: depressive symptoms, cognitive functioning, chronic medical morbidity, and age. In addition, we adjusted for the severity of the fall-related injury.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
;
Ormel et al. 1997,
Ormel et al. 1998).
For this cohort study, the general practitioners (GPs) reported patients who had sustained injuries to the extremities according to site as coded by the International Classification of Primary Care (ICPC;
Lamberts and Wood 1987). Patients who had completed the baseline assessment and had injuries according to ICPC Codes L72–L80 (hip fractures; fractures of wrist or forearm, ankle or lower leg, and hand or foot; and ankle sprains, knee sprains, or other sprains and dislocations) were included until December 31, 1997. Follow-up interviews were conducted 8 weeks after the injury (short-term impact) and 5 and 12 months after the injury (long-term recovery). The interviews were held by experienced middle-aged female interviewers at the respondents' homes. The interviewers did not know the interviewees in either a clinical or an administrative aspect. At the start of the follow-up interviews, participants completed a shortened version of Folstein's Mini-Mental State Examination (MMSE;
Folstein, Folstein, and McHugh 1975) so that we could evaluate respondents' cognitive capacities for completing the assessment (threshold score = 4;
Breakhus, Laake, and Engedaal 1992). The psychometric properties of this Dutch version of the MMSE have been tested earlier (
Kempen, Brilman, and Ormel 1995). If patients were too ill to complete the assessment at follow-up, proxy interviews (with a maximum of one for each patient) regarding perceptible aspects of participants' physical functioning were conducted with a well-informed person nearby.
Measurement
Disability in ADLs and IADLs was assessed at baseline (before the injury in 1993) and at the three follow-ups with the Groningen Activity Restriction Scale (GARS;
Kempen, Miedema, Ormel, and Molenaar 1996;
Kempen and Suurmeijer 1990). GARS was developed for assessment of disability in the domains of personal care and domestic activities. GARS comprises 18 items referring to both ADLs (personal care, 11 items) and IADLs (household chores, 7 items) with four response options per item (see Table 1 ). The results of previous studies showed that the 18-item GARS meets the stochastic cumulative scalability criteria of the Mokken Model and can thus be considered one dimensional (for detailed information, see
Kempen et al. 1996;
Kempen and Suurmeijer 1990). GARS scores range from 18 (no restriction) to 72 (maximum restriction). The internal reliability estimate in the present study was .91 (n = 171, see the Participants and Response section). Recovery was expressed as the difference between the GARS baseline score and one of the three postevent scores (the latter were subtracted from the first).
|
).
Winkleby, Jatulis, Frank, and Fortmann 1992 found that education was the best socioeconomic predictor of health behavior, disease risk factors, and health. We assessed the level of education according to the International Standard Classification of Education (ISCED;
UNESCO 1976). The index distinguishes six levels of education: no (elementary) school, elementary school, vocational training, high school, undergraduate degree, and graduate degree. The level of education is based on both standard formal education (a very long time ago for most older persons) and vocational courses during adult life. We recorded both during the baseline interview and coded the one that was highest.
Social support interactions were measured at baseline with the 12-item Social Support List (SSL 12-I;
Kempen and Van Eijk 1995). It reflects the extent of perceived support received through interactions with members of a person's primary social network. Examples of the items are "Does it ever happen to you that people ... drop in for a (pleasant) visit ... comfort you ... reassure you ... emphasize your strong point?" Scores on this 12-item scale range from 12 to 48; higher scores indicate more social support. The internal reliability estimate in the present study was .82.
Two scoring methods were used in this study for both educational level and social support. The original scores of both variables were used for descriptive purposes. We used categorical scores with three levels for each of the two variables to study changes in disability in relation to the level of education or social support, respectively. No (elementary) school or elementary school was considered low, vocational training or high school was considered medium, and undergraduate or graduate degree was considered high. The scores for social support were recoded in tertiles of participants in the present study so that we could study changes in disability in relation to the level of social support: A sum score of 23 or lower was considered low, a sum score of 24–27 was considered medium, and a score of 28 or higher was considered high.
Baseline levels of depressive symptoms, cognitive functioning, chronic medical morbidity, disability, age, and severity of the injury were included as covariates because they were found to be related to recovery after hip fracture in previous studies (see introductory section). Depressive symptoms were assessed at baseline with the depression subscale of the Hospital Anxiety and Depression Scale (HADS;
Zigmond and Snaith 1983). The psychometric properties of the Dutch version have been described by
Spinhoven and colleagues 1997. Items referring to symptoms that may have a physical cause (e.g., insomnia and weight loss) are not included in the scale. Therefore, HADS is considered to be unbiased by coexisting general medical conditions (
Spinhoven et al. 1997). Examples of the items are "I still enjoy the things I used to enjoy," "I can laugh and see the funny side of things," "I feel cheerful," and "I feel as if I am slowed down." The theoretical score range of this seven-item scale varies from 0 to 21; higher scores indicate more symptoms. The internal reliability estimate in the present study was .79.
Cognitive functioning was assessed at baseline with the original Mini-Mental State Examination (MMSE;
Folstein et al. 1975). The MMSE consists of 20 items concerning orientation in time and place, naming, repeating, reading, writing, copying, recall, short-term memory, spelling backwards, and performing a three-stage command. Scores range from 0 (severe cognitive impairment) to 30 (no cognitive impairment).
We use a checklist of 19 chronic medical conditions to assess chronic medical conditions at baseline: (a) asthma or chronic bronchitis, (b) pulmonary emphysema, (c) heart disease, (d) hypertension, (e) migraine or chronic headache, (f) (consequences of) stroke, (g) leg ulcer, (h) stomach ulcer, (i) rheumatoid arthritis, (j) (other) back problems or joint conditions, (k) diabetes mellitus, (l) liver disorder or gallstones, (m) prostate disease, (n) kidney disease, (o) thyroid gland disorder, (p) serious dermatological disorders like psoriasis and eczema, (q) cancer, (r) multiple sclerosis, and (s) Parkinson's disease or epilepsy. Participants were asked whether they had had a specific chronic medical condition in the 12 months prior to the interview. The same procedure is used by the Central Office for Statistics in the Netherlands in their periodic General Health Surveys. To reduce potential reporting bias by patients, we included only conditions for which a GP or specialist was consulted or for which medication was used in the 12 months prior to the interview. The number of chronic medical conditions was used as an index of chronic morbidity.
A three-level index of the severity of the injury was constructed on the basis of the ICPC codes used by the GPs. Hip fracture was considered most severe. The second level consisted of fractures other than hip fracture (fractures of wrist or forearm, ankle or lower leg, and hand or foot). The third level consisted of nonfracture injuries (ankle sprains, knee sprains, or other sprains and dislocations).
Participants and Response
During the inclusion period, GPs registered 287 patients who had sustained injuries to the extremities; patients were counted only once. Of these, 18 did not meet the inclusion criteria (short-version MMSE score < 5 [n = 2], or enrolled in another GLAS cohort [n = 16], 4 had died in the period between registration date and date of contact, and 5 people could not be located (number of eligible patients was 260). Another 59 patients refused to participate, 22 because they felt too ill and 37 for other reasons. We conducted proxy interviews (only at follow-up with a maximum of one for each patient) to determine the functional status of 10 patients who were hospitalized at the time of the assessment or felt too ill otherwise. Valid data were obtained from 201 patients (including the proxies) who participated in the first series of interviews (8 weeks postevent); of these, 186 participated in the second series (5 months postevent) and 181 in the third (12 months postevent). Attrition after the first series was caused by refusals to continue (n = 9), institutionalization (n = 1), and patients' death before the assessment (n = 2); for 3 people attrition was not well documented. Attrition after the second assessment was caused by refusal (n = 1), poor health (n = 1), and patients' death before the assessment (n = 3). Ten patients who participated in three follow-up assessments appeared to have completed the shorter telephone interviews at baseline, which did not include GARS (see the Data Source section). Only those patients with complete data for the dependent variable for all four measurements were included in the analyses (n = 171, response rate is 66% of 260 eligible patients; see above). The included patients sustained fractures of wrist or forearm (n = 44), fractures of ankle or lower leg (n = 14), fractures of hand or foot (n = 12), hip fractures (n = 34), other fractures (n = 32), ankle sprains (n = 14), knee sprains (n = 14), and other sprains and dislocations (n = 7). Of the 116 patients not in the study (including those who died and those who did not meet the inclusion criteria), 12 (10.3%) sustained fractures of wrist or forearm, 5 (4.3%) of ankle or lower leg, 9 (7.8%) of hand or foot bones, 37 (31.9%) of the hip, 27 (23.3%) other fractures, 14 (12.1%) sprains of the knee, 7 (6.0%) sprains of the ankle, and 5 (4.3%) other sprains and dislocations.
Analytical Strategy
First, descriptive baseline statistics were computed for the participants (n = 171) and nonparticipants (n = 116) in the study; differences between both groups were tested with Student's t test and chi-square. Second, mean levels of disability were estimated within the levels of education and social support for each observation period. We conducted paired t tests to test for significance in differences over time for each of the three levels. Finally, changes in GARS scores between baseline on the one hand and at 8 weeks, 5 months, and 12 months postevent on the other hand were computed for each of the three levels of education and social support. We used multivariate analysis of variance to assess possible differences between three levels of education and social support with respect to changes in disability scores from baseline to the three follow-up measurements while adjusting for the influence of the selected covariates (i.e., age, baseline levels of disability, depressive symptoms, cognitive functioning, chronic medical morbidity, and the severity of the injury). For all analyses, p < .05 was considered statistically significant.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). The percentages of low, medium, and high levels of social support in the GLAS baseline with the same cutoff scores as in the present sample were 34%, 34%, and 32%, respectively. There were significant differences between patients who participated in the study (n = 171) and those who did not (n = 116) with respect to age, severity of the injury, and disability at baseline: Nonparticipating patients were older, sustained more hip fractures, and reported higher levels of disability at baseline than participating patients. Other differences were not significant.
|
|
|
|
|
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The results of our study support the outcomes of
Cummings and colleagues 1988, who reported that hip fracture patients with more social support had more complete recovery of their prefracture level of functioning.
Magaziner and colleagues 1990 also found that the amount of patients' phone contacts with social network members 2 months after their hip fracture influenced their 1-year recovery. However,
Mossey and colleagues 1989 did not find any significant associations between social support and recovery. None of these studies actually measured the level of disability as well as predictors before the injury, and they included only hip fracture patients. Our prospective study shows evidence that the role of social support in recovery also holds for patients who sustain other fall-related injuries than hip fracture. We find some evidence that educational level plays a role in functional recovery after fall-related injuries even after adjusting for the influences of several covariates; however, the associations do not reach statistical significance. We find no substantial evidence for the hypothesis of differential vulnerability. The findings of our study point to the importance of social factors—particularly social support—for recovery.
Our prospective study includes patients with fall-related injuries to the extremities that occurred after the GLAS baseline assessment in 1993. The strength of this approach is that we assessed social support, education, disability, and the selected covariates at baseline before the injury occurred. However, this approach also has several limitations. The time interval between the start of the study (baseline in 1993) and the fall-related injuries varies from 1 month to 57 months (M = 23.9; SD = 16.1). Health status and social support may have changed during the interval. However, the correlation coefficients between the length of the time interval and the changes in disability between the baseline and the follow-ups in the patient sample are -.02, -.04, and .07, respectively (ns). In addition, the outcomes of Table 4 hardly change when the time of the interval is included as a covariate in the multivariate models. We therefore assume that the variation in the interval from baseline to injury did not substantially affect the outcomes of the study. Next, 116 patients who had sustained fall-related injuries according to their GP did not participate in the study (response rate 60%). Nonparticipants were older and reported higher levels of disability at baseline (age and disability were significantly related). In addition, the group of nonparticipants had a higher proportion of patients who sustained hip fractures compared with participants. Furthermore, 38% of the eligible source population at baseline did not participate in GLAS. Nonparticipants at baseline were older compared with participants. Computerized health care utilization records were available for 55% of the GLAS source population. Comparing these records for baseline participants and nonparticipants (on group level) suggests relatively high nonresponse in four groups of persons (
Ormel et al. 1997): those with advanced malignant neoplasms, those with significant cardiac surgery, those with a history of suicide attempt, and those who had not consulted their GP in the past 12 months. This suggests elevated nonresponse in the very sick (life-threatening disease, severe depression) and the very healthy. These results may have induced selection bias stemming from the nonresponse at different observation periods. Although attrition in aging studies may complicate the interpretation of descriptive outcomes, the impact of attrition seems to be less of a problem when associations between variables are analyzed. However, the selective response with respect to age, baseline disability, and the severity of the injury may have affected our outcomes. A last remark refers to the relatively low number of patients with a high level of education (n = 13). We conducted an additional analysis with just two levels of education: no (elementary) education, elementary education, or vocational training (n = 120) versus high school, undergraduate degree, or graduate degree (n = 51). The previous results were supported by this analysis: We found no significant effects of education on recovery.
In conclusion, we find that patients recovering from fall-related injuries who report high levels of social supports before their injury have a better recovery after 5 and 12 months; those patients with low and medium levels of social support recover less well. One could argue that patients with more social contacts might have a greater motivation to remain physically active and maintain their mobility and therefore put more effort into rehabilitation. However, this underlying mechanism is not clear. Although patients with high levels of education most closely approach their pre-event level of disability as well, differences do not reach statistical significance. Additional studies to elaborate these mechanisms in more detail and to obtain more knowledge about how social support and educational level of patients could be incorporated in rehabilitation programs are needed. In any case, encouragement and special attention by health professionals to low- and medium-educated older patients to maintain social support after fall-related injuries may be beneficial to these patients' rehabilitation.
|
Acknowledgments |
|---|
Received for publication November 17, 2000. Accepted for publication January 25, 2001.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. G. Taylor Timing, Accumulation, and the Black/White Disability Gap in Later Life: A Test of Weathering Research on Aging, March 1, 2008; 30(2): 226 - 250. [Abstract] [PDF]
|
|
|
|
 |
|
 E. J. MacKenzie and M. J. Bosse Factors Influencing Outcome Following Limb-Threatening Lower Limb Trauma: Lessons Learned From the Lower Extremity Assessment Project (LEAP) J. Am. Acad. Ortho. Surg., September 1, 2006; 14(10): S205 - S210. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. I. J. M. Kempen, J. Ormel, W. Scaf-Klomp, E. van Sonderen, A. V. Ranchor, and R. Sanderman The Role of Perceived Control in the Process of Older Peoples' Recovery of Physical Functions After Fall-Related Injuries: A Prospective Study J. Gerontol. B. Psychol. Sci. Soc. Sci., January 1, 2003; 58(1): P35 - 41. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Jang, W. E. Haley, B. J. Small, and J. A. Mortimer The Role of Mastery and Social Resources in the Associations Between Disability and Depression in Later Life Gerontologist, December 1, 2002; 42(6): 807 - 813. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
