| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
a Population Studies Center, Institute for Social Research, University of Michigan, Ann Arbor
b Departments of Epidemiology and Internal Medicine, University of Iowa, Ames
c Institute for Social Research, Institute of Gerontology, Department of Psychology, University of Michigan, Ann Arbor
Stephanie J. Fonda, New England Research Institutes, 9 Galen Street, Watertown, MA 02472 E-mail: sfonda{at}neri.edu.
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. The project used data from 3 waves of the Asset and Health Dynamics Among the Oldest Old (AHEAD) study. Depressive symptoms were assessed with an abbreviated Center for Epidemiologic Studies-Depression scale. Using 2 models, the project examined how driving cessation and reduction that occurred between Waves 1 and 2 contributed to increases in depressive symptoms between Waves 2 and 3. The first model included the entire sample (N = 5,239), and the second model focused on drivers only (n = 3,543). A third model added interaction terms to the analysis to consider whether respondents who stopped driving but had a spouse who drove were less at risk of worsening depressive symptoms.
Results. Respondents who stopped driving had greater risk of worsening depressive symptoms. Drivers who restricted their driving distances before the study began also had greater risk of worsening depressive symptoms, but seemingly less so than the respondents who stopped driving altogether. For respondents who stopped driving, having a spouse available to drive them did not mitigate the risk of worsening symptoms.
Discussion. Changes in driving patterns can be deleterious for older people's depressive symptoms. Initiatives for assisting older people should focus on strategies that help them retain driving skills, that prepare them for the possible transition from driver to ex-driver, and that ensure that they have access to mental health therapies if driving changes are imminent.
THE percentage of Americans aged 65 and older is projected to increase from 12.8% in 2000 to 20.7% by 2040 (U.S. Bureau of the Census 1993
). Concomitantly, the proportion of elderly drivers and ex-drivers will increase as well. This anticipated trend has led to tremendous interest in such topics as the safety of older drivers (Cox, Quillian, Thorndike, Kovatchev, and Hanna 1998; Odenheimer et al. 1994; Owsley et al. 1998; U.S. Department of Transportation 1997), programs providing older people with alternative means of transport (Freund and McKnight 1997), reasons for why older people stop driving (Campbell, Bush, and Hale 1993; Forrest, Bunker, Songer, Coben, and Cauley 1997; Kington, Reuben, Rogowski, and Lillard 1994; Marottoli et al. 1993), and the consequences of driving cessation for older adults' quality of life (Carp 1971; Cutler 1972, Cutler 1975; Marottoli et al. 1997). In this study we addressed whether and how changes in older adults' driving patterns affect a specific aspect of their quality of life—their depressive symptoms. Despite growing awareness that depression and depressive symptoms are a widespread public health quandary for older adults (Lebowitz et al. 1997), there are few recent studies of this issue.
There are a variety of reasons to expect that changes in driving patterns are associated with worsening depressive symptoms among older adults. One is that driving cessation and reduction lower people's mobility (Burkhardt, Berger, and McGavock 1996). Loss of mobility, in turn, probably reduces older people's access to resources known to protect their quality of life, such as paid and volunteer work, religious communities, friends, and health care services (Cutler 1974; Marottoli et al. 2000). Another explanation is based on evidence that loss of the ability to perform everyday self- and home-care tasks increases depressive symptoms and the risk of clinical depression (Bruce, Seeman, Merrill, and Blazer 1994; Gurland, Wilder, and Berkman 1988; Phifer and Murrell 1986). Given that driving is an everyday task for many people, it is likely that not being able to do it relates to depressive symptoms similarly. A third explanation is that changes in driving and depressive symptoms are linked through a disease that causes both. In this case, changes in driving would not directly and independently cause worsening depressive symptoms. A final explanation is found in the literature on life events and depression. This literature has suggested that certain lifestyle and identity-transforming events, such as the death of a loved one, lead to acute as well as persistent depressive symptoms in the older people who experience them (Glass, Kasl, and Berkman 1997; Hays et al. 1998; Turvey, Carney, Arndt, Wallace, and Herzog 1999). Inasmuch as the transition from driver to ex-driver changes one's lifestyle and identity (Rothe 1994), it too could be a risk factor for worsening depressive symptoms.
According to the model of selective optimization with compensation, however, changes in driving patterns need not be associated with worsening depressive symptoms. This model, developed originally by P. B. Baltes and Baltes 1990, defines selection as the process whereby older people redefine their goals in response to changes in personal and environmental resources and restrict their normal activities. To illustrate selection, M. M. Baltes and Carstensen 1996 described older people who, finding driving increasingly difficult, choose to avoid highways, drive only during daylight, and/or drive shorter distances. The model defines compensation as the process whereby older people achieve the same goals (i.e., they are not restricted as they are with selection), but through new means. Highly mobile people who turn to others for transportation when they can no longer drive, for example, are compensating, according to this model. Optimization refers to the process by which older people learn new ways of doing things and thereby regain functioning, health, well-being, and so forth. Older people who equip their automobiles with assistive devices or participate in driver reeducation programs are optimizing, according to the selective optimization with compensation framework. These three strategies theoretically enable older people to maintain their previous levels of satisfaction.
Existing research specifically inspecting the link between driving and quality of life has tended to support the expectation that driving limitations have negative consequences for older people's well-being. Cutler 1972, for example, found that older people with access to personal transportation reported greater life satisfaction than their counterparts without such means. Additionally, in a study of depression among stroke survivors, Legh-Smith, Wade, and Hewer 1986 found that depression was more common among survivors who no longer drove than it was among survivors who did. This relationship persisted for stroke survivors who had alternate means of transportation. In one of the few recent studies of changes in driving patterns and affective well-being, Marottoli and associates 1997 demonstrated that driving cessation predicted higher levels of depressive symptoms at least 1–3 years after it occurred. This study is especially suggestive because it was longitudinal, the effects of driving cessation exceeded those for most other factors included in the analysis, and it used a large community sample of older adults. None of the existing studies on consequences of changes in driving behavior, however, examined how the modification of driving without stopping it (selection) impinged on older adults' quality of life. Also, with the exception of the study by Legh-Smith and colleagues, previous research on changes in driving patterns has not looked at how the presence of other drivers (i.e., options for compensation) relate to older people's well-being.
In our study we contribute to existing research by assessing the effects of two types of changes in driving patterns on older Americans' depressive symptoms: driving cessation and reduction. We posed the following three hypotheses: Older people who stop driving will be at greater risk of worsening depressive symptoms than people who continue to drive. By contrast, people who continue to drive but reduce their driving distances will not be at greater risk of worsening depressive symptoms. Lastly, having easy access to automobile transport through a spouse who drives will lessen the risk of worsening depressive symptoms that results when older adults stop driving. To test these hypotheses, we utilized a national panel study of older adults. We examined how stopping or restricting driving in one time period contributed to changes in depressive symptoms at a later, chronologically distinct time period. We included in our analyses a comprehensive set of controls suggested by previous research as being important to driving patterns and/or mental health. This approach provided a more accurate representation of how changes in driving patterns affect depressive symptoms than has been previously available.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). The AHEAD study was a companion to the Health and Retirement Study (HRS) until 1998, when the studies merged and combined data collection efforts. Its target population was noninstitutionalized Americans born in 1923 or earlier (aged 70 years and older) and their spouses (also aged 70 years and older). Spouses aged less than 70 years were also interviewed if they were married to an age-eligible person, but they were not considered to be part of the target population. The first interview was conducted in 1993, a second in 1995, and a third in 1998. We used Version 2.10 of Wave 1 (1993), Version 1.0 of Wave 2 (1995), and Version 1 (early release) of Wave 3 (1998). The early release version of Wave 3 is preliminary in that it has not been cleaned. The variables we used from Wave 3 (interview status and depressive symptoms), however, are not subject to change in the cleaning process. Our study included only age-eligible AHEAD self-respondents. At Wave 1, there were 6,580 age-eligible White, Black, and Hispanic self-respondents who provided information about their driving status. As of Wave 2, 577 had died, 369 refused a second interview, and 395 required proxies. We excluded these respondents from the analytic sample because we did not know if their driving patterns changed after Wave 1 and/or we lacked information about their depressive symptoms at Wave 2. This yielded an analytic sample of 5,239 respondents. Only 4,102 of the 5,239 respondents, however, provided information on their depressive symptoms at both Waves 2 and 3; the others had died between Waves 2 and 3 (n = 591), required proxies (n = 324), or refused a third interview (n = 222).
Because the rate at which AHEAD respondents dropped from the study in Waves 1–3 was nontrivial, we took two steps to account for and limit bias arising from sample attrition. First, we explored how the Wave 1 driving behavior and depressive symptoms of decedents and nonrespondents at Wave 2 differed from the respondents still in the sample at Wave 2. We found that 46.9%, 54.0%, and 35.1% of the respondents who died between Waves 1 and 2, required proxies at Wave 2, or refused a second interview, respectively, did not drive at Wave 1. By comparison, 26.1% of the analytic sample did not drive at Wave 1. We also found that the average number of depressive symptoms at Wave 1 for respondents who died between Waves 1 and 2, required proxies at Wave 2, or refused a second interview was 2.51, 2.18, and 1.88. The average number of depressive symptoms at Wave 1 for respondents in the analytic sample was 1.55. From these results we inferred that omitting the AHEAD respondents who died between Waves 1 and 2, required proxies at Wave 2, or refused a second interview would produce conservative estimates of the driving-depression relation. Second, because we had information about whether their driving behavior changed between Waves 1 and 2, we retained the respondents in the sample who had died, required proxies, or refused to participate in the study at Wave 3. We modeled these outcomes as competing risks to worsening depressive symptoms at Wave 3 so as to minimize systematic error resulting from further sample attrition.
Measures
Dependent variable.
Depressive symptoms were measured in the AHEAD study with an 8-item version of the original, 20-item Center for Epidemiologic Studies-Depression (CES-D) scale. The abbreviated CES-D scale was designed to reduce the length of survey interviews for elderly respondents while measuring the continuum of symptoms captured by the original scale (Kohout, Berkman, Evans, and Cornoni-Huntley 1993). It has comparable reliability, validity, and dimensionality to the full version (Turvey, Wallace, and Herzog 1999).
For these analyses we used a summary measure of respondents' answers to the abbreviated CES-D scale at Waves 2 and 3. All items were coded so as to indicate the presence of symptoms, ranging from 0 to 8; higher scores indicated worse affective functioning. We assessed change in depressive symptoms by subtracting the Wave 2 CES-D scores from the Wave 3 CES-D scores, with the change score theoretically ranging from -8 to +8. The average change in scores between Waves 2 and 3 was .49 (weighted mean), with 68.26% of respondents who provided information on their depressive symptoms at both waves changing by less than two symptoms (SD = ±1.76). As a result, we defined increases or decreases of at least 2 points in the CES-D score as true change. Also, after confirming that driving changes were unrelated to improvement in depressive symptoms (i.e., decrease in the CES-D score), in the final analyses we combined respondents who improved with those who did not change or changed by one symptom.
Competing risks to worsening depressive symptoms between the second and third waves were loss to follow-up and death. Loss to follow-up pertained to respondents who required proxies or refused to complete the third interview. Both were included as outcomes in our four-category dependent variable (also see the Participants section).
Driving patterns.
Each interview contained questions about respondents' driving status and modification of driving distances. Juxtaposition of respondents' answers across interviews yielded the following categories: (a) respondents who could drive at Waves 1 and 2, (b) respondents who never drove or stopped before Wave 1, and (c) respondents who stopped driving between Waves 1 and 2. This categorization is consistent with that used by Marottoli and associates 1997 in their analysis of driving and depressive symptoms among New Haven Established Populations for Epidemiologic Studies of the Elderly respondents. Among respondents who drove at Waves 1 and 2, there were the following groups: (a) respondents who drove unrestricted at Waves 1 and 2, (b) respondents who restricted their driving distances before Wave 1, and (c) respondents who restricted their driving distances between Waves 1 and 2. Few AHEAD respondents started driving between Waves 1 and 2 or increased their driving distances, so these respondents were grouped with those who drove and drove without restrictions at Waves 1 and 2.
Social-demographic controls.
The analyses controlled for several background characteristics: respondents' age, gender, race, educational attainment, and whether they lived in an urban area.
Controls for health and functioning at Wave 1.
The analyses also controlled for respondents' debilitating and life-threatening health conditions, physical limitations, and cognitive functioning at Wave 1. The indicator for debilitating health conditions measured the presence of high blood pressure, arthritis, incontinence, vision problems, and hearing problems (Fonda and Herzog 2001; House, Lepkowski, and Williams 2000). The indicator of life-threatening health conditions measured whether respondents had cancer, stroke, lung disease, cardiac disease, and diabetes (Fonda and Herzog 2001; House et al. 2000; Musick, Herzog, and House 1999). The measure for physical limitations summed the number of strength and mobility tasks (i.e., walk several blocks, climb a flight of stairs, lift/carry over 10 lb, pull/push large objects, pick up a dime) with which respondents had difficulty (Wallace and Herzog 1995). We measured cognitive functioning using a continuous indicator of respondents' total number of correct answers to tests assessing their orientation to the current date (e.g., day, month), ability to count backward from 20, word/name recognition, ability to subtract by 7 starting from 100 (Serial 7s), and ability to recall 10 common nouns immediately and then later in the interview (memory; Herzog and Wallace 1997).
Controls for changes in health and functioning.
The analyses also added measures that might account for the relationship between changes in driving patterns and depressive symptoms. One measure indicated whether respondents reported having a new debilitating health condition at Wave 2; one indicated whether they had a new life-threatening health condition; another indicated if their physical limitations worsened; and the last indicated whether respondents' cognitive scores declined by at least 4 points.
Spouse's driving status.
The study incorporated measures of whether the respondents had a spouse who drove at Wave 2, had a spouse who did not drive, had a spouse who did not say whether they drove because they did not complete the interview, or were unmarried. Other drivers may also have been available to respondents, but the AHEAD data do not provide clear information about this.
Statistical Analysis
The analyses involved estimation of three multinomial logistic regression models. The first compared the relationships of driving, never having driven or stopped before Wave 1, and driving cessation between Waves 1 and 2 to worsening depressive symptoms and its competing risks at Wave 3. The comparison group for the Wave 3 outcomes consisted of respondents whose depressive symptoms did not change or improved at Wave 3. The model adjusted for respondents' social-demographics, baseline health, changes in health that occurred between Waves 1 and 2, and spouse's driving characteristics at Wave 2. We added a measure of respondents' preexisting symptoms to both models to adjust for the fact that respondents with few symptoms had more opportunity to worsen than those with many symptoms. The second model focused on drivers only and contrasted the effects of driving unrestricted, restricting driving distances between Waves 1 and 2, and restricting driving distances before Wave 1. It used the same set of controls as the first model. The third model added multiplicative interaction terms (e.g., Stopped Driving x Married and Spouse Drove, Never Drove or Stopped Before Wave 1 x Married and Spouse Drove) to the first model to test whether respondents who stopped driving but had alternative transportation through a spouse were protected against the negative affective consequences of driving cessation. All of the models used the survey estimation program in STATA 7 to adjust for the unequal probabilities of selection to participate in the AHEAD study and nonresponse.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Several particularities of this study warrant mention before we discuss our findings. First, we had pragmatic and substantive grounds for focusing on increases of two or more depressive symptoms. Pragmatically, the relatively high frequency of one-symptom changes (32.9% of respondents) suggested that one-symptom increases could be due disproportionately to measurement error and confound the results. Moreover, limiting increases to those of three or more depressive symptoms probably would have precluded understanding the subtler effects of driving changes. From a substantive standpoint, previous research has shown that four or more symptoms on the abbreviated CES-D scale in the AHEAD study is roughly equivalent to the 16-plus cutoff on the full CES-D scale used to differentiate depressed people from nondepressed people (Turvey, Wallace, and Herzog 1999). Hence, an increase of two symptoms on the abbreviated scale would indicate at least a subsyndromal depressive episode, which is known to affect daily physical functioning in a manner similar to clinical depression (Rollman and Reynolds 1999; Whooley and Browner 1998; Wulsin, Vaillant, and Wells 1999).
Second, we do not know what precipitated the changes in driving patterns among the AHEAD respondents. Family intervention, advice from a physician, a crash, state intervention, the costs of operating a vehicle, and a decision by the older driver could each lead to changes in driving patterns, but with very different implications for how the drivers interpret these changes and, ultimately, for their affective well-being. For example, events that convey to older drivers that they lack control over their lives could deleteriously affect their self-concept (M. M. Baltes and Skinner 1983). But drivers who select for themselves whether to drive and under what circumstances might retain their feelings of self-efficacy and affective well-being (P. B. Baltes and Baltes 1990; M. M. Baltes and Carstensen 1996). Information about what precipitated changes in driving could be useful for refining the general hypotheses regarding the relationship between changes in driving and depressive symptoms.
Third, aside from information about whether respondents lived in an urban area and had a spouse who drove, we did not have information about the ecological aspects of the respondents' settings. We did not know who had easy access to public transportation or whether other household members could drive respondents. This study, therefore, indicates how changes in driving related to changes in depressive symptoms for AHEAD respondents living in average settings. Examination of how environment mediates the relationship between driving limitations and well-being could be an area of further study.
Consistent with previous research (Carp 1971; Cutler 1972, Cutler 1974; Legh-Smith, et al. 1986; Marottoli et al. 1997) and our hypothesis, in this study we find that respondents who stop driving are more likely to later report worsening depressive symptoms. Driving cessation also predicts death and loss to follow-up, which, although not the focus of this study, is interesting because it suggests that driving cessation could be a sentinel event for these outcomes. The findings lead us to speculate that driving cessation signifies—in ways that are especially tangible—the attainment of old age and its stigma of dependency and/or the constriction of access to necessary and recreational activities. Empirical tests of our speculations may be an area for further study.
Our findings for driving modification, or selection, are mixed with respect to the model of selective optimization with compensation and our hypotheses. According to the model of selective optimization with compensation, selection through an act such as driving modification is a strategy older people use to continue the basic themes in their lives and thereby maintain satisfaction (P. B. Baltes and Baltes 1990). Yet we find that respondents who restricted their driving distances before the study began (i.e., for at least 2 years) are at greater risk of worsening depressive symptoms. This result might mean that respondents who have restricted their driving behavior for a while view driving cessation as an increasing possibility, which is an event that many drivers dread (Burkhardt, et al. 1996; Carp 1971). We also find, however, that respondents who restricted their driving distances relatively recently (i.e., between Waves 1 and 2) are not at greater risk of worsening depressive symptoms and respondents who restricted their driving before the study began do not experience an equally large increase in depressive symptoms as those who stopped driving altogether. These findings support the idea that modification of driving, at least for shorter durations, is one way that older people can achieve their transportation goals and maintain affective well-being.
Lastly, we find that the effects of driving cessation on worsening depressive symptoms are not mitigated by the presence of a spouse who drives. This finding is contrary to what we expected given the model of selective optimization with compensation. Our finding suggests that driving cessation per se—not just access to transportation—might be the more crucial risk factor for many older people and raises questions about how effective transportation programs are likely to be for older people's affective well-being. It is possible such programs will not be especially effective, assuming it is true that older adults prefer to receive assistance from people with whom they have the closest relationships, followed by people with whom they have the next closest relationships, and so on, with professionals being the least desirable sources of assistance (Cantor 1979; Shanas 1979). This remains an open question.
In light of these results, we propose that initiatives to address the driving concerns of older people should promote strategies to prevent driving cessation. Such strategies might include the development of driver self-evaluation workbooks to encourage safe behavior (Eby, Shope, Molnar, Vivoda, and Fordyce 2000), advanced traveler information systems that relay timely information (e.g., navigation instructions) to older drivers while they are driving (Eby and Kostyniuk 1998), and driving simulators that reeducate older drivers and enable them to practice driving in a safe setting (Ward 1996). Even educating older people about how they can effectively modify their driving (e.g., limiting driving distances) is a possible strategy for preventing (or at least delaying) driving cessation, although the protective effects of driving modifications may be short term. If driving cessation remains the only choice for some people, transportation programs for older people should include efforts to prepare them for the life transition from driver to ex-driver. This transition could be couched in terms of new opportunities (e.g., to save money by no longer maintaining a car) rather than as exclusively a loss. Finally, family members and care providers should know that changes in driving patterns may put the older driver at risk of worsening depressive symptoms and take steps to ensure that the older person has access to various mental health therapies.
|
Acknowledgments |
|---|
Received for publication February 16, 2001. Accepted for publication May 2, 2001.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. Caragata Nasvadi and A. Wister Do Restricted Driver's Licenses Lower Crash Risk Among Older Drivers? A Survival Analysis of Insurance Data From British Columbia Gerontologist, August 1, 2009; 49(4): 474 - 484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Turano, B. Munoz, S. E. Hassan, D. D. Duncan, E. W. Gower, K. B. Roche, L. Keay, C. A. Munro, and S. K. West Poor Sense of Direction Is Associated With Constricted Driving Space in Older Drivers J Gerontol B Psychol Sci Soc Sci, May 1, 2009; 64B(3): 348 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. M. F. Owen, D. P. Crabb, E. T. White, A. C. Viswanathan, D. F. Garway-Heath, and R. A. Hitchings Glaucoma and Fitness to Drive: Using Binocular Visual Fields to Predict a Milestone to Blindness Invest. Ophthalmol. Vis. Sci., June 1, 2008; 49(6): 2449 - 2455. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Edwards, L. A. Ross, M. L. Ackerman, B. J. Small, K. K. Ball, S. Bradley, and J. E. Dodson Longitudinal Predictors of Driving Cessation Among Older Adults From the ACTIVE Clinical Trial J. Gerontol. B. Psychol. Sci. Soc. Sci., January 1, 2008; 63(1): P6 - P12. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Leproust, E. Lagarde, S. Suissa, and L R. Salmi Association between road vehicle collisions and recent medical contact in older drivers: a case-crossover study Inj. Prev., December 1, 2007; 13(6): 382 - 387. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. Massof, J. T. Deremeik, W. L. Park, and L. L. Grover Self-Reported Importance and Difficulty of Driving in a Low-Vision Clinic Population Invest. Ophthalmol. Vis. Sci., November 1, 2007; 48(11): 4955 - 4962. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Marottoli, P. H. V. Ness, K. L. B. Araujo, L. P. Iannone, D. Acampora, P. Charpentier, and P. Peduzzi A Randomized Trial of an Education Program to Enhance Older Driver Performance J. Gerontol. A Biol. Sci. Med. Sci., October 1, 2007; 62(10): 1113 - 1119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Dickerson, L. J. Molnar, D. W. Eby, G. Adler, M. Bedard, M. Berg-Weger, S. Classen, D. Foley, A. Horowitz, H. Kerschner, et al. Transportation and Aging: A Research Agenda for Advancing Safe Mobility Gerontologist, October 1, 2007; 47(5): 578 - 590. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. Windsor, K. J. Anstey, P. Butterworth, M. A. Luszcz, and G. R. Andrews The Role of Perceived Control in Explaining Depressive Symptoms Associated With Driving Cessation in a Longitudinal Study Gerontologist, April 1, 2007; 47(2): 215 - 223. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. E. Freeman, S. J. Gange, B. Munoz, and S. K. West Driving Status and Risk of Entry Into Long-Term Care in Older Adults Am J Public Health, July 1, 2006; 96(7): 1254 - 1259. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Ragland, W. A. Satariano, and K. E. MacLeod Driving Cessation and Increased Depressive Symptoms J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2005; 60(3): 399 - 403. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
