Age and Ageing

Age and Ageing Advance Access originally published online on November 21, 2007
Age and Ageing 2008 37(1):51-56; doi:10.1093/ageing/afm162

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 37/1/51    most recent afm162v1 E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Disclaimer Google Scholar Articles by Martin, L. Articles by Fries, B. E. Search for Related Content PubMed PubMed Citation Articles by Martin, L. Articles by Fries, B. E. Social Bookmarking          What's this?

Copyright © The Author 2007. Published by Oxford University Press on behalf of the British Geriatrics Society.

Predictors of a new depression diagnosis among older adults admitted to complex continuing care: implications for the depression rating scale (DRS)

Lynn Martin, Jeff W. Poss, John P. Hirdes, Richard N. Jones, Michael J. Stones and Brant E. Fries

Health Studies and Gerontology, University of Waterloo, Homewood Health Centre, Homewood Research Institute, Waterloo, Ontario, Canada

Address correspondence to: John P. Hirdes. Email: hirdes{at}healthy.uwaterloo.ca

	Abstract

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Background: depression is a major disabling condition among older adults, where it may be under-diagnosed for a number of reasons, including a different presentation for younger people with depression. The Minimum Data Set 2.0 (MDS 2.0) assessment system provides a measurement scale for depression, the Depression Rating Scale (DRS), in addition to other items that may represent depressive phenomenology.

Objective: the ability of the DRS to predict the presence of new depression diagnoses at follow-up, among hospitalised older adults admitted without depression, is examined.

Methods: the study sample consists of all persons aged 65 years or more admitted between 1996 and 2003 to a complex continuing care (CCC) bed in Ontario without a recorded depression diagnosis. The sample was restricted to those who remained in hospital for about 3 months (n = 7,818) in order to obtain follow-up assessment information. Logistic regression was used to explore the relationship between admission characteristics (i.e. DRS scale items, other MDS 2.0 items related to DSM-IV criteria for depression) and receipt of a depression diagnosis on the follow-up assessment.

Results: a new depression diagnosis at follow-up was present in 7.5% of the individuals. The multivariate model predicting depression diagnosis included only the DRS scale, sadness over past roles, and withdrawal from activities.

Conclusions: the DRS score at admission was predictive of receiving a depression diagnosis on a follow-up assessment among older adults admitted to the CCC. Further, the predictive ability of the DRS is only modestly improved by the addition of other items related to DSM-IV criteria.

Keywords: depression, geriatric assessment, interRAI, DSM-IV, elderly

	Background

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Depression has been identified as the leading cause of disability in developed nations [1], and it affects as many as 2 to 5% older adults [2]. However, prevalence estimates differ by residential setting, ranging from 10 to 15% in community settings [3] and over 20% in nursing homes [4, 5]. Under-detection of depression is an important problem resulting from a variety of possible causes at the patient-level (e.g. communication and cognitive impairment), physician-level (e.g. focus on medical conditions, normalisation of depression in later life), and system-level (e.g. availability of mental health professionals in non-psychiatric hospitals).

Depression is even more common among older adults with physical illness [6], with prevalence rates as high as 50% reported in hospital settings [7]. Comorbid medical conditions further complicate the detection of depression in this population [8] due to the misattribution of symptoms to a physical problem rather than a depressive disorder. In addition to the obvious implications for quality of life, depression has been linked to medical burden [9], morbidity [10], mortality [11, 12], length of stay in hospital [13], physician visits, hospitalisation, and nursing home placement [14]. Although early detection and treatment of depression in this population is vital, current approaches to care too often leave depression undiagnosed.

The definition of depression itself poses a problem that contributes to under-detection in older adults. The DSM-IV provides a standard set of criteria upon which a formal diagnosis of major depression is based. Older adults may be less likely to satisfy the formal criteria for DSM-IV major depression because, even with similar levels of underlying depressive syndrome severity, they are less likely to endorse one of the DSM-IV required symptoms—depressed mood [15]. The notion of ‘depression without sadness’ refers to the tendency of older adults to show anhedonic and somatic rather than dysphoric symptoms [16]. The differential presentation of depressive symptomology by older adults and the tendency of care providers to under-detect and under-treat depression in older adults points to the need for systematic screening in long-term care and hospital settings.

	Measurement of depression

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Among the instruments developed to screen for depression, the Hamilton Rating Scale for Depression (HAM-D) is the most widely used [17, 18]. However, it is not routinely used with all older adults in all care settings.

In many jurisdictions the use of the Resident Assessment Instrument or Minimum Data Set (MDS) represents an important opportunity to conduct census level screening on persons in long-term care and hospital settings [19]. Large-scale implementation of the MDS as the standard assessment approach in these settings has occurred in many countries, including the United States, Canada, Iceland, Finland and Italy. Several clinical scales are embedded within the MDS, including a measure of depression (Depression Rating Scale, or the DRS). The developers of this scale found it to be highly correlated to the HAM-D (r = 0.70), where a cut-off score of 3 or more maximised sensitivity (94%) with minimal loss of specificity (72%) in a sample of nursing home residents [20]. They also found sensitivity of 91% and specificity of 69% for the DRS when tested against psychiatrists’ diagnoses of depression. Kohler and colleagues [21] found that the DRS and the Geriatric Depression Scale (GDS) are weakly correlated with each other, because they measure depression in different ways. However, both were associated with the presence of a depression diagnosis and, unlike the GDS, the DRS was unaffected by item non-response. In addition, research on a general adult sample provided evidence of the validity of the DRS in psychiatric hospital settings [22]. Nonetheless, some researchers have argued that further work is required to improve the sensitivity of the DRS for identifying older adults with depression [23].

	Objectives

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
The focus of the current study is on new admissions to complex continuing care (CCC) hospitals where no prior depression diagnosis has been made. The specific objectives are: (i) to examine the ability of the DRS to predict the presence of new depression diagnoses 3 months post admission; and (ii) to determine whether the sensitivity of the DRS in predicting future depression diagnoses could be enhanced with the use of other MDS 2.0 items.

	Methodology

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Participants
In 1996, the Ministry of Health and Long-Term Care in Ontario, Canada mandated the use of the MDS in all beds of freestanding CCC hospitals or CCC units of acute hospitals.

MDS assessments are completed for all patients admitted to all designated CCC beds for 14 days or longer, and the data are submitted on a quarterly basis to the Canadian Institute for Health Information (CIHI). CIHI's Continuing Care Reporting System (CCRS)—www.cihi.ca/ccrs is a national data warehouse supporting the implementation of the MDS in eight provinces/territories. Given that it was the first province to mandate the use of the instrument in Canada, Ontario's CCC population represents the largest sub-group in the CCRS dataset. Reassessment typically occurs every quarterly for as long as the patient remains in hospital, so that patients may have multiple assessments.

This study examines factors associated with the transition to a new diagnosis of depression among CCC patients who did not have a diagnosis of depression at admission. The analyses excluded admissions: (i) under 65 years of age (12,255 cases); (ii) with an existing depression diagnosis (11,048 additional cases); (iii) already taking anti-depressant medication (9,107 additional cases); (iv) with very severe cognitive impairment (6,858 additional cases); (v) with no follow-up assessment, usually due to discharge before 90 days (38,608 additional cases); and (vi) with reassessments less than 14 days or more than 120 days later (755 additional cases). A final sample of 7,818 patients fit all of the study criteria.

Measures

Depression diagnosis
As part of the MDS assessment, assessors record the presence of a known diagnosis of depression (e.g. documented in patient record, communication from other clinicians); however, they do not differentiate between major depression, minor depression or dysthymia. Previous research has shown that facility MDS assessors record the presence of a depression diagnosis with good reliability ( = 0.65) [24], a level of precision that compares favourably with estimates for other tools used in psychiatric epidemiology [25].

Depression rating scale (DRS)
The DRS is an observer-rated scale that assesses depressive symptoms based on the presence of mood disturbance indicators available in the MDS 2.0. Ignoring the context or assumed cause, each indicator is coded from 0 to 2, based on its observed frequency (not in the last 30 days, up to 5 days per week, and on 6 or 7 days per week, respectively).

Burrows and colleagues [20] conducted a series of analyses to identify which MDS items were most powerfully correlated with both the Hamilton Depression Rating Scale (HADS) and the Cornell Scale for Depression in Dementia. Factor analysis was conducted using those indicators, and five distinct concepts emerged (disturbed mood, anxiety, fear, loss of meaning and affect). Each of the factors was then modelled individually to find the ideal combination of items to predict HADS and Cornell scale scores. The resulting DRS is a seven-item summated scale where scores range between 0 and 14. A cut-off score of 3 best identifies persons experiencing at least mild depression [20]. Acceptable levels of internal consistency have been reported in samples of nursing home residents (Cronbach's alpha = 0.69–0.85) [20, 21] and inpatient mental health patients (Cronbach's alpha = 0.77) [22].

DSM-IV related items
The MDS 2.0 does not have items to address all DSM-IV related criteria for depression. However, several items are available to measure anhedonia, change in weight and appetite, sleep problems, psychomotor activity and concentration.

	Results

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Sample characteristics
Table 1 shows selected characteristics of the study sample at admission. The average age of patients was about 81 years, with most of them being admitted from an acute care hospital, the slight majority was female, and about half were widowed. Common health conditions included hypertension (28.4%), CVA (27.4%), dementia (26.6%), diabetes (22.3%), arthritis (21.6%) and cancer (20.2%).

View this table: [in this window] [in a new window]   Table 1. Patient characteristics

 
The reliability of the DRS was consistent with previous reports, with a Cronbach's alpha value of 0.74. In this sub-group of persons with no existing diagnosis of depression, 47.4% had a DRS score of 1 or more and 20.8% reached the conventional threshold of 3 or more.

Table 2 shows the baseline prevalence for items in the DRS, the range of DRS scores, as well as for other items related to DSM-IV criteria for depression. Though fewer patients had experienced weight gain (2.8%), made self-deprecating remarks (5.8%) or negative statements (7.7%), or showed signs of unrealistic fears (9.5%), the remaining depressive symptoms were exhibited by more than 10% of patients. Many had exhibited impairment in decision-making (74.4%), been easily distracted (35.2%) or restless (32.0%), exhibited sad, pained or worried facial expressions (32.0%), and expressed sadness over past roles (30.1%).

View this table: [in this window] [in a new window]   Table 2. Prevalence and individual ORs predicting new depression diagnosis at follow-up

 
At follow-up, a new diagnosis depression was recorded for 7.5% of the study sample (n = 584). Table 2 shows that, controlling for age and sex, all seven DRS items were significantly associated with increased odds of having a new depression diagnosis at follow-up (OR = 1.25–1.60) for each point increase in the two-point response set. For example, compared with patients who had no repetitive health complaints or concerns at baseline, the odds of a new depression diagnosis at follow-up was 1.60 and 2.56 (i.e. 1.60²) times greater for those manifesting with these complaints on a lesser than daily basis or a daily basis in the last week, respectively. For each single point increment in its fourteen point range, the full DRS was associated with a 1.14 increase in the odds of having a new depression diagnosis at follow-up, (e.g. those with baseline scores of 3 and 6 on the DRS have ORs of 1.48 and 2.19, respectively). The c statistic for the model containing the DRS algorithm was 0.60.

In order to evaluate their predictive ability above and beyond that of the DRS, analyses were also done for additional MDS 2.0 items related to DSM-IV diagnostic criteria and the transition to a new depression diagnosis after adjusting with age, sex and DRS. Only sadness over past roles (OR = 1.70), reduced social interaction (OR = 1.18), poor appetite (OR = 1.29) and withdrawal from activities (OR = 1.25) were significant at the 0.05 level.

A full model was developed that considered all significant covariates in the bivariate regression models (Table 3). In this case, the DRS collapsed on four groups with values of: (i) zero indicating the presence of no symptoms as the reference group; (ii) 1–2 representing a group manifesting some mood indicators but not at the threshold level advocated by the scale developers [22]; (iii) 3–5 reaching the conventional DRS cut-point for depression; and (iv) six or more, in which preliminary research in mental health settings has been found as a useful cut-point for more severe depression. All three DRS subgroups had significantly higher odds of a subsequent new diagnosis of depression compared with the asymptomatic reference group. The patients with a DRS of 1–2 have historically not been considered as requiring attention for depression; however, these results indicate that they are significantly more likely to be diagnosed with depression at follow-up than the reference group. It was also notable that there was no additional gain in predictive power when the higher DRS threshold of 6+ was used as a cut-point. While controlling for the DRS, patients who expressed sadness over lost roles or who withdrew from activities at baseline had associated ORs of 1.41 and 1.24, respectively. The c statistic for the full model was 0.62, slightly higher than what was reported for the DRS alone.

View this table: [in this window] [in a new window]   Table 3. Multivariate model predicting a new depression diagnosis at follow-up

 

	Discussion

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Both the DRS scale and its individual items were significant predictors of the presence of a new depression diagnosis at follow-up among older adults in CCC hospitals. However, two additional MDS 2.0 items that are related to DSM-IV criteria (i.e. sadness over past roles and anhedonia) were significant predictors of that transition after adjusting the DRS. The contribution of these items was not unexpected [26].

The present findings have some important implications for clinical practice and future research. First, the results confirm that there is a substantial subpopulation of older adults who enter CCC with undetected depression. Second, the DRS is a useful tool during admission time which can assist in identifying persons who, at follow-up, would be considered to have an active diagnosis present. For these individuals, the care plan should include specific actions to clarify the nature of and respond to any underlying mood problems affecting the individual. That said, not all persons who had a new diagnosis present at follow-up were detected by the DRS, and there is evidence that a lower threshold score should be used for preliminary screening purposes. While this would improve the sensitivity of the DRS, it would also reduce its specificity. Third, at least two additional MDS items appear to have value in identifying persons with previously undetected depression—sadness over lost roles and withdrawal from activities. These items should be explored further as potential candidates for use in an expanded version of the DRS.

Although the reported ORs of 2.00 or more for values above the conventional cut-off for the DRS are not insubstantial, it is worth noting that the results of this research are probably attenuated by continued under-diagnosis at follow-up. Had the study design be based on prediction of gold-standard detailed diagnostic evaluations aimed specifically at seeking out depression, the associations reported here would likely have been stronger. Rather, the present research models a normal clinical practice over time, and normal clinical practice has not been particularly effective in identifying depression.

This study has some notable limitations. While a variety of DSM-IV symptoms were available within the MDS instrument, some were not (i.e. hypersomnia, psychomotor retardation, fatigue, loss of energy, indecisiveness, suicidal ideation, recurrent thoughts of death or suicide attempt). The availability of some or all of these measures might alter the performance of the DRS and additional DSM-related covariates in the prediction of depression diagnosis. However, some of these criteria, for example suicide attempt, are very rare in institutionalised/hospitalised seniors.

The study design also creates certain inherent limitations. First, the expected follow-up assessment interval of 3 months limits the intervention period in which suspected diagnoses can be identified by nurses or physicians, and may result in under-diagnosis of depression in the study population. Second, the sequence of symptom presentation, intervention, medication and diagnosis is unclear over the 3-months assessment interval. For example, a patient may exhibit symptoms in the 14-day admission period, receive effective treatment, and subsequently have no requirement for further mental health evaluation. As a result the person would be identified as having neither symptoms nor diagnosis at follow-up. On the other hand, a patient exhibiting no symptoms in the 14-day admission period, but developing them subsequent to the MDS assessment, may have a diagnosis at follow-up. Because of the study's design, the latter case cannot contribute to the predictive model. Given enough follow-up assessments, it might be possible to identify whether there is an ‘admission effect’ or whether symptoms persist throughout longer stays. However, in this study population, episodes yielding three or more reassessments are rare [27].

It would also be useful to replicate this research in other care settings and with other populations. For example, the province of Ontario has also mandated compatible assessment instruments that include the DRS for all long-stay home care clients and for all adults in inpatient psychiatry [28]. It would also be useful to employ longitudinal data with multiple waves of follow-up over multiple years to examine the incidence, sequencing and etiology of physical and mental illness and their relationship to diagnostic decision-making. Such research would provide further insights to the applicability of the DRS in different clinical environments and with different populations.

	Conclusions

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
The DRS is a useful screener for potentially undetected depression that can be used to inform care planning in any long-term care or hospital setting that uses the MDS routinely. The present results suggest that persons with a DRS of 1–2 also warrant further review for potential depression. Two items related to anhedonia improved the predictive ability of the DRS, suggesting that there would be value in updating the DRS.

	Key points

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 

• Depression is an important quality of life problem that may be under-diagnosed among the frail elderly.
  
• The DRS can be obtained from the MDS 2.0, a comprehensive assessment instrument that is used widely in nursing homes and continuing care settings.
  
• The DRS is predictive of future depression diagnoses among CCC patients who did not have a depression diagnosis at admission.
  

	Conflicts of interest

 
No financial competing interests. JPH, BEF and TR are fellows of interRAI, a non-profit group dedicated to improving the health care for persons who are elderly, frail or disabled. interRAI owns the copyright to the MDS for nursing homes and many other care settings. For more information please visit http://www.interrai.org.

	Funding

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Financial support for this work is drawn in part from Canada's Primary Health Care Transition Fund and is gratefully acknowledged. JPH's participation was supported by an Investigator Award from the Canadian Institutes for Health Research.

	Ethics clearance

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 
Access to these dataset was provided through a data-sharing agreement between the Canadian Institute of Health Information (CIHI) and interRAI. Ethics approval for secondary analyses of these anonymised data was provided by the University of Waterloo, Office of Research.

	References

Top Abstract Background Measurement of depression Objectives Methodology Results Discussion Conclusions Key points Funding Ethics clearance References

 

1. Murray CJ, Lopez AD. Global mortality, disability, and the contribution of risk factors: Global burden of disease study. Lancet (1997) 349:1436–42.[CrossRef][Web of Science][Medline]
2. Mottram P, Wilson K, Strobl J. Antidepressants for depressed elderly. Cochrane Database Syst Rev (2006) Art. No.: CD003491, DOI: 10.1002/14651858.CD003491.pub2.
3. Alexopoulos GS, Buckwalter K, Olin J, et al. Comorbidity of late life depression: an opportunity for research on mechanisms and treatment. Biol Psychiatry (2002) 52:543–58.[CrossRef][Web of Science][Medline]
4. Alexopoulos GS, Vrontu C, Kakuma T, et al. Disability in geriatric depression. Am J Psychiatry (1996) 153:877–85.[Abstract/Free Full Text]
5. Duff S. Comorbid Depression and Anxiety in Canadian Nursing Homes. (2003) Waterloo, ON: University of Waterloo.
6. Murphy E. Social origins of depression in old age. Br J Psychiatry (1982) 141:135–42.[Abstract/Free Full Text]
7. Koenig HG. Differences in psychosocial and health correlates of major and minor depression in medically ill older adults. J Am Geriatr Soc (1997) 45:1487–95.[Web of Science][Medline]
8. Hooyman NR, Kiyak HA. Social Gerontology: A Multidisciplinary Perspective. (1999) Needham Heights, MA: Allyn and Bacon.
9. Luber MP, Meyers BS, Williams-Russo PG, et al. Depression and service utilization in elderly primary care patients. Am J Geriatr Psychiatry (2001) 9:169–76.[CrossRef][Web of Science][Medline]
10. Meeks S, Murrell SA, Mehl RC. Longitudinal relationships between depressive symptoms and health in normal older and middle-aged adults. Psychol Aging (2000) 15:100–9.[CrossRef][Web of Science][Medline]
11. Bula CJ, Wietlisbach V, Burnand B, et al. Depressive symptoms as a predictor of 6-month outcomes and service utilization in elderly medical inpatients. Arch Intern Med (2001) 161:2609–15.[Abstract/Free Full Text]
12. Covinsky KE, Kahana E, Chin MH, et al. Depressive symptoms and 3-year mortality in older hospitalized medical patients. Ann Intern Med (1999) 130:563–9.[Abstract/Free Full Text]
13. Verbosky LA, Franco KN, Zrull JP. The relationship between depression and length of stay in the general hospital patient. J Clin Psychiatry (1993) 54:177–81.[Web of Science][Medline]
14. Koenig HG, Kuchibhatla M. Use of health services by hospitalized medically ill depressed elderly patients. Am J Psychiatry (1998) 155:871–7.[Abstract/Free Full Text]
15. Gallo JJ, Rabins PV. Depression without sadness: alternative presentations of depression in late life. Am Family Physician (1999) 60:820–6.[Web of Science][Medline]
16. Gallo JJ, Rabins PV, Lyketsos CG, et al. Depression without sadness: functional outcomes of nondysphoric depression in later life. J Am Geriatr Soc (1997) 45:570–8.[Web of Science][Medline]
17. Kapfhammer HP. Somatic symptoms in depression. Dialogues Clin Neurosci (2006) 8:227–39.[Medline]
18. Sajatovic M, Ramirez LF. Rating Scales in Mental Health. (2003) Cleaveland, OH: Lexi-Comp Inc. 2nd edition.
19. Hawes C, Morris JN, Phillips CD, et al. Development of the nursing home resident assessment instrument in the USA. Age Ageing (1997) 26:19–25.[Abstract/Free Full Text]
20. Burrows AB, Morris JN, Simon SE, et al. Development of a minimum data set-based depression rating scale for use in nursing homes. Age Ageing (2000) 29:165–72.[Abstract/Free Full Text]
21. Koehler M, Rabinowitz T, Hirdes J, et al. Measuring depression in nursing home residents with the MDS and GDS: an observational psychometric study. BioMed Central Geriatr (2005) 5:1.
22. Hirdes JP, Smith TF, Rabinowitz T, et al. The Resident Assessment Instrument-Mental Health (RAI-MH): Inter-rater reliability and convergent validity. J Behav Health Serv Res (2002) 13:419–32.
23. Anderson RL, Buckwalter KC, Buchanan RJ, et al. Validity and reliability of the Minimum Data Set Depression Rating Scale (MDSDRS) for older adults in nursing homes. Age Ageing (2003) 32:435–8.[Abstract/Free Full Text]
24. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics (1977) 33:159–74.[CrossRef][Web of Science][Medline]
25. Eaton WW, Neufeld K, Chen LS, et al. A comparison of self-report and clinical diagnostic interviews for depression diagnosis interview schedule and schedules for clinical assessment in neuropsychiatry in the Baltimore epidemiologic catchment area follow-up. Arch Gen Psychiatry (2000) 57:217–22.[Abstract/Free Full Text]
26. Reker GT. Personal meaning, optimism, and choice: existential predictors of depression in community and institutional elderly. Gerontologist (1997) 37:709–16.[Abstract]
27. Hospital Report Research Collaborative. Hospital Report, 2001 Complex Continuing Care (2001) http://www.hospitalreport.ca/pdf/FINALCCCReportDec202001.pdf.
28. Hirdes JP. Addressing the health needs of frail elderly people: Ontario's experience with an integrated health information system. Age Ageing (2006) 35:329–31.[Free Full Text]

Received 7 November 2006; accepted in revised form 19 June 2007.

CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 37/1/51    most recent afm162v1 E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Disclaimer Google Scholar Articles by Martin, L. Articles by Fries, B. E. Search for Related Content PubMed PubMed Citation Articles by Martin, L. Articles by Fries, B. E. Social Bookmarking          What's this?

Online ISSN 1468-2834 - Print ISSN 0002-0729

Copyright © 2009 British Geriatrics Society

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics