Physical and mental health

In a review article one of the present authors (Ames, Reference Ames1993) noted that physical disability was strongly associated with depression, but emphasized the need for further studies to identify and define other risk factors. Medical comorbidities and functional impairment have emerged as important risk factors in a number of subsequent studies (Henderson et al., Reference Henderson, Korten, Jorm, Christensen, Mackinnon and Scott1994; Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005; Segal, Reference Segal2005; Watson et al., Reference Watson2006). Persistent pain also has been identified as a major risk factor, affecting both mood and quality of life (Cuijpers and Lammeren, Reference Cuijpers and Lammeren1999; Asghari et al., Reference Asghari, Ghaderi and Ashory2006). Medications for a number of comorbid conditions (e.g. anti-hypertensive/cardiac drugs; treatment for Parkinson's disease, antineoplastic drugs and certain antibiotics) may also cause depression or produce depressive symptoms (Garavaglia, Reference Garavaglia2004).

The medical conditions listed by Hyer et al. (Reference Hyer, Carpenter, Bishmann and Wu2005) include those whose symptoms may overlap with those of depression, and indicate a high correlation between dementia and depression. This correlation is reported in a number of other studies (Gruber-Baldini et al., Reference Gruber-Baldini, Zimmerman, Boustani, Watson, Williams and Reed2005; Sheehan et al., Reference Sheehan, D'Souza, Nwe and O'Malley2007), though it is difficult to determine cause and effect. A history of previous depression is one of the best predictors of manifest depression in residents of care homes for older people (Payne et al., Reference Payne2002), and better cognitive functioning increases levels of risk of depression when associated with loss of physical function and capacity to live independently (Jones et al., Reference Jones, Marcantonio and Rabinowitz2003; Tsai et al., Reference Tsai, Chung, Wong and Huang2005). Jones et al. (Reference Jones, Marcantonio and Rabinowitz2003) also identified Parkinson's disease and heart disease as risk factors, although medication for these conditions may be implicated too (Garavaglia, Reference Garavaglia2004).

Other studies have noted relationships between depression and loss of perception, as in hearing loss (Eisses et al., Reference Eisses, Kluiter, Jongenelis, Pot, Beekman and Ormel2004) and uncorrected refractive error affecting vision (Owsley et al., Reference Owsley, McGwin, Scilley, Meek, Seker and Dyer2007). A further study identified a relationship between depression and malnutrition (Smoliner et al., Reference Smoliner, Norman, Wagner, Hartig, Lochs and Pirlich2009) though no clear pathway of causality was apparent.

Individual personality, social history, attitudes and coping strategies

Loss, a major factor in admission into residential care homes, is strongly associated with depression. Personal loss encompasses grief over the death of loved ones, loss of home, pets, social support and close friends; loss of function and control over the body; loss of independence and autonomy; and loss of the familiar (Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005; Pot et al., Reference Pot, Deeg, Twisk, Beekman and Zarit2005; Zeiss, Reference Zeiss2005; Choi et al., Reference Choi, Ransom and Wyllie2008). An important factor is the loss of control over one's own destiny (Krach et al., Reference Krach, DeVaney, DeTurk and Zink1996). Personal coping styles and attitudes are important in explaining why some people are more negatively affected than others by loss and the presence of death. Cataldo (Reference Cataldo1994) found “non-hardiness and health-limiting death attitudes” to be reliable predictors of depression, and staff in another study identified social withdrawal (related to despair, loss, grief, loneliness and isolation) as a key indicator of depression (Choi et al., Reference Choi, Wyllie and Ransom2009). Lack of assertiveness is likely to lower defenses against institutional risk factors (Segal, Reference Segal2005). Higher education levels also correlate with greater likelihood of being depressed in long-term care (Eisses et al., Reference Eisses, Kluiter, Jongenelis, Pot, Beekman and Ormel2004), especially when associated with loss of mobility, function and independence.

Residential care home environments

Institutional factors that may contribute to depression include loss of privacy and frustration over shared rooms, noise, institutional furniture and odors, lack of stimulating social programs, lack of close relationships, high turnover rate of staff with little training and many cultural and educational differences, and other frustrations of living in close quarters with strangers under an institutional regime (Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005; Choi et al., Reference Choi, Ransom and Wyllie2008). The medicalized environment of residential home settings and lack of alternative approaches to care have also been identified as depression risk factors (Garavaglia, Reference Garavaglia2004; Zeiss, Reference Zeiss2005), as has the constant presence of death (Cataldo, Reference Cataldo1994).

Length of time in care

The evidence suggests that the influence of length of time in care is different for cognitively impaired residents, and more likely to be a negative for the well-being of those without cognitive impairment (Stout et al., Reference Stout, Wilkin and Jolley1993; Pot et al., Reference Pot, Deeg, Twisk, Beekman and Zarit2005). Many residents admitted with dementia have undiagnosed depression (Sheehan et al., Reference Sheehan, D'Souza, Nwe and O'Malley2007), which is likely to improve if diagnosed and treated on and after admission (Masand, Reference Masand1995; Payne et al., Reference Payne2002). For residents with better cognitive function, the risk of becoming depressed increases with length of stay and an increased sense of hopelessness over health status and lack of autonomy (Jones et al., Reference Jones, Marcantonio and Rabinowitz2003; Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005; Pot et al., Reference Pot, Deeg, Twisk, Beekman and Zarit2005; Tsai et al., Reference Tsai, Chung, Wong and Huang2005).

Screening and detection instruments

The GDS has been widely used in residential care homes and several shorter versions have been developed, the most popular being the 15-item version. The original 30-item version (using a cut-off point of ≥11) has good psychometric properties. Gerety and colleagues (Reference Gerety1994) found moderate to substantial agreement with the DSM-III-R criterion standard for major depression and recommended its use in residential care homes, as both a case finding and severity instrument. Jongenelis and colleagues (Reference Jongenelis2005) found the GDS-30 to have sensitivity of 96.3% for major depression and 85.1% for minor depression compared with psychiatric interview and DSM-IV criteria. The 15-item version is generally preferred over the original as it is quicker to complete (6 minutes versus 12) and has good psychometric properties compared with the original version (McCurren, Reference McCurren2002) and with DSM criteria (Gerety et al., Reference Gerety1994). A five-item version of the GDS is also effective as a screening tool for depression in cognitively intact older adults, with 0.94 sensitivity and 0.81 specificity compared with an evaluation by a geriatrician (Rinaldi et al., Reference Rinaldi2003). As the GDS relies on self-report, there is debate as to whether it is suitable for use with people with mild to moderate dementia. A recent study comparing the GDS-15 with DSM-IV-TR criteria in people with and without dementia found adequate specificity and sensitivity, suggesting that it may be an appropriate instrument for use in these populations (Lach et al., Reference Lach, Chang and Edwards2010).

The CSDD has been extensively evaluated for use in residential care homes. The CSDD covers five areas: mood related signs, behavioral disturbances, physical signs, cyclic functioning and ideational disturbances. It takes approximately 30 minutes to complete: 20 minutes with a carer (or other informant) and 10 minutes with the older person. It is reliable, valid and sensitive, and can differentiate the entire range of severity, including mild to no depression (Alexopoulos et al., Reference Alexopoulos, Abrams, Young and Shamoian1988). Its limitations are that it takes 30 minutes to complete, requires trained personnel, does not correlate well with a psychiatrist's assessment, and is not always accurate in people with advanced dementia (de Bellis and Williams, Reference de Bellis and Williams2008). Despite these limitations, de Bellis and Williams concluded that it is the most comprehensive depression screening tool for older people with and without dementia, making it suitable for use in residential care homes. The main advantage of the tool is its use of multiple sources of information. Koritsas et al. (Reference Koritsas, Davidson, Clarke and O'Connor2006) suggest that routine use of the CSDD in residential care homes could improve doctor, nurse and personal carer communication, as personal care staff can be informants in the assessment of depression. Poor communication between professionals has been identified as a barrier to identification of depression in residential care homes (Brown et al., Reference Brown, Raue, Schulberg and Bruce2006).

Seven items from the mandated MDS from the Resident Assessment Instrument in the USA have been validated for use as a depression screening tool known as the MDS Depression Rating Scale (DRS) (Burrows et al., Reference Burrows, Morris, Simon, Hirdes and Phillips2000). In the original study the DRS compared favorably with the GDS-15 and the DRS has since been found to have adequate sensitivity and specificity and to be able to be reliably administered via self-report to residents with a MMSE score ≥12 (Ruckdeschel et al., Reference Ruckdeschel, Thompson, Datto, Streim and Katz2004). However, other studies have found poor correlation between the DRS and the GDS-15 (McCurren, Reference McCurren2002; Jones et al., Reference Jones2004; Meeks, Reference Meeks2004). This is probably due to the use of different information sources, as the GDS relies on self-report and the DRS is an observer-rated instrument. It may also be that the tools are measuring different elements of depression. One study found that the MDS identified greater depression among people with cognitive impairment and the GDS-15 identified more depression in people with better cognitive functioning (Jones et al., Reference Jones2004). Meeks (Reference Meeks2004) did not recommend the use of the MDS DRS as a screening tool for residential care homes because of its poor correlation and poor psychometric properties compared with the GDS.

A single screening question “Do you feel that your life is empty?” was trialed as a screening tool for depression in residential care homes in London and compared with the CSDD. A “yes” response indicated that the resident was twice as likely to have depression and a “no” response meant that there was a 75% chance that they did not have depression. Half of the study sample (n = 209) scored less than 15 on the Mini-Mental State Examination (MMSE) but there was no difference in sensitivity or specificity for the screening question for different levels of cognitive impairment. Watson et al. (Reference Watson, Zimmerman, Cohen and Dominik2009) trialed five strategies for depression screening in residential aged care. The best strategy was a two-item version of the Patient Health Questionnaire (PHQ-2) with a sensitivity of 0.80 and specificity of 0.75 as it was brief and easy to administer.

Under-detection

Nurses and personal care staff should be well placed to detect depression because of their close involvement with residents, but most are not very good at recognizing the symptoms of depression (Leo et al., Reference Leo, Sherry, DiMartino and Karuza2002; Ayalon et al., Reference Ayalon, Arean and Bornfeld2008). There is a poor relationship between resident and nurse CSDD ratings (Burrows, Reference Burrows1995). Brühl and colleagues (Reference Brühl, Hendrika and Martien2007) found the GDS detected 50% more depression than did nurses. Nurses recognized depression in only 55% of residents diagnosed via a psychiatric interview using DSM-IV criteria and they detected depression in 40% of residents not diagnosed as depressed. In the study by Watson and colleagues (2009) mentioned above, measures completed by care staff (modified version of the CSDD and a one-item screen) failed to detect depression adequately. A UK study conducted by Bagley and colleagues (Reference Bagley2000) found that only 15–27% of 308 newly admitted residents with depression were identified as depressed by staff (GDS compared with staff informant interviews). They concluded that more education was needed, as less than 2% of staff in their study had received in-service training on depression in older people. Ayalon et al. (Reference Ayalon, Arean and Bornfeld2008) found that paraprofessional staff working in residential care homes were more likely to view depression as a normal phenomenon, had less accurate beliefs about signs and symptoms and were less familiar with effectiveness of treatments for depression than nurses, social workers and activity staff. Education should be targeted to meet the needs of this group, as they provide the bulk of the care to residents.

Education

A number of studies have investigated the impact of education on detection and treatment of depression in residential care homes (McCabe et al., Reference McCabe, Russo, Mellor, Davison and George2008). In an American study investigating nursing home physicians’ beliefs about how well they can detect and treat depression, excellent training (versus good, fair, poor or none) and the use of screening tools were associated with better recognition and treatment skills, and practice guideline awareness was associated with greater self-reported treatment competency (Banazak et al., Reference Banazak, Mullan, Gardiner and Rajagopalan1999). A single education session for general practitioners (GPs) on late life depression was associated with improved GP recognition of depression in nursing home residents in an Australian study (Davidson et al., Reference Davidson, Koritsas, O'Connnor and Clarke2006). In another Australian study, staff education was one component of a successful multifactorial intervention for improving detection and treatment of depression in residential care homes (Llewellyn-Jones et al., Reference Llewellyn-Jones, Baikie, Smithers, Cohen, Snowdon and Tennant1999). Dutch nursing home staff randomized to a 4-hour training session on recognition of depression through observation of ADLs (using the Behavior Rating Scale for Psychogeriatric Inpatients) were better able to recognize depression than those in the control group. The ability to identify residents who were not depressed did not differ between groups. The use of video-based training also improves nursing staff's ability to detect mood symptoms in nursing home residents (Wood et al., Reference Wood, Cummings, Schnelle and Stephens2002). Having a mandated screening tool also improves recognition and initiation of treatment for depression in residential care homes (Boyle et al., Reference Boyle, Roychoudhury, Beniak, Cohn, Bayer and Katz2004).

Treatment approaches

The main approaches in treatment of depression anywhere at any age can be divided into biological (e.g. antidepressants and electroconvulsive therapy (ECT)), psychological (e.g. cognitive behavior therapy) and social (e.g. recreational activities). Treatments may be used singly or in combination depending on circumstances. Table 2 displays a summary of representative treatment studies in residential care homes.

Table 2. Representative studies of treatment for depression in residential homes for the elderly

Note: BDI = Beck Depression Inventory, CGI-S = Clinical Global Impression of Severity, CMAI = Cohen-Mansfield Agitation Inventory, CSDD = Cornell Scale for Depression in Dementia, GAS = Global Assessment Scale, GDS = Geriatric Depressive Scale, GDS-S = Geriatric Depressive Scale-Short Edition, HDRS = Hamilton Depression Rating Scale, LSI-Z = Life Satisfaction Index Z, MLDT = Macneill-Lichtenberg Decision Tree, MMSE = Mini-Mental State Examination, PAS = Psycho-geriatric Assessment Scales, PES-AD = Pleasant Events Schedule For Alzheimer's Disease, QLI = Quality of Life Index, RSE = Rosenberg Self-Esteem Survey, SADS = Schedule for Affective Disorders and Schizophrenia

Antidepressants are the most common treatment for depressed residents (Brown et al., Reference Brown, Lapane and Luisi2002; Carlson and Snowden, Reference Carlson and Snowden2007). There is evidence that antidepressants are effective for treating depression, in particular, major depression in care home residents (Rovner and Katz, Reference Rovner and Katz1993; Snowden et al., Reference Snowden, Sato and Roy-Byrne2003). It is suggested that antidepressants should be included in the first-line treatment for residents with major depression (American Geriatrics Society and American Association for Geriatric Psychiatry, 2003). Several different classes of antidepressant are available (Baldwin, Reference Baldwin, Jacoby, Oppenheimer, Dening and Thomas2008) and the choice of medication should be based on the side effect profile, potential interactions with other medications and the resident's comorbid illnesses (Masand, Reference Masand1995; Rosen et al., Reference Rosen, Mulsant and Pollock2000; Kallenbach and Rigler, Reference Kallenbach and Rigler2006). Antidepressants should be started at a low dose and be increased slowly (Snowdon et al., Reference Snowdon, Burgess, Vaughan and Miller1996; Bell and Goss, Reference Bell and Goss2001; American Geriatrics Society and American Association for Geriatric Psychiatry, 2003; Kallenbach and Rigler, Reference Kallenbach and Rigler2006). ECT is usually administered in hospital and is reserved for severe depression, but because it works quickly ECT should be considered as the first-line treatment for residents at high risk of suicide (Kallenbach and Rigler, Reference Kallenbach and Rigler2006).

Numerous psychological approaches are appropriate to treat depression in residential care homes. These include behavior therapy (Meeks and Depp, Reference Meeks and Depp2003; Meeks et al., Reference Meeks, Looney, Haitsma and Teri2008; Reference Meeks, Shah and Ramsey2009), cognitive therapy (Zerhusen et al., Reference Zerhusen, Boyle and Wilson1991; Tsai et al., Reference Tsai, Wong, Tsai and Ku2008), cognitive behavior therapy (CBT) (Hyer et al., Reference Hyer, Yeager, Hilton and Sacks2008), and life review approaches (Chao et al., Reference Chao2006; Plastow, Reference Plastow2006). Social interventions include various types of recreational activities (Snowden et al., Reference Snowden, Sato and Roy-Byrne2003), e.g. therapeutic biking programs (Fitzsimmons, Reference Fitzsimmons2001; Buettner and Fitzsimmons, Reference Buettner and Fitzsimmons2002). Most of the studies cited above reported significant beneficial effects and some review papers have concluded that psychosocial approaches are effective in treating depression, in particular minor depression, in residential care homes (Snowden et al., Reference Snowden, Sato and Roy-Byrne2003; Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005). Consistently, it is suggested that non-pharmacological methods should be considered as first-line treatment for residents with minor depression (American Geriatrics Society and American Association for Geriatric Psychiatry, 2003). It is important to note that many of these approaches were delivered in a group format (Bharucha et al., Reference Bharucha, Dew, Miller, Borson and Reynolds2006). Group therapy is a more practical and cost-effective strategy in the residential home environment (Carlson and Snowden, Reference Carlson and Snowden2007). Individual psychotherapy is not available in most residential care homes (Snowden et al., Reference Snowden, Sato and Roy-Byrne2003) due to a shortage of trained psychotherapists, cost, and, in many countries, a lack of any tradition of psychotherapists working in residential care homes.

There is also some evidence that other approaches, such as staff training (Glaister and Blair, Reference Glaister and Blair2008), changes in health service model (Snowden et al., Reference Snowden, Sato and Roy-Byrne2003) and residential home environment (Bell and Goss, Reference Bell and Goss2001) might also be effective in reducing residents’ depressive symptoms.

It seems likely that combining medical and psychosocial approaches would be the best way to manage depression in residential care homes (McCurren et al., Reference McCurren, Dowe, Rattle and Looney1999; American Geriatrics Society and American Association for Geriatric Psychiatry, 2003; Morris, Reference Morris2008). However, no recent studies have systematically investigated the possible benefits of such a combined approach.

Over the past 12 years there has been increasing interest in multifaceted approaches to the treatment of depressed residents (Llewellyn-Jones et al., Reference Llewellyn-Jones, Baikie, Smithers, Cohen, Snowdon and Tennant1999; Reference Llewellyn-Jones2001; Brooker and Woolley, Reference Brooker and Woolley2007; Brooker et al., Reference Brooker, Woolley and Lee2007; Hyer et al., Reference Hyer, Yeager, Hilton and Sacks2008; Jordan et al., Reference Jordan, Byrne and Bushell2009). For example, the multifaceted shared care intervention used by Llewellyn-Jones et al. (Reference Llewellyn-Jones, Baikie, Smithers, Cohen, Snowdon and Tennant1999) included: (a) multidisciplinary consultation and collaboration to remove barriers to care, (b) training of GPs and carers in detection and management of depression, (c) depression-related health education and activity programs for residents. Similarly the British Enriched Opportunities program (Brooker and Woolley, Reference Brooker and Woolley2007; Brooker et al., Reference Brooker, Woolley and Lee2007) included individualized assessment of residents, case work, an activity and occupation program, staff training, management and leadership. These studies reported significant improvement in residents’ depression and suggest that integrating multiple modalities should be the focus of future research.

As discussed above, it can be difficult to identify depressed residents. It is also challenging to treat them effectively. Some antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), may have limited efficacy in residents with comorbid dementia (Burrows et al., Reference Burrows, Salzman, Satlin, Noble, Pollock and Gersh2002). In addition, some residents may not be able to participate effectively in psychotherapy because of their limited cognitive function (Kallenbach and Rigler, Reference Kallenbach and Rigler2006; Carlson and Snowden, Reference Carlson and Snowden2007).

There are important methodological limitations in many treatment studies. These include small sample sizes (Bharucha et al., Reference Bharucha, Dew, Miller, Borson and Reynolds2006), variable study inclusion criteria (Snowden et al., Reference Snowden, Sato and Roy-Byrne2003; Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005), short treatment duration (Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005), few randomized controlled trials (Carlson and Snowden, Reference Carlson and Snowden2007) and heterogeneous outcome measures (Bharucha et al., Reference Bharucha, Dew, Miller, Borson and Reynolds2006). However, these studies do provide strong evidence that depression is treatable in nursing home residents (Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005).

Issues in treatment of depression

In spite of the evidence that depression is treatable in residential homes, many studies have found that depression is under-treated in this group (Masand, Reference Masand1995; Hyer et al., Reference Hyer, Carpenter, Bishmann and Wu2005). About 50%, or even up to 75% of residents with depression were receiving no treatment at all (Rovner et al., Reference Rovner, German, Brant, Clark, Burton and Folstein1991; George et al., Reference George, Davison, McCabe, Mellor and Moore2007).

Although there is evidence that the initiation of treatment for depression has improved in recent years (Boyle et al., Reference Boyle, Roychoudhury, Beniak, Cohn, Bayer and Katz2004), the treatment is often inadequate or inappropriate. First, there is a reliance on antidepressants and very few residents receive non-pharmacological treatment for depression (George et al., Reference George, Davison, McCabe, Mellor and Moore2007; Choi et al., Reference Choi, Wyllie and Ransom2009). Second, many residents receiving antidepressants are on sub-therapeutic doses (Draper et al., Reference Draper2001; Brown et al., Reference Brown, Lapane and Luisi2002). There is also evidence that antidepressants are rarely reviewed by physicians (O'Connor et al., Reference O'Connor, Griffith and McSweeney2010). Given these problems, it is not surprising that only a minority of residents on antidepressants show improvements in depressive symptoms (Draper et al., Reference Draper2001; Boyle et al., Reference Boyle, Roychoudhury, Beniak, Cohn, Bayer and Katz2004). The findings of inadequate treatment of depression in residential care homes have major implications for the management of depression in these environments. Many researchers have suggested that the management should go beyond first generation problems, such as identification and initial treatment, to include second generation problems, such as strategies for partial- and non-responders, and ongoing management through the maintenance phases (Draper et al., Reference Draper2001; Datto et al., Reference Datto, Oslin, Streim, Scheinthal, DiFilippo and Katz2002; Weintraub et al., Reference Weintraub, Datto, Streim and Katz2002; Boyle et al., Reference Boyle, Roychoudhury, Beniak, Cohn, Bayer and Katz2004; Carlson and Snowden, Reference Carlson and Snowden2007).