Abstract
Depressive symptoms are expected to become a leading cause of the global burden of disease, second only to cardiovascular disease, by 2020 (Chapman & Perry, 2008). Late-life depressive symptoms are an important public health problem. In older adults, depressive symptoms impede independence in activities of daily living (ADLs), hinder quality of life, and can affect physical functioning (Blazer, 2003; Jang, Bergman, Schonfeld, & Molinari, 2006; Li & Conwell, 2009, Penninx, Leveille, Ferrucci, van Eijk, & Guralnik, 1999; Watson et al., 2006). Consequently, interventions that address affect are an important component of occupational therapy practice (American Occupational Therapy Association, 2008). Lack of social support has been associated with a higher prevalence of depressive symptoms, as has an older adult’s place of residence (Glass, De Leon, Bassuk, & Berkman, 2006; Isaac, Stewart, Artero, Ancelin, & Richie, 2009). Both factors are critical components of a client’s context relevant to occupational therapy practice.
Senior residences are considered midway along a continuum of living arrangements that range from independent living at home to dependence on others for basic self-care in a nursing home. Services provided can include assistance with ADLs and instrumental activities of daily living (IADLs), provision of meals, medical services, and recreational programs (Janeski & Pruchnicki, 2006; Jang et al., 2006). As many as 27% of older adults residing in senior residences exhibit depressive symptoms (Jang et al., 2006; Watson, Garrett, Sloane, Gruber-Baldini, & Zimmerman, 2003; Watson et al., 2006). Psychosocial interventions used successfully with older adults to ameliorate depressive symptoms include cognitive–behavioral therapy, problem-solving therapy, and life review and reminiscence therapy (Adamek & Slater, 2008). Life review is an occupation-based psychosocial intervention that has mounting empirical support (Bohlmeijer, Roemer, Cuijpers, & Smit, 2007) and warrants the consideration of occupational therapists practicing in geriatrics. The intervention can be conducted in an individual or group format and can be performed orally or in writing. In this article, we explore the evidence for both oral and written life review.
Occupational engagement in life review through oral dialogue can be described as recalling and telling stories about one’s life from childhood to the present day. The process can take place using an individual approach, that is, between the reviewer and a therapeutic listener, or through group interaction with a trained group leader facilitating the discussion (Arean et al., 1993; Haight, Michel, & Hendrix, 1998). In most cases, life review interventions include verbal prompts or discussion topics to help guide the client through the process (Haight et al., 1998; Mastel-Smith et al., 2006). Engagement in the occupation-based intervention of life review through dialogue has therapeutic benefits that include decreasing depressive symptoms (Arean et al., 1993; Davis, 2004; Haight, 1992).
Life review through dialogue using an individual treatment approach has been conducted in a number of treatment settings and with a variety of different populations (Davis, 2004; Haight et al., 1998; Mastel-Smith et al., 2006). For example, Mastel-Smith et al. (2006) investigated the effectiveness of life review offered by home care workers with a homebound older adult population aged 65–92, whereas Davis (2004) focused on the inpatient stroke population aged 45–87. In Haight et al.’s (1998) study, newly relocated nursing home residents were the focus of the intervention.
In addition to its use with a variety of patient populations, life review has been carried out using different intervention protocols, including Haight’s Life Review and Experiencing Form (Davis, 2004; Haight et al., 1998) and interventions developed by Mastel-Smith et al. (2006) for use in a particular study. Although a variety of populations have been studied and different interventions have been used, positive outcomes have been found; namely, a statistically significant decrease in depressive symptoms (Davis, 2004; Haight et al., 1998; Mastel-Smith et al., 2006).
Life review through oral dialogue has also been conducted as a group intervention, which involves the same systematic review of life events but in a group format. In some cases, interventions include group discussion and other techniques, such as rounds and role-playing (Chiang Lu, Chu, Chang, & Chou, 2008). Group interventions offer the added therapeutic benefits of social support from peers.
As was the case for the individual approach, several randomized controlled studies (RCTs) with older adults in both community and institutional settings have shown group life review through dialogue to be an effective intervention (Arean et al., 1993; Chiang et al., 2008; Hanaoka & Okamura, 2004). Moreover, the intervention has been demonstrated to be effective when used with different cultural groups, including Taiwanese and Japanese participants residing in their respective home countries (Chiang et al., 2008; Hanaoka & Okamura, 2004). Although an oral-based life review has been demonstrated to be effective in decreasing depressive symptoms, the act of writing about life events may offer enhanced therapeutic benefit.
Therapeutic writing, a form of expressive therapy that uses the act of writing and processing the written word, can take several different forms. Emotional disclosure includes writing about emotional or traumatic life events without additional prompts, whereas guided autobiography, or life review through writing, includes written or verbal prompts to encourage writing about one’s life in a systematic, chronological way irrespective of emotional content. Several RCTs have demonstrated support for the therapeutic benefits of emotional disclosure through writing in both healthy adults and adults with chronic illness. Outcome measures in these studies have focused primarily on biomedical measures (Pennebaker, Kiecolt-Glaser, & Glaser, 1988; Petrie, Fontanilla, Thomas, Booth, & Pennebaker, 2004; Smyth, Stone, Hurewitz, & Kaell, 1999; Wright & Chung, 2001).
Evidence has also been found for the therapeutic benefits of engagement in life review through writing for older adults regardless of the emotional content of the writings. The research in this area has focused on psychosocial well-being as opposed to biomedical outcome measures, and depressive symptoms are often the primary outcome measure (Birren & Cochran, 2001; Elford et al., 2005; Mastel-Smith, McFarlane, Sierpina, Malecha, & Haile, 2007).
Life review through writing may be more beneficial than life review using dialogue alone (Sherman, 1995). For example, Sherman (1995) conducted a comparative study of life review through oral dialogue and life review through oral dialogue plus a written component with 25 older adults (ages 65–79) attending a community center. The results of the study revealed no mean differences between the groups with regard to affect, but Sherman noted a statistically significant increase in a measure of late-life adjustment for the oral plus written life review group.
Subsequent studies with community-dwelling older adults provided additional support for the therapeutic benefits of life review through writing (Mastel-Smith et al., 2007; Richeson & Thorson, 2002). Richeson and Thorson (2002) used a quasi-experimental design to examine the effects of life-review-through-writing classes compared with liberal arts classes offered through the Elderhostel program. The results revealed nonsignificant increases in life satisfaction for both groups. However, those in the life-review-through-writing classes showed a statistically significant greater improvement in negative affect than those in the liberal arts classes.
In an RCT, Mastel-Smith et al. (2007) examined the effectiveness of a group life-review-through-writing intervention, the Share Your Life Story workshop (Sierpina, 2002), on depressive symptoms in community-dwelling adults >60 yr old. Consistent with the results of the Richeson and Thorson (2002) study, a positive effect on participants’ mood was found: The participants in the treatment group reported fewer depressive symptoms than participants in the control group. Limitations of the Mastel-Smith et al. study included the homogeneity of the participants, that is, the lack of racial diversity (94% White) and high educational level of the group as a whole (100% with high school or more).
To date, the literature has focused on life review through oral dialogue, which has mounting empirical support for its effectiveness in decreasing depressive symptoms (Ando, Morita, & Akechi, 2010; Bohlmeijer et al., 2007; Davis, 2004; Mastel-Smith et al., 2006). Few studies, however, have examined the effectiveness of protocols of life review that include the added therapeutic benefits of writing (Mastel-Smith et al., 2007; Richeson & Thorson, 2002; Sherman, 1995). Life review through writing, as opposed to life review through oral dialogue, may be a more beneficial approach (Sherman, 1995). Moreover, some researchers have called for a refinement and validation of intervention protocols, and a study of interventions in a diverse setting has also been suggested to generalize the results (Bohlmeijer et al., 2007; Mastel-Smith et al., 2007).
Therefore, we evaluated the effectiveness of life review through writing with respect to decreasing depressive symptoms in older adults ≥65 yr old residing in senior residences. On the basis of information from facility administrators, we anticipated that the senior residences chosen for the study would provide a more ethnically and educationally diverse group of study participants with a greater level of functional impairment than participants in previous studies in this area of inquiry. The research question under investigation in this study was, “What is the effect of life review through writing on depressive symptoms in older adults residing in senior residences?”
Method
Research Design
The research study was an RCT conducted in four senior residences. The institutional review board (IRB) application was submitted and approved by both New York University and the one intervention site that had its own IRB. Informed consent was obtained from all interested and eligible participants.
Participants
Participants were recruited by Tracy Chippendale from four senior residences in New York City using IRB-approved flyers and announcements. The four research sites offered similar services, including restaurant-style meals, linen service, housekeeping, daily onsite recreation programs, and an emergency call system; three of the four sites offered scheduled transportation to local area shops. Chippendale invited older adults ≥65 yr old to participate. Additional inclusion criteria were the ability speak and write English and a negative screen for probable dementia on the Mini-Cog cognitive screening tool (Carolan Doerflinger, 2007). People with probable dementia were excluded to establish cognitive capacity for consent and to ensure that participants would be able to recall life events to participate fully in the intervention. No inclusion criteria were used with regard to a minimum level of depressive symptoms because we felt that participants could benefit from the intervention to improve their mood despite not meeting the criteria for mild depression.
Eligible participants were randomly assigned to the treatment or wait-list control group using a table of random numbers (zeros and ones) generated by the Microsoft Excel computer program. Within each residence, participants were assigned using randomization with forced equal sample size because of the anticipated small sample size at each site. The participants in each group were in close proximity to each other; however, data on diffusion were collected at posttest.
Intervention
Chippendale, who was also the workshop leader, conducted the intervention using the Share Your Life Story workshop protocol, which took place on a weekly basis for 8 wk (Sierpina, 2002). The 90-min sessions were conducted at each of the four senior residences. The intervention included presentation of information on writing techniques by the group leader, timed writing exercises, and encouragement to write stories about their life outside of workshop time. During each session, participants read their written work aloud and received positive feedback on their writing from other participants and the group leader. Specific writing prompts that related to a specific period of one’s life were given each week. For example, the first session involved writing prompts that relate to ages 8, 9, or 10, and the subsequent session focused on ages 18, 19, or 20. Participants were instructed not to share their experiences of being in the group with residents outside the group.
At the time of posttesting, the control group was asked two questions: (1) “Have you discussed the workshop with any of those currently participating?” and (2) “Have you been telling others or writing stories about your life?” Responses to these questions revealed that contamination had not taken place. Also noteworthy is that no contact was made with the wait-list participants during the first 8-wk workshop. The wait-list control group participated in the same 8-wk writing workshop after the posttests were completed. No additional data were collected after the second 8-wk workshop.
Measures
Cognitive Screening.
We used the Mini-Cog to rule out older adults with probable dementia to establish eligibility for the study. The Mini-Cog has demonstrated strong psychometric properties. Compared with the Mini-Mental State Examination (MMSE; Folstein, Folstein, & McHugh, 1975) and the Cognitive Abilities Screening Instrument (Teng et al., 1994), the Mini-Cog had the highest sensitivity (99%) and correctly classified the greatest percentage of participants (96%). Unlike the MMSE, the diagnostic value of the Mini-Cog is not influenced by education or language (Borson, Scanlan, Brush, Vitaliano, & Dokmak, 2000).
Depressive Symptoms.
Depressive symptoms were measured using the Geriatric Depression Scale (GDS; 30-item scale; Lelito, Palumbo, & Hanley, 2001). The tool was initially validated in adults ≥55 yr old recruited from housing projects or senior centers or hospitalized for depression in one U.S city. According to Yesavage et al. (1982–1983), the GDS internal consistency measures have a correlation of .56 with the total score, an interitem correlation of .36, and an α coefficient of .94. Its test–retest reliability is .85. Correlations between the GDS and two other valid measures of depression, the Zung Self-Rating Depression Scale (r = .84) and the Hamilton Rating Scale for Depression (r = .83), indicate strong convergent validity (Yesavage et al., 1982–1983). The GDS is also known for its high degree of reliability and validity across different ethnicities and cultural groups (Mui, Kang, Chen, & Domanski, 2003; Yesavage et al., 1982–1983).
Demographics and Key Covariates.
The baseline and posttest questionnaires ascertained demographic information including age, education level, gender, and ethnicity as well as possible confounders, such as self-rated health, current treatment of depression, independence in ADLs and IADLs, levels of leisure participation, and social support. Three questions on the emotional aspect of social support were generated from existing tools, including the Duke Social Support Index and the Social Support Appraisals Scale (Koenig et al., 1993; Vaux et al., 1986). The questions, rated on a Likert scale, were “How often do you feel lonely?” “How often do you wish you had more people to spend time with?” and “I feel valued and important.” We selected the first two questions because we felt that the group could enhance the amount of perceived social support. We selected the third question because the opportunity to share and receive feedback on important life experiences could affect the sense of feeling valued by others. We did not include a standardized measure of social support out of concern for respondent burden.
Data Collection
Data were collected at baseline and then within a week after completion of the first 8-wk workshop. All data collection questionnaires and assessments were administered orally by Chippendale; note that she is an occupational therapist with extensive experience using standardized geriatric assessment tools in a clinical setting.
Analysis
Meta-analysis of life review and reminiscence therapy revealed a large effect size of 0.84 (Adamek & Slater, 2008). Moreover, when reminiscence was differentiated from life review, an effect size of 1.04 was reported for life review compared with 0.40 for reminiscence (Bohlmeijer et al., 2007). For a more conservative estimate, we used .84 for power analysis and for the calculation of an appropriate sample size. Baseline data were examined using χ2 tests of independence and independent-samples t tests to ensure that no initial differences occurred between groups. The hypothesis was tested using a repeated-measures analysis of variance (RMANOVA). We used a descriptive cross-tabs analysis to compare change in GDS category between the treatment and control groups. Data were analyzed using SPSS Version 17 (SPSS Inc., Chicago) at a .05 significance level.
Results
Approximately one-third of the people who were approached agreed to participate. Those who declined stated that they either could not commit to the 8-wk program or had no interest in writing. Two prospective participants were excluded on the basis of their Mini-Cog scores. The initial sample size of eligible participants was 47, recruited from four different senior residences. Only 2 participants were lost to attrition, resulting in a final sample of 45 with 22 in the control group (6 participants from Site A, 4 from Site B, 6 from Site C, and 6 from Site D) and 23 in the treatment group (5 participants from Site A, 6 from Site B, 6 from Site C, and 6 from Site D). Demographic and other baseline characteristics are presented in Table 1.
Participant Characteristics for the Total Sample (N = 45)
Note. GDS = Geriatric Depression Scale 30-question version; M = mean; SD = standard deviation.
Workshop Attendance
Of the 23 participants in the treatment group, 7 (30.4%) participated in all eight workshop sessions, 6 (26.1%) participated in seven sessions; 2 (8.7%), in six sessions; 5 (21.7%), in five sessions; 2 (8.7%), in four sessions; and 1 (4.3%), in only three sessions (percentages do not total 100% due to rounding). With the exception of the 2 participants lost to attrition, all participants were included in the analyses despite the number of sessions attended to ensure that the sickest and most frail participants were not excluded from the analysis, given that doctors’ appointments and illness accounted for most missed sessions.
Improvement in Depressive Symptoms
Initial analyses using χ2 tests of independence for categorical variables and t tests for continuous variables revealed no statistically significant differences between the control and treatment groups at pretest with regard to gender, ethnicity, education, self-rated health, medications taken for depression, previous participation in life review through writing, number of ADLs or IADLs requiring assistance, and self-rated social support on the basis of the three questions derived from existing measures. Independent-samples t tests suggested no difference between the treatment and control groups with regard to number of leisure interests, social support, or GDS scores. We found a statistically significant difference for age, with the treatment group being older (mean [M] = 87.2, standard deviation [SD] = 6.3) than the control group (M = 80.8, SD = 7.5), t(1, 43) = −3.106, p = .003. However, age was not significantly correlated with depressive symptoms (r = .03, p > .05).
Descriptive statistics showed that the mean change in GDS score for the treatment group was 2.70 (SD = 4.09), whereas for the control group, the mean was 0.32 (SD = 2.41). The results of the 2 × 2 factorial design RMANOVA indicated a statistically significant main effect of time, F(1, 43) = 8.86, p = .005, and a statistically significant Time × Group interaction, F(1, 43) = 5.1, p = .029 (Table 2). We noted a significant change in GDS score between pretest and posttest, influenced by group and treatment (Cohen’s d = 0.70, which is considered a large effect).
Repeated-Measures Analysis of Variance Using GDS Scores as Repeated Measure and Group as Fixed Factor
Note. GDS = Geriatric Depression Scale, 30-question version.
Exploratory Analysis of Geriatric Depression Scale Results
Although the general trend was toward a higher percentage of participants with a decrease in depressive symptoms for those who attended more workshop sessions, the relationship was not entirely linear. In other words, the relationship between the number of sessions attended and the percentage of participants who had a decrease in depressive symptoms was not consistent. Cross-tabs evaluation to compare the treatment and control groups with regard to the number of participants who presented in each GDS category at pretest and posttest demonstrated that only the treatment group showed any change from one category of depressive symptoms to another (Table 3).
Cross-Tabs Comparing GDS Categories at Pretest and Posttest
Note. GDS = Geriatric Depression Scale, 30-question version.
Discussion
In this RCT, we examined the effects on depressive symptoms of the occupation-based intervention of life review through writing. The results of hypothesis testing revealed a statistically significant improvement in depressive symptoms for the treatment group participants compared with the control group participants. The cross-tabs exploratory analysis supports the study’s clinical significance because the results indicate that only participants in the treatment group had a change in depression category according to the GDS criteria.
The informal feedback from participants provided additional support for the psychosocial benefits of the intervention. Several participants reported that they did not participate in any other group leisure activities at their residence except this writing workshop. Noteworthy is that a subgroup of the participants was highly educated, including a retired medical doctor and university professors. The writing workshop provided an intellectually stimulating activity of interest to this group of people. Multiple requests to continue the group were made after its completion, and several people reported that the workshop was good for the mind and soul. Participants reported looking forward to the group every week. Several participants reported sharing their written work with family and friends, some of whom requested copies of their work.
The results of the study are consistent with the literature on depressive symptoms and life review. In this study, 31.1% of study participants expressed depressive symptoms, which is slightly higher than the expected 24%–27% prevalence rate for depressive symptoms in the general senior residence population (Jang et al., 2006; Watson et al., 2006). However, the self-selected participants in this study may have specifically joined the study and workshop because of a history of depressive symptoms.
The results of hypothesis testing are consistent with the findings of Mastel-Smith et al. (2006) and Richeson and Thorson (2002). These researchers, who examined the use of life review through writing rather than other forms of life review, reported significantly improved mood for the participants who were in the treatment group compared with those who were in the wait-list control group or who participated in other education groups.
Limitations
One limitation of the study is that the group leader collected the posttest data from each participant, an arrangement that could have resulted in respondent bias in favor of the workshop’s success. However, the gap between pretest and posttest was >2 mo, so participants were unlikely to have remembered their previous responses in detail. Although Chippendale attempted to recruit the needed number of participants at two or three sites, in the end, four sites were necessary, which could have resulted in a clustering effect at each site. However, the fact that the same group leader ran groups at all four sites contributed to consistency across groups. Although the results provide support for the benefits of the intervention, note that posttests were administered within a week of workshop completion; therefore, the long-term effects of the intervention were not addressed in this study and should be addressed in future research.
The observed effects in the treatment group could be the result of experimenter attention. However, experimenter attention versus life review was tested in a previous study, and the life review group demonstrated a statistically significant improvement in depressive symptoms compared with the group that received experimenter attention alone (Davis, 2004).
Future Research
Suggestions for future research include recruitment of participants with a minimum level of depressive symptoms (e.g., ≥10 on the GDS), which would enable a clearer picture of the clinical significance of the results. According to some studies, personal meaning and spiritual well-being may help to explain how the intervention works or benefits participants (Ando et al., 2010; Westerhof, Bohlmeijer, Beljouw, & Pot, 2010). Therefore, inclusion of this variable in future studies is warranted. Finally, because evidence for the therapeutic benefits of life review is mounting, a comparative study with another intervention is warranted to help determine the most effective nonpharmacological interventions for depressive symptoms in older adults.
Implications for Occupational Therapy Practice
Depressive symptoms not only hinder quality of life but have an impact on functional status and increase the use of health care services among older adults (Blazer, 2003). Although occupational therapy services for older adults tend to focus predominately on the physical effects of aging, addressing psychosocial needs is critical to improving quality of life and sustaining independence in daily activities. In fact, the integration of physical and psychosocial attention has been suggested to reduce disablement and even premature death among people with chronic health conditions (Friedman, Furst, & Williams, 2010). Moreover, a holistic approach to care is consistent with the tenets of the profession of occupational therapy.
Occupational engagement in life review through writing is a valuable intervention that can be included in geriatric occupational therapy practice. At this writing, group interventions in the subacute setting that meet certain criteria are reimbursable through Medicare. Older adults who reside in senior residences and who have depressive symptoms could benefit from the occupational therapy psychosocial intervention of life review. In addition to the positive effect on depressive symptoms, the life review workshop also appears to meet an unmet need for leisure participation. Several study participants claimed that among the scheduled activities, the writing workshop was the only one that interested them.
Occupational therapists have the unique skill set to adapt activities, thereby enabling participation in occupations in spite of intervening functional limitations. Many of the study participants required physical modifications to participate. Adaptations included bold-lined paper, high-contrast pens, and glare reduction for people with visual impairments; lumbar supports for low back pain; seating arrangements to facilitate lip reading for participants with hearing impairment; and pens with adapted grips for participants with limited joint range of motion. Occupational therapists, with their training in adaptation, can facilitate participation in writing groups by developing alternate postures and writing tools to enable handwriting.
In summary:
Addressing psychosocial needs is critical to sustaining independence in daily living skills.
Life review through writing workshops is an evidence-based intervention to address depressive symptoms in older adults.
Life review through writing is a valuable intervention that can be included in geriatric occupational therapy practice.
Conclusion
A life-review writing workshop (Sierpina, 2002) conducted among older adults residing in senior residences resulted in a statistically significant improvement in depressive symptoms for treatment group participants compared with control group participants. Descriptive cross-tabs analysis suggests that the results are also clinically significant. Life-review writing workshops are an effective, easy-to-use intervention to address depressive symptoms in older adults that can be implemented by occupational therapists in clinical practice settings.
Footnotes
Acknowledgments
We thank the American Occupational Therapy Foundation for their financial support for this study through the dissertation research grant program. Tracy Chippendale acknowledges and thanks the members of her dissertation committee at New York University for their assistance with this research project: Jane Bear-Lehman (chair), Colleen Gillespie, and Elizabeth Capezuti. She especially thanks the participants, whose life stories have captivated and inspired her, provided her with history lessons, and ultimately enriched her life.