Family Support During Hospitalization Buffers Depressive Symptoms Among Independent Older Adults

ABSTRACT

Objectives

Hospitalization is a stressful event that may lead to deterioration in older adults’ mental health. Drawing on the stress-buffering hypothesis, we examined whether family support during hospitalization would moderate the relations between level of independence and in-hospital depressive symptoms.

Method

This research was a secondary analysis of a cohort study conducted with a sample of 370 hospitalized older adults. Acutely ill older adults reported their level of independence at time of hospitalization and their level of depressive symptoms three days into the hospital stay. Family support was estimated by a daily report of hours family members stayed with the hospitalized older adult.

Results

Independent older adults whose family members stayed longer hours in the hospital had fewer depressive symptoms than independent older adults with shorter family visits. Relations between depressive symptoms and family support were not apparent for dependent older adults, even though their family members stayed more hours.

Conclusions

This study partially supports the stress-buffering hypothesis, in that social support ameliorated depressive symptoms among hospitalized independent older adults.

Clinical implications

Assessing depressive symptoms and functional ability and creating an environment conducive to family support for older adults may be beneficial to hospitalized older adults’ mental health.

KEYWORDS:

• Basic activities of daily living
• Depressive symptoms
• Family support
• Hospitalization
• Instrumental activities of daily living
• Older adults

Introduction

Depressive symptoms are common among hospitalized older adults, with a prevalence ranging from 16% to 60% among older inpatients (Ciro et al., 2012; Pederson et al., 2016; Reichardt et al., 2019; Shastri et al., 2019; Zisberg, 2017). Depressive symptoms during hospitalization are related to longer hospital stays, lack of response to treatment, falls, and decreased functional capacity (Laura et al., 2022; Reichardt et al., 2019). Depressive symptoms have far-reaching effects, including deterioration in health status during and after hospitalization (Covinsky et al., 1997) and increased risk of re-hospitalization, institutionalization, and death (Beeler et al., 2020; Pederson et al., 2016).

The appearance or intensification of depressive symptoms during hospitalization has been associated with both internal and external factors (Rosa et al., 2018; Vink et al., 2008). Sociodemographic characteristics such as older age, female sex, absence of spouse, and low level of education have been related to depressive symptoms (Auslander, 2012; Choi et al., 2019; Nascimento et al., 2017; Rosa et al., 2018; Van Orden et al., 2015; Zou et al., 2018). Other factors are the health condition and the functional ability of the older adult, including levels of comorbidity, the medical condition for which he or she was hospitalized, and the existence of frailty syndrome (Marconcin et al., 2022; Tavares et al., 2018).

Level of independence and depressive symptoms

Both functional status and cognitive impairment were found to be associated with older adults’ depression and even suicide (Wang & Blazer, 2015). In cohorts of hospitalized older adults, decline in basic and instrumental activities of daily living (BADL and IADL, respectively) was significantly associated with depressive symptoms, after adjustment for confounding variables (Reichardt et al., 2020). Even one year after hospitalization, level of depression (adjusted for age, sex, and depressive symptoms at baseline) was significantly predicted by cognitive decline and functional impairment (Helvik et al., 2013).

Family support during hospitalization

An important resource theorized to protect against depressive symptoms in stressful situations is family support. The stress-buffering hypothesis (S. Cohen & Wills, 1985) posits social support is a factor that mitigates the detrimental effect of stressors on health condition through physiological and psychological mechanisms (S. Cohen & Hoberman, 1983). Studies testing the buffering hypothesis have found social support creates an effective shield that helps people deal with stressful situations (Chukwuorji et al., 2017), thus contributing to positive outcomes in the contexts of mental and physical health (Harasemiw et al., 2019; Li & Tung, 2020; Xiao et al., 2019; An et al., 2019). A study examining the buffering hypothesis in the context of depressive symptoms in older adults found those who perceived more social support reported fewer depressive symptoms (Santini et al., 2015). Family support can buffer risk factors of depression, such as disability (Chan et al., 2011) or negative attitudes to aging (Liu et al., 2020).

Family support can also be an important buffer during hospitalization (Mackie et al., 2018), and a lack of such support has been associated with depressive symptoms among hospitalized older adults (Kong et al., 2020). The engagement of a familiar and supportive person may help the inpatient meet the challenges of ongoing hospital care, while maintaining his or her dignity and self-esteem (Morrow & Nicholson, 2016). Family support contributes to comfort, reduces loneliness, and provides a source of encouragement and confidence (Gur-Yaish et al., 2013). As such, it can function as a protective factor against depressive symptoms in the hospital and has been linked to a reduction in the incidence of depressive symptoms three months after hospitalization (Ciro et al., 2012). Family support during hospitalization is common in Israel: studies show that 96% of older adults in Israel are accompanied by a family member during a hospital stay (Gur-Yaish et al., 2021), often for a substantial amount of time (Ambrosi et al., 2017), ranging between 5.6 and 8 hours a day (Auslander, 2012; Gur-Yaish et al., 2013).

Although the above-mentioned studies (Ciro et al., 2012; Gur-Yaish et al., 2013, 2021) established a link between family support and depressive symptoms, findings from Gur-Yaish et al. (2013) suggest level of independence at time of admission to the hospital might also play a significant role. In this particular study, independent older adults reported higher levels of depressive symptoms when the family gave them instrumental support. In a study of non-hospitalized older adults, researchers found an association between social support and lower depression among those with no or low levels of functional impairment but did not find an association among those with severe impairment (Van Orden et al., 2015).

To the best of our knowledge, no study has investigated a possible interactive effect between the amount of time family members spend in the hospital and level of independence on in-hospital depressive symptoms. To explore this issue, based on the literature, we formulated the following hypotheses:

Hypothesis 1:

The amount of time family members spend in the hospital with the hospitalized older adult will moderate the relations between cognitive status and in-hospital depressive symptoms.

Hypothesis 2:

The amount of time family members spend in the hospital with the hospitalized older adult will moderate the relations between level of independence in BADL and in-hospital depressive symptoms.

Hypothesis 3:

The amount of time family members spend in the hospital with the hospitalized older adult will moderate the relations between level of independence in IADL and in-hospital depressive symptoms.

In the analysis, we accounted for potential confounders that may affect in-hospital depressive symptoms, including level of depressive symptoms at admission, severity of acute illness, chronic comorbid conditions, length of hospital stay, and background characteristics.

Method

Study design and setting

The study was a secondary analysis of data from the WALK-FOR cohort study (Y. Cohen et al., 2019). WALK-FOR was designed to assess the effect of an interventional program promoting in-hospital mobility of older patients. The WALK-FOR study recruited patients aged 65 and older admitted into two internal medicine units at an academic medical center in northern Israel between October 1, 2015, and September 1, 2016. Participants were recruited before and after the introduction of changes in clinical practice encouraging patients to be more mobile during their hospital stay. Only cognitively competent patients (scoring 5 or above on the Pfeiffer Short Portable Mental Status Questionnaire (SPMSQ) (Pfeiffer, 1975) with preserved ability to walk two weeks before hospitalization participated in the study.

The study was approved by the University of Haifa ethics board (Approval no. 139/16) and by HaEmek Medical Center’s Helsinki committee (Approval no. 0100–15-EMC). All participants provided informed consent, and participation was voluntary. The study was registered in the Australian New Zealand Clinical Trials Registry (registration record ACTRN12616001274460).

Participants

Four hundred and one older patients who met the inclusion criteria of the WALK-FOR study and agreed to participate were considered for the present analysis (see full description of recruitment process in previous publication Y. Cohen et al., 2019. Of these, 14 were excluded because their hospital stays were short (less than 24 h), and 17 were excluded because of missing data on psychological symptoms, leaving a final sample of 370 participants. The functional and cognitive status, age, education level, number of children, level of depression on entry, and co-morbidity index of excluded participants were comparable to those of participants retained in the final sample.

Instruments and measures

Independent variables

Level of Independence

BADL and IADL were assessed within the first 24 hours of hospital admission (1) using the Modified Barthel Index for Activities of Daily Living (Mahoney & Barthel, 1965) for basic activities such as eating, dressing, and hygienic activities (BADL), and (2) using Lawton’s Instrumental Activities of Daily Living (Lawton & Brody, 1969) for activities such as shopping, transportation, and cooking (IADL). Higher scores in both instruments indicate higher levels of independence. Cognitive status was measured using Pfeiffer’s Short Portable Mental Status Questionnaire (SPMSQ) (Pfeiffer, 1975). Total scores range from 0 to 10, with higher scores indicating better cognitive status.

Outcome measure

Level of depressive symptoms

The level of depressive symptoms was assessed using the seven-item depression subscale from the Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983). HADS has a Likert-type response scale, ranging from 0 (not at all) to 3 (all the time). Level of depressive symptoms was assessed twice: the first time within the first 24 hours of hospital admission and the second time after three days. The second assessment was used as the outcome measure (in-hospital depression), accounting for first time (baseline) measurement. Reliability of the depressive symptom subscale was good: α = 0.85for the baseline level, and α = 0.86for the in-hospital level.

Moderator variable

Family support during hospitalization

In-hospital family support was assessed through in-person daily interviews (up to three follow-up interviews during the hospital stay). In these interviews, participants were asked about the number of hours a family member stayed with them, ranging from 0 to 12 hours. An average score was calculated for up to three days of hospitalization. Family relations with visitors were self-reported by patients. When participants had more than one visitor, we asked about the primary caregiver among all visitors.

Control variables

Chronic morbidity

To measure chronic morbidity, we used the Charlson Comorbidity Index (CCI) (Charlson et al., 1987). The CCI is a method of categorizing comorbidities of patients based on the International Classification of Diseases (ICD) diagnosis codes retrieved from medical records Each of 19 comorbidity categories has an associated weight (from 1 to 6), based on the adjusted risk of mortality, and the sum of all the weights results in a single comorbidity score for a patient. A score of zero indicates that no comorbidities were found. The higher the score, the more likely the predicted outcome will result in mortality.

Severity of illness

Acute illness severity was assessed using the National Early Warning Score (NEWS) (Williams, 2022), a well-validated, track-and-trigger, early warning score system used to identify and respond to patients at risk of deteriorating. It is based on a simple scoring system in which a score is given to physiological measurements already being monitored when patients are in healthcare settings. A summary score ranges from 0 (low risk) to 20 (high risk).

Length of stay in hospital

Data on hospitalization duration were retrieved from medical records.

Background characteristics

Age and sex were retrieved from medical records; ethnicity was self-reported and then categorized dichotomously as “Jew” or “non-Jew.”

Statistical analyses

We performed descriptive analysis of study variables, including standard tests of regression assumptions (Osborne & Waters, 2002; test results indicated that all assumptions were met.

To test the hypothesis that the amount of time family members spend in the hospital will moderate the relations between level of independence (i.e., cognitive status and independence in BADL and IADL) and in-hospital depressive symptoms, in the first stage, we tested the direct effect of level of independence at admission on in-hospital depressive symptoms. We performed a series of multivariable linear regressions, including level of independence (separately, cognitive status and independence in BADL and IADL), baseline level of depressive symptoms, severity of illness, comorbidities, length of hospitalization, background characteristics (age, sex, ethnicity), and participation in the WALK-FOR intervention.

To determine whether family support moderated the association between depressive symptoms and functional status during hospitalization, we used moderator analysis. A moderator is a variable that affects the direction and/or strength of the relations between an independent or predictor variable and a dependent or criterion variable (Baron & Kenny, 1986). We used Hayes’s (2013) SPSS PROCESS routine for moderation data analysis IBM SPSS Statistical package version 25.0. The moderation was evaluated by applying model 1 to our data and controlling for chronic morbidity, severity of illness, length of stay, background characteristics, participation in WALK-FOR intervention, and baseline level of depressive symptoms. We probed the interaction between level of independence (i.e., cognitive status and independence in BADL and IADL) and family support hours by testing the conditional effects of family support at three levels of family visit duration: one standard deviation below the mean, at the mean, and one standard deviation above the mean.

Results

Sample characteristics

Participants’ (N = 370) mean age was 75.43 years (SD = 7.1); 221 (60.1%) participants were men; 245 (66.2%) participants were Jews. The average cognitive and functional statuses were relatively high, SPMSQ = 9.1 (SD = 1.4), and participants were mostly independent: BADL = 89.9 (SD = 15.4); IADL = 11.3 (SD = 4.8). Mean length of hospital stay was 6.1 days (SD = 3.5). The CCI score was similar to previous research in this population, 2.1 (SD = 1.8), and severity of illness was relatively low (M = 1.5, SD = 1.7), pointing to stable acuity level. Participants reported low to medium depressive symptoms at admission, 9.2 (SD = 3.4). Half (N = 187, 50.5%) participated in the WALK-FOR intervention.

Children were identified as the most common figures staying with older adults during hospitalization (N = 211, 57.0%); spouses also stayed (N = 112, 30.3%). In the remaining cases, other family members or friends stayed in the hospital (N = 45, 12.2%). Two respondents had no visitors. The mean number of hours a family member or friend stayed in the hospital was rather high (M = 7.8, SD = 3.3) (see Table 1).

Table 1. Participants’ Characteristics (N = 370).

	M (SD)/N(%)
Age	75.4 (7.1)
Gender: Women	149 (40.3%)
Ethnicity: Jews	245 (66.2%)
Family status: Married or living together	224 (60.5%)
Length of hospital stay (days)	6.1 (3.5)
Participation in WALK-FOR intervention: Included	187 (50.5%)
Comorbidity: CCI* (0–22)	2.1 (1.8)
Severity of illness: NEWS*(0–24)	1.5 (1.7)
Cognitive status: SPMSQ* (0–10)	9.1 (1.4)
Activities of daily living: BADL* (0–100)	89.9 (15.4)
IADL* (0–16)	11.3 (4.8)
Depressive symptoms at entrance: HADS*(0–21)	9.2 (3.4)
Family support (hours per day)	7.8 (3.3)

*CCI – Charlson Comorbidity score; NEWS - National Early Warning Score (Williams, 2022); SPMSQ - Pfeiffer Short Portable Mental Status Questionnaire (Pfeiffer, 1975); BADL - Modified Barthel Index for Activities of Daily Living (Mahoney & Barthel, 1965); IADL - Lawton’s Instrumental Activities of Daily Living (Lawton & Brody, 1969); HADS - Hospital Anxiety and Depression Scale (Zigmond & Snaith, 1983).

We used multivariable linear regression to test if cognitive status significantly predicted in-hospital depressive symptoms, after controlling health and background characteristics. The overall regression was statistically significant (R² = .341, F _{(10, 358)} = 18.535, p < .001), with lower cognitive status a significant predictor of increased in-hospital depressive symptoms (β = −0.137, p = .005).

A similar regression procedure tested whether level of independence in BADL significantly predicted in-hospital depressive symptoms, after controlling health and background characteristics. The overall regression was statistically significant (R² = .379, F_{(10, 358)} = 21.822, p < .001), with decreases in level of independence in BADL significantly predicting increases in in-hospital depressive symptoms (β = −0.274, p < .001).

The overall regression testing independence in IADL was statistically significant (R² = .382, F_{(10, 358)} = 22.160, p < .001). Decreases in level of independence in IADL significantly predicted increases in hospital depression (β = −0.317, p < .001).

We performed moderation analysis to test our hypotheses on the interplay between the amount of time family members spend with the hospitalized patient, level of independence, i.e., cognitive status (hypothesis 1), BADL (hypothesis 2), and IADL (hypothesis 3), and levels of depressive symptoms. In these analyses, we also controlled for baseline depression and background characteristics.

Hypothesis 1:Cognitive Status

To test hypothesis 1, we probed interactional effects between average family visit duration and cognitive status on levels of depressive symptoms. As shown in Table 2, family support significantly moderated the relations between cognitive status and depressive symptoms (t = −2.16, p = .032).

Table 2. Interactive effect of cognitive status and family support on level of in-hospital depressive symptoms.

			95% CI
Effect	Estimate	SE	LL	UL	p
Constant	3.696	.450	−5.605	8.932	.653
Cognitive status	.315	.819	−.361	.876	.413
Family support	.310	1.981	.004	1.223	.048
Interaction Cognitive Status x Family Support	.034	−2.155	−.139	−.006	.032
Depressive symptoms at entrance	.048	9.050	.341	.530	<.001
Participation in WALK-FOR intervention	.319	2.277	.099	1.353	.023
Severity of illness	.091	2.888	.084	.443	.004
Sex	.322	1.688	−.090	1.176	.092
Jews	.366	−.796	−1.010	.428	.426
Age	.023	−.450	−.054	.034	.653
Comorbidity	.087	.730	−.108	.236	.466
Length of hospital stay	.046	2.271	.014	.193	.024

There was no relationship between cognitive status and depressive symptoms when duration of family visit was one standard deviation below the mean (β = −0.06, p = ns), but there was an association when it was one standard deviation above (β = 0.55, p < .001) or at the mean (β = −0.31, p = .016). The interaction is illustrated in Figure 1.

Figure 1. Interaction between family support and cognitive status.

Hypothesis 2: Basic Activities of Daily Living

The second interaction analysis showed family support did not significantly moderate the relations between BADL status and depressive symptoms (t = 0.005, p = .087), after controlling for possible cofounders (see Table 3).

Table 3. Interactive effect of independence in basic activities of daily living and family support on level of in-hospital depressive symptoms.

			95% CI
Effect	Estimate	SE	LL	UL	p
Constant	2.982	2.515	1.636	13.366	0.012
Basic activities daily living	.025	−1.000	−.075	.025	.318
Family support	.249	1.502	−.116	.864	.134
Interaction BADL x Family Support	.003	−1.715	−.010	.001	.087
Depressive symptoms at entrance	.050	7.000	.254	.452	.000
Participation in WALK-FOR intervention	.309	2.574	.188	1.402	.011
Severity of illness	.090	2.649	.061	.414	.008
Sex	.311	1.540	−.133	1.092	.124
Ethnicity	.354	−.701	−.943	.447	.484
Age	.021	−.633	−.056	.028	.527
Comorbidity	.086	.105	−.160	.178	.916
Length of hospital stay	.044	2.401	.019	.193	.017

Hypothesis 3: Instrumental Activities of Daily Living

Analysis of IADL as the main predicting variable in the interaction between average family visit duration and level of depressive symptoms, as shown in Table 4, indicated family support significantly moderated the relations between level of IADL and depressive symptoms (t = −2.25, p = .025).

Table 4. Interactive effect of independence in instrumental activities of daily living and family support on level of in-hospital depressive symptoms.

			95% CI
Effect	Estimate	SE	LL	UL	P
Constant	3.696	.450	−5.605	8.932	.653
Instrumental activities daily living	.315	.819	−.361	.876	.413
Family support	.310	1.981	.004	1.223	.048
Interaction IADL x Family Support	.034	−2.155	−.139	−.006	.032
Depressive symptoms at entrance	.048	9.050	.341	.530	.000
Participation in WALK-FOR intervention	.319	2.277	.099	1.353	.023
Severity of illness	.091	2.888	.084	.443	.004
Sex	.322	1.688	−.090	1.176	.092
Ethnicity	.366	−.796	−1.010	.428	.426
Age	.023	−.450	−.054	.034	.653
Comorbidity	.087	.730	−.108	.236	.466
Length of hospital stay	.046	2.271	.014	.193	.024

We found interaction between independence in IADL and family support hours was significantly related to depression on all three levels: one standard deviation below the mean (β = −.17, p < .001), at the mean (β = −.24, p < .001), and one standard deviation above the mean (β = −.31, p < .001). The interaction is illustrated in Figure 2.

Figure 2. Interaction between family support and instrumental activities of daily living status.

Discussion

In this study, we examined the relations between in-hospital depression and family support during hospitalization in a sample of older adults. We found significant associations between cognitive status and independence in IADL and level of in-hospital depression. Our results are in line with previous investigations demonstrating the relationship between IADL and depression in the hospital setting (Helvik et al., 2013; Reichardt et al., 2020). They extend previous findings on associations between cognition and depression in the community (Wang & Blazer, 2015) by demonstrating these associations in a hospital setting.

The stress-buffering hypothesis was partially supported, depending on levels of cognition and IADL levels. More independent older adults whose family members stayed longer hours with them in the hospital had fewer depressive symptoms than independent older adults whose visitors stayed for shorter periods. Hours of family visits in the hospital stay might capture the idiosyncratic ways family provides the emotional support found by previous studies to lead to lower levels of depressive symptoms (Gur-Yaish et al., 2013). Future observational studies could document the things family members do with older adults during the hospital stay that protect them from depressive symptoms.

However, relations between depressive symptoms and family support were not apparent for dependent older adults, even though family members stayed slightly more hours. Our results are consistent with a study of non-hospitalized older adults (Van Orden et al., 2015), and extend these findings to the hospital setting. We offer three possible explanations of the support for the buffering hypothesis only among independent older adults. First, the dependent older adults in our sample had higher levels of depressive symptoms at admission. Thus, their depressive symptoms may have been part of their long-life experience (Heser et al., 2020; Morin et al., 2020; Taylor & Lynch, 2004). Dependent older adults’ depression may be a chronic condition, and family support might be more beneficial in acute situations, when depressive symptoms are more circumstantial. Second, it is plausible to assume that independent older adults arrive at the hospital with more reserves (Coventry et al., 2020; WHO, 2017b) and thus may benefit more from family members’ presence. Third, family caregivers of more dependent older adults report about greater caregiving burden (Ringer et al., 2017). This increased burden might affect their ability to provide emotional care and support during the hospital stay and overshadow the beneficial effects of family visits.

We found a significant interaction between functional independence and family support for IADL but not for BADL. IADL refers to activities to support daily life within the home and community that often require more complex interactions with the environment and other people than those used in BADL. Moreover, IADL autonomy is associated with executive functions, such as planning, working memory, attention, problem solving, verbal reasoning, and mental flexibility (Pashmdarfard & Azad, 2020). It is plausible that older patients can still perform basic activities, such as walking and toileting, but will not be able to shop or make their meals independently. In addition, older adults who are intact in IADL might benefit more from family visits, as they are more able to be engaged in activities that reduce depressive symptoms. Results from our study suggest it is important to assess IADL at admission to the hospital.

This study demonstrates the importance of investigating in-hospital depression and family support, considering initial levels of independence. It seems increased family support for more independent hospitalized older adults has a greater effect on depressive symptoms than similar support given to their more dependent comparators. Research and practice should acknowledge and measure the differences in older adults’ level of independence.

Most of our sample (65%) reported having depressive symptoms, and about a third reported abnormal levels (34%) of these symptoms. These percentages are somewhat higher than those found in previous studies (Ciro et al., 2012; Pederson et al., 2016; Reichardt et al., 2019; Shastri et al., 2019; Zisberg, 2017). Although this might reflect differences in measurement (Koenig et al., 1997; Luppa et al., 2012), the finding has important implications. Depression is a widely acknowledged condition, but it is under-diagnosed and often inadequately treated in primary care (WHO, 2017a). Detection of depressive symptoms during hospitalization can offer a starting point for treatment.

The mean number of hours family members stayed in the hospital in our sample was rather high but was similar to a study conducted in Italy (Ambrosi et al., 2017) and to other Israeli studies (Auslander, 2012; Gur-Yaish et al., 2013). Moreover, only two of our participants had no family visits during the hospital stay. These findings for Israel and Italy are not surprising: Italy and Israel are Mediterranean countries with a familial culture and close contact is common (Katz et al., 2010). Future studies should investigate the influence of family support in different countries with different family traditions of caregiving.

Limitations

The average cognitive and functional statuses of our participants were relatively high, and they were mostly independent with rather low levels of comorbidities and illness severity. Future research could include older adults with lower levels of independence and higher levels of cognitive decline. Second, we did not include measures of family relationship quality. Although most of the literature indicates the majority of older adults receive social support from their families and enjoy the moderating effect of this support on their mental health, in some situations, conflicts and negative aspects of family relationships may fail to create a protective effect (Harada et al., 2018; Rook, 2003), and in extreme cases, family visits can affect negatively hospitalization outcomes. Third, even though the vast majority of older adults are accompanied by family members during hospitalization, it is imperative to investigate the experience of older adults who have no family visits.

Conclusion

This study partially supports the stress-buffering hypothesis by demonstrating the ameliorative effect of social support from family members on relatively independent older adults’ levels of depressive symptoms during hospitalization. This finding highlights the importance of family support and suggests the need to create a comfortable environment conducive to such support. Emphasis should be placed on identifying depressive symptoms and level of independence at the time of admission to hospital and on raising awareness of the different needs of dependent and independent older adults.

Clinical implications

• Family support has a differential moderating effect on depressive symptoms of older adults during hospitalization, and an assessment of level of family support, cognitive status, and IADL independence of older hospitalized adults may be used to improve their mental outcomes.

• Interprofessional collaboration among hospital staff, including doctors, nurses, and nursing assistants, should acknowledge the value of family visits for hospitalized older adults, even if they are independent.

• Collaborative practice with distribution of staff roles and responsibilities should enable the creation of an environment in which family support is optimally provided for the benefit of older inpatients.

• Detection of depressive symptoms during hospitalization is important and might be a starting point for treatment.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets used during the current study available from the corresponding author on reasonable request.

Additional information

Funding

The study was partially supported by seed money from HaEmek Medical Center and the Cheryl Spencer Research Center at the Nursing Department, University of Haifa. The funding organizations were not involved in the design of the study, the collection, analysis, and interpretation of data, or in the writing of the manuscript.