Older age (Pirkis et al., Reference Pirkis2009; Zhang et al., Reference Zhang, Xu, Nie, Zhang and Wu2012) and sexual orientation (Cochran and Mays, Reference Cochran and Mays2009) are risk factors for depression among men. Pirkis et al. (Reference Pirkis2009) reported the age-adjusted prevalence rate of clinically significant depression to be 8.6% and a major depressive episode to be 1.9% among a community sample of 9,013 Australian men aged 60 years and over. In relation to sexual orientation, Cochran and Mays (Reference Cochran and Mays2009) reported that 21.5% of their sample of gay men were likely to meet the criteria for major depressive disorder in the previous 12 months and were 2.3 times more likely to have a probable diagnosis than heterosexual men. It is difficult to obtain prevalence rates specifically for older gay men. Available evidence indicates that 30% of a sample of gay men aged 50–82 years self-reported feeling depressed (Shippy et al., Reference Shippy, Cantor and Brennan2004) and that the average score for depressive symptoms among a sample of gay men aged 44–75 years demonstrated that the sample experienced depressive symptoms “some or a little of the time” (Wight et al., Reference Wight, LeBlanc, de Vries and Detels2012). Given that older gay men are exposed to the risks associated with age and sexual orientation, it is likely to be a significant problem for this population.

A risk factor for depression among older adults is living alone (Chou et al., Reference Chou, Ho and Chi2006; Tintle et al., Reference Tintle, Bacon, Kostyuchenko, Gutovich and Bromet2011; Oh et al., Reference Oh, Park, Lee, Kim, Choi and Nam2015). This relationship has been explained at least in part by a lack of belonging. Living alone has been identified as a risk factor for thwarted belongingness, which occurs when the basic human need to belong is unmet (Van Orden et al., Reference Van Orden, Witte, Cukrowicz, Braithwaite, Selby and Joiner2010). Baumeister and Leary (Reference Baumeister and Leary1995) proposed that people “have a pervasive drive to form and maintain. . .lasting, positive, and significant interpersonal relationships” (p. 1). These interpersonal relationships need to be frequent, positive, and stable (Baumeister and Leary, Reference Baumeister and Leary1995). Others have proposed that people need a psychological sense of belonging for mental health (Hagerty et al., Reference Hagerty, Lynch-Sauer, Patusky, Bouwsema and Collier1992). Hagerty et al. (Reference Hagerty, Lynch-Sauer, Patusky, Bouwsema and Collier1992) defined a sense of belonging as an individual's experience of being valued or important, and as though one fits in with those around them. Low levels of a sense of belonging have been associated with depressive symptoms among older adults (McLaren et al., Reference McLaren, Gomez, Bailey and Van Der Horst2007) and among gay men (McLaren et al., Reference McLaren, Jude and McLachlan2008). The relationship has not been investigated among older gay men.

Although there is a theoretical explanation to link living alone with depression among older adults, empirical evidence is mixed, with the results varying according to gender, age, and nationality of the participants. For example, Oh et al. (Reference Oh, Park, Lee, Kim, Choi and Nam2015) supported the relationship between living alone and higher levels of depressive symptoms among older men and women in a Korean sample aged 60 years and older. Tintle et al. (Reference Tintle, Bacon, Kostyuchenko, Gutovich and Bromet2011) found the relationship among a sample of Ukrainian older men but not older women (aged 50 years and older) and Chou et al. (Reference Chou, Ho and Chi2006) found the link in older women but not older men in their sample of Chinese adults aged 60 years and older. There appears to be just one study that has examined the link between living alone and depression among gay men. An Australian study of gay men aged 50 years and older (Lyons, Reference Lyons2015) did not find an association between living alone and depression, raising the possibility that the relationship is not evident among older gay men.

The mixed empirical evidence on the link between living alone and depression suggests that the relationship is likely to be more complex than a simple direct relationship. Purcell et al. (Reference Purcell, Heisel, Speice, Franus, Conwell and Duberstein2012) conducted a study with 130 adults aged 50 years and older who suffered from a mood disorder and found that living alone was not associated with suicidal ideation. Results indicated, however, an interaction between living arrangements and family connectedness. Specifically, they found that older adults living with others and having higher levels of family connectedness reported significantly lower levels of suicidal ideation than those living alone. Whether family connectedness weakens the relationship between living alone and suicidal ideation was not detailed by the authors. It is likely, however, that relationships that assist to meet belongingness needs weaken the relationship between living alone and suicidal ideation. Given the strong relationship between depression and suicide (Nock et al., Reference Nock2008), it is probable that this moderation effect would be evident for the living arrangements–depressive symptoms relation.

Further examining the possible relationship among living alone and depression among older gay men is important, given that as gay men age, they are more likely to live alone (Lyons et al., Reference Lyons, Pitts and Grierson2012). Lyons et al. (Reference Lyons, Pitts and Grierson2012) found that 30% of gay men aged 40–49 lived alone, 38% aged 50–59 lived alone, and 44% of gay men aged 60 and older lived alone. Other research has indicated that 65.8% of gay men aged 50–82 years lived alone (Shippy et al., Reference Shippy, Cantor and Brennan2004). In addition, older gay men (40.66%) are more likely to live alone than older heterosexual men (15.15%; Fredriksen-Goldsen et al., Reference Fredriksen-Goldsen, Kim, Barkan, Muraco and Hoy-Ellis2013). Several factors contribute to the high proportion of older gay men living alone. In countries such as Australia, gay men cannot marry, and although they can be “partnered,” older gay men are less likely to have a partner than older heterosexual men (Stonewall, 2011; Fredriksen-Goldsen et al., Reference Fredriksen-Goldsen, Kim, Barkan, Muraco and Hoy-Ellis2013). In addition, older gay men are less likely to have children (28% compared with 88% of older heterosexual men) and to see their biological family on a regular basis (Stonewall, 2011). Consequently, older gay men are less likely than older heterosexual men to have a spouse/partner, children, or other family members to live with as they age.

The high proportion of older gay men living alone highlights the importance of further investigating the link between living alone and depressive symptoms among this sample, and in particular, whether thwarted belongingness associated with living alone can be compensated by belonging in other contexts. Research has demonstrated that belonging to the gay community can be a protective factor for gay men (McLaren et al., Reference McLaren, Jude and McLachlan2008; McLaren et al., Reference McLaren, Gibbs and Watts2013). Research indicates that one way gay and bisexual men experience gay community is through their relationships with their gay friends, both individually and as a network of friends (Woolwine, Reference Woolwine2000; LeBeau and Jellison, Reference LeBeau and Jellison2009). Indeed, 63.5% of LeBeau and Jellison's sample reported that social life was a key advantage of being involved in the gay community. The men reported that romantic relationships and friendships were key reasons to be connected with the gay community. Gay friends provide a space for trust, shared interests, acceptance, mutual understanding, and emotional closeness (de Vries and Megathlin, Reference de Vries and Megathlin2009) and can perform many family functions (Woolwine, Reference Woolwine2000; de Vries and Megathlin, Reference de Vries and Megathlin2009). Woolwine proposed that the friendships established between gay men are likely to be protective. Recent research confirmed this, indicating that belonging with gay friends was directly related to lower levels of depressive symptoms among a sample of gay men aged from 18 to 79 years (Morris et al., Reference Morris, McLaren, McLachlan and Jenkins2015). Whether belonging with gay friends can negate the thwarted belongingness associated with living alone among older gay men is unknown.

The proposal that belonging with gay friends can weaken the relationship between living alone and depressive symptoms is consistent with the risk-protective model of resiliency (Hollister-Wagner et al., Reference Hollister-Wagner, Foshee and Jackson2001). The model proposes that the strength of the relationship between a risk factor (living alone) and an outcome factor (depressive symptoms) depends upon the presence of protective factors (belonging with gay friends), with the protective factors weakening the relationship between risk and outcome.

In summary, living alone has been identified as a risk factor for depression in some samples of older adults. It has been theorized that this relationship can be explained, at least in part, by belongingness needs being unmet. It is unclear, however, whether living alone is related to depression among older gay men. It is also unclear whether belongingness needs being met through other mechanisms, such as gay friends, can compensate for thwarted belongingness associated with living alone and weaken the relationship between living alone and depressive symptoms among older gay men. The aim of the current study was, therefore, to investigate whether living alone is associated with depressive symptoms among older gay men, and whether belonging with gay friends moderates the relation between living alone and depressive symptoms among older gay men.

Methods

Participants

A convenience sample of 160 Australian self-identified gay men aged from 65 to 92 years (M = 72.74, SD = 7.99) participated in the study. The participants lived in the state of Victoria, Australia. Details of the sample can be seen in Table 1. The majority of men were partnered, lived with another person, had completed a university degree, and had disclosed their sexual orientation to the majority of people they knew.

Table 1. Frequency of demographic characteristics in sample (N = 164)

Measures

Potential participants were invited to participate in the study through an information letter that provided a brief written description of the research and the contact details for the researcher and relevant resources that could be accessed by the participant if they experienced any distress during or after completing the questionnaire, including a free 24-h telephone counseling service for gay men.

The demographic section of the questionnaire sought participants’ details including their age, relationship status, living arrangements, highest education level achieved, and the degree of disclosure of their sexual orientation. The men indicated their sexual orientation by choosing from a number of options, including heterosexual, gay male, bisexual, and do not know/unsure. Only men who identified as gay were included in the study.

The 20-item Centre for Epidemiologic Studies-Depression Scale (CES-D) was used to measure depressive symptomatology (Radloff, Reference Radloff1977). Participants used a 4-point scale (0 = less than one day; 3 = 5–7 days) to indicate how often they had experienced depressive symptoms over the past week (e.g. I felt that everything I did was an effort). The scale has a range of 0–60, with higher scores indicating more depressive symptoms. A cut-off score of 12 has been demonstrated to have maximum sensitivity and specificity for identifying depression among older adults (Lewinsohn et al., Reference Lewinsohn, Seeley, Roberts and Allen1997). The CES-D has shown strong internal consistency for a sample of gay men (Cronbach's ∝ = 0.92, Morris et al., Reference Morris, McLaren, McLachlan and Jenkins2015) and for the current sample (Cronbach's ∝ = 0.96).

A sense of belonging with gay friends was assessed using two visual analogue scales (see Morris et al., Reference Morris, McLaren, McLachlan and Jenkins2015), which measured two psychological aspects of a sense of belonging (feeling needed or valued, and fit). Both visual analogue scales consisted of a 10-cm line anchored at each end with extreme positions (0 “Not needed or valued at all” to 10 “Completely needed or valued”; 0 = Do not fit in at all to 10 = Completely fit in”). Respondents placed a vertical line through both lines to indicate the extent to which they felt valued by and fitted in with their gay friends. The placement of the vertical line was measured to the nearest centimeter, and the two scores were summed. Total scores ranged between 0 and 20, with higher scores indicating a greater sense of belonging with gay friends. The measure showed high internal consistency for the current sample (Cronbach's ∝ = 0.97).

Procedure

Several strategies were used to recruit the sample. Older gay men were recruited at gay community events and functions, where the research was promoted and paper copies of the questionnaire were distributed. The questionnaire was also available online, and a link was placed on websites and in newsletters aimed at older gay men. Advertisements were also placed in agencies (e.g. healthcare services) that were utilized by older gay men. Snowball sampling was also used.

The men completed either a paper-based (n = 76, 47.5%) or online questionnaire (n = 84, 52.5%). A total of 150 paper questionnaires were distributed, resulting in a response rate of 50.7%. Most men (91%) who accessed the online survey completed it. The order of the measures was counterbalanced to minimize order effects.

Data analysis

Initially, descriptive statistics and correlations between the key variables were conducted. The moderation model was tested using Process (Model 1; Hayes, Reference Hayes2013) in SPSS. The number of bootstrap samples for bias corrected bootstrap confidence intervals (CIs) was 10,000, and a bootstrapped 95% CI was used to infer significance. Significance is supported if the CI does not include zero. Given that the variables of age (Polku et al., Reference Polku2015), relationship status, and education level (Kamiya et al., Reference Kamiya, Doyle, Henretta and Timonen2013) are related to depression among older adults, questionnaire type (online/paper) is related to a sense of belonging among gay men and lesbians (McLaren et al., Reference McLaren, Gibbs and Watts2013), and the disclosure of sexual orientation is related to suicide among gay men and lesbians (Plöderl et al., Reference Plöderl, Sellmeier, Fartacek, Pichler, Fartacek and Kralovec2014); these variables were controlled for in the analyses.

Results

A total of 31.2% of the older gay men scored 12 or above on the CES-D, indicating that almost one-third of the sample reported clinically significant levels of depressive symptoms. The mean scores and standard deviations for each variable, and the bivariate and partial correlations between variables, are shown in Table 2. Results of the correlations indicated that living alone was associated with lower levels of belonging with gay friends and higher levels of depressive symptoms. Higher levels of belonging with gay friends were associated with fewer depressive symptoms.

Table 2. Mean scores and standard deviations of variables, and correlations between variables (N = 160)

Note. Bivariate correlations are above the diagonal, and partial correlations, controlling for age, relationship status, education level, disclosure of sexual orientation, and type of questionnaire, are below the diagonal.

^a 1 = living alone, 2 = living with another.

*p < 0.05. **p < 0.01. ***p < 0.001.

Results testing the moderation model can be seen in Table 3. The interaction term was significantly related to depressive symptoms, providing support for the moderation model. The interaction term explained an additional 1.3% of the variance in depressive symptoms, F(1, 151) = 5.25, p = 0.023. Figure 1 displays the interaction between living arrangement and belonging with gay friends predicting depressive symptoms. For the graph, the effects of living arrangement and belonging with gay friends on depressive symptoms were plotted at three points: high, average, and low. High, average, and low values for belonging with gay friends were +1 SD, 0 SD, and −1 SD, respectively, of the centered mean of zero. The direct effect between living arrangement and depressive symptoms was significant for low, b = −13.55, p < 0.001 (95% CI = −18.80, −8.29), average, b = −9.76, p < 0.001 (95% CI = −13.47, −6.05), and high, b = −5.98, p = 0.011 (95% CI = −10.59, −1.37), levels of belonging with gay friends. Results indicate that the relationship between living alone and depressive symptoms weakened with increasing levels of belonging with gay friends.

Table 3. Results of regression analysis testing moderation model (N = 164)

Note. ^a1 partnered, 2 = unpartnered; ^b1 = paper, 2 = online; ^c1 = living alone, 2 = living with another.

*p < 0.05. **p < 0.01. ***p < 0.001.

Figure 1. Interaction between living arrangement and belonging with gay friends predicting depressive symptoms. Note. Living Arrangement: 1 = living alone, 2 = living with another.