Social contacts and depression in middle and advanced adulthood: Findings from a US national survey,2005

Abstract

Aim:

Empirical evidence has linked social contacts with mental stability. The aim is to assess how social contacts are associated with depression among the general population.

Methods:

We analysed the data of 5,681 adults aged 40 or older, who completed a depression screening as a part of the National Health and Nutrition Examination Survey, 2005–2008. Depression was ascertained using the Patient Health Questionnaire (PHQ), a nine-item screening instrument asking about the frequency of depression symptoms over the past two weeks. A PHQ score of 10 or higher was defined as depression.

Results:

The prevalence of depression was 5.54% (SE = 0.64) and 8.49% (SE = 0.71), respectively, among men and women. The association between social contacts and depression were more salient among men than women. The odds ratios (ORs) of depression were 2.43 (95% CI = 1.44−4.10), 2.42 (95% CI = 1.37−4.28), 1.00 (reference) and 1.94 (95% CI = 0.80−4.71), respectively, among men who never attended church, attended occasionally, weekly and more than weekly. The corresponding ORs were 1.79 (95% CI = 1.10−2.91), 1.72 (95% CI = 1.06−2.80), 1.00(reference) and 0.98 (95% CI = 0.52−1.84) for women. Compared with the respondents who had 10 or more friends, the ORs of depression were 4.01 (95% CI = 1.89−8.50) and 1.86 (95% CI = 0.92−3.79), respectively, for men and women who had no close friends.

Conclusions:

The current study concluded that traditional social contacts remain strongly associated with depression in the digital era. Digital social networking is one of the biggest growing industries, creating a new platform to make social contacts. There is an urgent need to explore how to maximize the potential of digital social networking to strengthen social bonds while minimizing its negative effects.

Keywords

Social network depression gender religion marriage

Background

Depression has been cited as one of the leading causes of disability worldwide (WHO, 2003). Although the precise cause of depression remains unclear, a large array of factors have been identified as risks to developing or triggering depression. In addition to biological and psychological components, social isolation has long been postulated as a major risk factor, and the support received from social contacts is believed to buffer against the effects of depression (Cheng, Lee & Chow, 2010; Heaney & Israel, 2002; Rosenquist, Fowler & Christakis, 2011; Stice, Ragan & Randall, 2004; Vollmann et al., 2010; Zhang, McKeown, Hussey, Thompson & Woods, 2005). Social contacts may include church membership, family support, network of friends, colleagues from work, and others. In addition to religious teachings and beliefs, which provide a coherent framework to interpret life’s events, the physical dimension of religiosity, namely attending services, has been hypothesized to grant access to an extensive social network among fellow members and be an important source of social integration (Miller & Thoresen, 2003; Skarupski, Fitchett, Evans & Mendes de Leon, 2010). Marriage as part of the family support, and the major node of the social networks, helps people fend off psychological assaults.

Although the postulated protective effects of social contacts, largely anecdotal, have surfaced in the realm of public health, only recently have researchers began to vigorously evaluate the clinical implication of social networks to improve the mental well-being of the general populations (Dew et al., 2010; Miller & Thoresen, 2003; Rosenquist et al., 2011). The majority of a relatively small number of previous studies were conducted either in clinical settings (Cheng et al., 2010; Lett et al., 2009) or among populations with co-mobility from various mental or physical conditions (Barefoot, Brummett, Clapp-Channing et al., 2000; Barefoot, Brummett, Helms et al., 2000; Brummett et al., 1998; Frasure-Smith et al., 2000; Holahan, Moos, Holahan & Brennan, 1995, 1997; Kilbourne, Cummings & Levine, 2009; Koenig, 2007), suffering from small sample sizes (Cheng et al., 2010; Dew et al., 2010; Stice et al., 2004). A resultant limitation is that confounding effects were not adequately controlled for due to insufficient statistical power (Kendler, Myers & Prescott, 2005); therefore, these studies cannot determine the relative strength of social contacts compared to other variables, and whether the relationship is largely spurious (Kilbourne et al., 2009). In addition, the conclusions from these investigations among populations with various co-existing conditions were not generalizable. Equally important, the significance of both religiosity and marriage has been diminishing dramatically in recent years due to the radical shift in society involving technology. The conclusions from the pre-digital era need to be re-evaluated. Focusing on three aspects of social contacts – church visitation, marital status and the number of close friends – the aim of the current study is to examine whether social contact remains a protective buffering factor against depression in the digital age. The relative richness of information collected in a national survey enables us to control for the confounding effects from social deprivation, lifestyle and history of major medical illness.

Methods

Study population

The National Health and Nutrition Examination Survey (NHANES) is an ongoing stratified, multi-stage probability sample of the non-institutionalized population designed to represent the health and nutritional status of general populations of all ages in the USA. The data from the surveys 2005–2008 were used. The study population for the current analyses was limited to 10,283 adults aged 40 and older who were eligible for the social support component of NHANES. After excluding 4,106 respondents whose data on church attendance or depression were missing, 35 with missing data on the number of close friends, two without information on marital status, 434 having no data on family income, and 25 missing the history of major medical illnesses, we analysed the data from the remaining 2,836 men and 2,845 women.

Variable definitions and measurements

Depression

Depression was assessed using the Patient Health Questionnaire (PHQ), a commonly used tool to assist primary care clinicians in diagnosing depression (Kroenke, Spitzer & Williams, 2001; Kroenke, Spitzer, Williams & Lowe, 2010). Based on the diagnostic criteria for major depressive disorder in the Diagnostic and Statistical Manual Fourth Edition (APA 2000), the PHQ consists of nine descriptive statements of depressive mood addressing items such as feelings of worthlessness, hopelessness, loneliness, loss of appetite, sleep disturbances, concentration problems, and psychomotor retardation. Interview questions were asked using the computer-assisted personal interview (CAPI) at mobile examination centres. Participants were asked to rate each item according to the frequency experienced in the last two weeks on a standard four-point scale from 0 to 3 (0 = not at all, 1 = several days, 2 = more than half the day, 3 = nearly every day). The total score of nine items had a potential range of 0–27, with the higher representing responses in the depressed range. Depression scores on the PHQ can be categorized as follows: 0–4 as no depression; 5–9 as minimal; 10–14 as mild; 15–19 as moderate; 20–27 as severe (Kroenke et al., 2001; Kroenke et al., 2010; Spitzer, Kroenke & Williams, 1999). To identify distinct levels of depressive symptoms and provide adequate numbers for each level of risk, the current study used the cut-off of 10 to dichotomize the study participants. Individuals with a depression score below 10 were classified as depression-free; otherwise as the group with depression. Participants who missed any one of the nine items on the questionnaire were excluded from the current analyses as having missing values.

Social contacts

Social contacts were assessed using the Social Support Questionnaire (SSQ) with a computer-assisted interviewer system. The questions in SSQ were selected from the Yale Health and Aging Study (MacArthur Studies of Successful Aging) (Berkman et al. 1993; Kubzansky, Berkman, Glass & Seeman, 1998) and the Social Network Index – Alameda County Study, and have been validated by previous investigations (Seeman, Kaplan, Knudsen, Cohen & Guralnik, 1987; Seeman, Lusignolo, Albert & Berkman, 2001). The current study abstracted three indicators of social contact: church or religious service attendance; marital status; and the number of close friends. Church or religious service attendance was measured by the number of times attending over the course of weeks, or months, and re-coded as never, occasionally, weekly and more than weekly. Marital status included currently married (including individuals married and living together, living with partners), previously married (including widowed, divorced and married but living separately) and never married. The number of close friends was counted, including relatives and non-relatives with whom a participant could talk with about private matters or call upon for help. Due to the left-shifted distribution, the number of close friends was re-coded into four categories to identify distinct sizes of network and to provide adequate numbers at risk in each level: no close friends; 1–4; 5–9 and 10 or more close friends.

Major confounders and effect modifiers

The covariates for both genders included age, socio-economic status (SES), risk behaviours and history of major medical illnesses. The respondents were grouped by age as 40–64 years, 65–74 years and 75+ years. This categorization was chosen to divide those before and after the age of average retirement, in addition to those living past the average life expectancy. The National Center for Health Statistics (NCHS) standard definition for ethnicity was used, and was coded as non-Hispanic whites, non-Hispanic blacks, Mexican and other Hispanic Americans, and others (CDCP, 2009). Educational attainment was measured as the highest completed grade of school and categorized as below high school, high school/equivalent, some college years and college graduate or above. Poverty status was defined by the poverty income ratio (PIR) (HHESD & US Census Bureau, 2009), including poor (PIR < 1.0), near poor (1 ≤ PIR < 2), middle income (2 ≤ PIR < 4) and high income (PIR ≥ 4) (Bloom, Cohen & Freeman, 2009). The category for alcohol consumption separated heavy (every/just about every day) and moderate drinkers (2 to 3 times a week) from less regular drinkers or non-drinkers. Current smoking status rather than smoking history is more relevant to depressed mood (Zhang, McKeown, Hussey, Thompson & Woods, 2005; Zhang et al., 2005), therefore participants were asked: ‘During the past 30 days, on the days that you smoked, how many cigarettes did you smoke per day?’ We categorized the participants into four groups, heavy smokers (20+ cigarettes per day), moderate smokers (10–19 cigarettes per day), light smokers (1–9 cigarettes per day) and rare/never. The medical condition interview conducted in the mobile examination centres provided self-reported data on a broad range of health conditions. Respondents were asked: ‘Has a doctor or other health professional ever told you that you had any of the following conditions?’ This was followed by a list of various chronic and acute conditions, representing conditions that are both prevalent and associated with substantial morbidity in the US population (Zhang, McKeown, Hussey, Thompson & Woods, 2005). Based on the literature (Inaba et al., 2005; Zhang, McKeown, Hussey, Thompson & Woods, 2005), we selected the history of stroke, heart attack and cancer as potential confounders of the association to be examined.

Statistical analysis

We used the SAS procedures for survey (SAS, Research Triangle Park, NC) with appropriate weighting and nesting variables to produce accurate national estimates and to adjust for the over-sampling of specific populations. Odds ratios (ORs) with 95% confidential intervals (CIs) were calculated to determine whether the prevalence of depression in various levels of social contacts significantly differed from the prevalence at corresponding referent levels. Two-side p values less than .05 were considered as significant. It has been documented that depression occurs differentially between genders in terms of neurobiology and sociology (Zhang, McKeown, Hussey, Thompson & Woods, 2005; Zhang et al., 2005); therefore separate models were run for men and women. The estimates for all three social contact variables were made simultaneously from one multi-variable-adjusted regression model with all covariates and three social contact variables included in the right side. We did not use the -2 log likelihood test to simplify the regression model; instead, a fully saturated model was retained, including all the covariates described above regardless of p values. To avoid over-specification of the regression, no interaction terms were included in the gender-specific model. Sensitivity analyses were also performed using various cut-offs of PHQ scores to assess potential biases associated with the cut-offs to define depression.

Results

Table 1 presents the selected characteristics of the weighted study population by gender. Overall, no significant differences were identified between men and women in terms of age at interview, education attainment and race/ethnicity. However, women were more likely to live with a poor family, while men were more likely to be heavy smokers, drinkers and have a history of heart disease. A total of 181 men and 289 women were categorized as depressed, with a prevalence of 5.54% (SE = 0.64) and 8.94% (SE = 0.71), respectively. Significant differences were also observed between men and women in social contacts. More women than men attended church on a weekly basis: 29.57% (SE = 1.58) vs 21.44% (SE = 1.36). However, the percentage of women previously married was almost twice that of men: 31.79% (SE = 1.35) vs 17.96% (SE = 0.99).

Table 1.

Selected characteristics of weighted study population (sample of 5,681 adults aged 40+, NHANES 2005–2008).

Characteristics		Men (n = 2,836)		Women (n = 2,845)
		Unweighted Sample size^a	% (SE)	Unweighted Sample size^a	% (SE)	P ^b
Depression	Yes	181	5.54 (0.64)	289	8.49 (0.71)	< .01
Age group	40–64	1,763	76.03 (1.19)	1,825	73.03 (1.31)	< .01
	65–74	568	14.01 (0.75)	573	15.80 (0.89)
	75+	505	9.96 (0.69)	447	11.17 (0.88)
Race/ethnicity	Non-Hispanic white	1,553	77.50 (2.02)	1,488	76.65 (2.11)	.10
	Non-Hispanic black	587	9.42 (1.17)	634	10.95 (1.52)
	Mexican or other Hispanics	613	8.68 (1.04)	635	8.17 (0.91)
	Others	83	4.40 (0.61)	88	4.23 (0.52)
Income^c	High income	863	44.24 (2.18)	770	39.46 (1.73)	< .01
	Mid-income	817	29.76 (1.11)	794	29.81 (1.00)
	Near poor	720	17.26 (1.17)	793	20.57 (1.04)
	Poor	436	8.74 (0.74)	488	10.16 (0.77)
Education	Above high school	1,282	55.94 (2.06)	1,303	55.53 (1.79)	.38
	High school graduate	665	25.09 (1.24)	718	26.59 (0.99)
	Below high school	889	18.98 (1.43)	824	17.88 (1.31)
Cigarette	Heavy	271	11.80 (1.32)	191	8.03 (0.93)	< .01
smoking	Moderate	160	5.00 (0.59)	158	6.18 (0.57)
	Light	182	4.89 (0.44)	153	4.45 (0.62)
	Rare/never	2,223	78.30 (1.39)	2,343	81.34 (1.14)
Alcohol	Heavy	393	15.65 (1.33)	141	6.07 (0.73)	< .01
drinking	Moderate	217	9.45 (0.86)	131	5.93 (0.86)
	Light	482	19.04 (1.12)	269	12.86 (0.95)	< .01
	Rare/never	1,744	55.86 (1.40)	2,304	75.15 (1.43)
Heart disease	Yes	381	10.70 (0.69)	243	6.43 (0.57)	< .01
Church	Never	1,105	41.02 (1.84)	813	32.63 (1.79)	< .01
visitation	Occasionally	770	28.09 (0.90)	742	25.92 (0.82)
	Weekly	660	21.44 (1.36)	866	29.57 (1.58)
	More than weekly	301	9.45 (0.70)	424	11.88 (0.75)
Number of	No friends	148	3.09 (0.33)	105	2.39 (0.29)	.05
friends	1-4 friends	1,126	36.62 (1.44)	1,263	39.39 (1.75)
	5-9 friends	795	32.82 (1.20)	834	33.46 (1.29)
	10+ friends	767	27.47 (1.58)	643	24.77 (1.25)
Marital	Never	192	6.06 (0.66)	191	6.02 (0.66)	< .01
status	Currently married	2,045	75.97 (1.21)	1,517	62.19 (1.59)
	Previous married	599	17.96 (0.99)	1,137	31.79 (1.35)

SE = standard error; NHANES = National Health Examination and Nutrition Survey.

Sample sizes (n) were presented as unweighted but percentages (%) as weighted.

χ² p values for the difference between men and women.

Respondents were categorized based on the Poverty Income Ratio (PIR): poor (PIR < 1.0), near poor (1 ≤ PIR < 2), middle income (2 ≤ PIR < 4) and high income (PIR ≥ 4).

The reverse J-shaped association between church visitation and depression is illustrated in Figure 1. For men and women, both occasional and more than weekly visiting had a higher risk of depression than never and lower than weekly visiting. Among men, the crude ORs of depression were 3.93 (SE = 2.31−6.70), 2.65 (SE = 1.59−4.41), 1 (reference) and 1.94 (SE = 0.80 − 4.71), respectively, for never, occasionally, weekly and more than weekly church attendance (data not shown). A similar pattern of OR appeared among women: 2.57 (SE = 1.52−4.34), 2.43 (SE = 1.35−4.33), 1 (reference) and 2.16 (SE = 0.88−5.31) (data not shown). When all selected covariates were adjusted simultaneously, gender-specific associations were observed. The reverse J-shaped association remained among men, but an L-shaped association emerged among women. The adjusted ORs among women who attended church never, occasionally, weekly and more than weekly were 1.79 (SE = 1.10−2.91), 1.72 (SE = 1.06−2.80), 1 (reference) and 0.98 (SE = 0.52−1.84) (Table 2).

Figure 1.

Social contacts and prevalence of depression among men and women (sample of 5,681 adults aged 40+, NHANES 2005–2008).

Table 2

Odds ratios of depression by level of social contacts (sample of 5,681 adults aged 40+, NHANES 2005–2008)^a.

Social contact		Men (n = 2,836)		Women (n = 2,845)
Indicator	Level	OR	95% CI	OR	95% CI
Church visitation	Never	2.43	1.44–4.10	1.79	1.10–2.91
	Occasionally	2.42	1.37–4.28	1.72	1.06–2.80
	Weekly	1.00	(reference)	1.00	(reference)
	More than weekly	1.94	0.80–4.71	0.98	0.52–1.84
Number of	No friends	4.01	1.89–8.50	1.86	0.92–3.79
friends	1–4 friends	2.10	1.18–3.74	1.54	0.99–2.40
	5–9 friends	1.32	0.76–2.29	1.10	0.69–1.76
	10+ friends	1.00	(reference)	1.00	(reference)
Marital status^b	Never	1.51	0.79–2.89	0.64	0.36–1.13
	Previously	2.15	1.46–3.15	1.62	1.06–2.48
	Currently	1.00	(reference)	1.00	(reference)

CI = confidence interval; OR = odds ratio; SE = standard error; NHANES = National Health Examination and Nutrition Survey.

The variables adjusted included age, race, family income, education, smoking and drinking, total number of people living in the household, and history of major medical illness (history of doctor-diagnosed cardiovascular disease, stroke and cancer).

The indicator of total number of people living in the household was excluded from the right side of the regression models for this outcome variable to avoid the over-adjustment due to the strong correlation between marital status and the number of people living in the household.

Being currently married was significantly associated with low odds of depression compared with never or previously married; this association was much more salient among men (Figure 1). The unadjusted ORs associated with previous marriage in comparison with current marriage were 3.04 (SE = 2.19−4.22) for men and 1.82 (SE = 1.34−2.47) for women (data not shown). After multi-variable adjustment, the association between being never married disappeared for both genders, but the elevated odds associated with previous marriage remained, in particular, among men (with OR of 2.15 (SE = 1.46−3.15) (Table 2).

A linear relationship between the number of close friends and depression was observed for both men and women (Figure 1), where the lowest prevalence of depression occurred in those with 10 or more close friends, and the highest was seen among those who self-reported no close friends. No appreciable differences were identified between men and women from crude data: ORs of depression for respondents with no friends, 1–4 friends, 5–9 friends and 10 or more friends, respectively, were 4.70 (SE = 2.56−8.98), 2.72 (SE = 1.61−4.57), 1.30 (SE = 0.77−2.22) and 1 (reference) for men, and 4.82 (SE = 2.57−9.05), 2.30 (SE = 1.60−3.31), 1.09 (SE = 0.71−1.66) and 1 (reference) for women (data not shown). After adjustment for covariates, the linear association became much weaker among women, but remained pronounced for men (Table 2).

In the sensitivity analyses, we repeated the analyses described above using a cut-off of 5 and 15, respectively, to assess the potential biases introduced by the choice of cut-offs. All of the principal findings were retained but were less significant largely due to diminishing numbers of depression cases (for the cut-off = 15) or decreased specificity of definition (for the cut-off = 5). For the cut-off of 5, the ORs were 1.50 (SE = 0.85−2.64), 1.32 (SE = 0.90−1.93), 1 (reference) and 1.16 (SE = 0.82−1.66), respectively, for men never attending church, attending occasionally, weekly and more than weekly. For the cut-off of 15, the ORs were 6.55 (SE = 0.47−91.8), 4.69 (SE = 0.33−66.20), 1 (reference) and 2.11 (SE = 0.18−25.2), respectively.

Discussion

From a large nationally representative sample, we observed several important relationships. Traditional social contacts, through religious activity and marriage, remain strongly associated with a reduced risk of depression, and these relationships were more salient among men compared with women. The inherent limitation of cross-sectional design, however, prevented the authors from assessing the nature of these associations; the discussions below is limited to the correlation rather than causality.

The overall trend that the frequency of church visitation was reversely associated with the odds of depression is consistent with previous investigations (Braam et al., 2004; Kilbourne et al., 2009; Maselko, Gilman & Buka, 2009; Strawbridge, Shema, Cohen & Kaplan, 2001). One cross-sectional study conducted among diabetic patients found that church attendance weakened the severity of depression (Kilbourne et al., 2009). A cohort study also observed that those who attended any religious services experienced an approximately 30% lower odds of major depressive episode as compared to those who never attended services (Maselko et al., 2009). Surprisingly, a reversed J-shaped association was detected in the current study, showing that attending church more than weekly was associated with increased odds of depression compared with weekly visitation. It is likely that a depressed individual relies more heavily on religious coping methods that, in turn, could lead to increased religious service attendance. This interpretation is supported by research showing that religious coping was very prevalent among clinically ill populations, who may feel that God has abandoned them (Pargament, Koenig, Tarakeshwar & Hahn, 2004). The frequency of religious service attendance is measuring not only the physical dimension of religiosity but also the strength of a person’s relationship to a higher power or the quality of a person’s ‘relationship with God’ (Maselko et al., 2009). Attending church more than weekly might be an indicator of higher levels of religiosity. Evidence has indicated that higher levels of religiosity in some circumstances may be a marker of insecure attachment, which is associated with an increased risk of depression (Maselko et al., 2009).

Radical change has been seen in the past several decades regarding the importance and social function of marriage. However, its protective effect against depression seemingly remains. The current findings that the respondents who were currently married were more likely to be less depressed, the formerly married were the most burdened, and the never married fell squarely between these two, are consistent well with a study published almost 40 years ago (Pearlin & Johnson, 1977) and with investigations from non-western cultures (Inaba et al., 2005; Jang et al., 2009). Similar results were observed by Davies, Avison and McAlpine (1997), whose study was conducted among single and married mothers. The single mothers, whether never married, separated or divorced, were almost three times more likely than the married mothers to experience a major depressive disorder. Interpretations of Davies’s study, however, were subject to confounding effects due to a possible artifact of low SES of single mothers, along with extra stresses as a result of raising children alone (Inaba et al., 2005). The results from the current study confirmed that the protective effects from marriage were independent of low SES.

The observation that depression was significantly associated with the number of close friends in a dose–response fashion in middle and advanced adulthood in the general population was consistent with findings obtained among adolescents (Cornwell, 2003; Stice et al., 2004), adults with co-morbidity from physical illnesses (Frasure-Smith et al., 2000; Lett et al., 2009) and from oriental culture where familism is strongly valued (Cheng et al., 2010). Cheng et al. (2010) found that the size of a social network was positively associated with general well-being in a study conducted among nursing home residents in China. Studies also reported that the size of a perceived social support system, such as close contacts, was reversely related to depression (Cornwell, 2003; Stice et al., 2004; Yuh et al., 2008).

Perhaps the surprising finding of the current analyses was that, after adjustment for covariates, almost all the associations became much weaker among women but remained pronounced among men. Further efforts to identify the major confounder for the association in women revealed that there was no single variable among the selected covariates that predominately explained the significant reduction of the estimates among women but not men. This is opposite to the vast literature indicating that women are more likely to seek support in their social network than men, and women are more sensitive than men to the depressogenic effects of low levels of social contacts (Elliott, 2001; Kendler et al., 2005; Schraedley, Gotlib & Hayward, 1999). Girls showed a stronger association between social contacts and depressive symptoms compared with boys (Schraedley, Gotlib & Hayward, 1999). The level of social contacts significantly predicted the level of depression among middle- and advanced-aged women, while among men with the same age range, no significant association was found (Elliott, 2001). In a large opposite-sex dizygotic twin study, a global social contact robustly predicted the risk of subsequent episodes of major depression in the female twins in the opposite-sex pairs, while in the male twins the association was modest and non-significant (Kendler et al., 2005). At present, we are not able to plausibly explain the discrepancy between our results and previous observations. However, it has been suggested that social contact is much more reliably assessed in women than men (Kendler et al., 2005); therefore, the weaker associations between social contact and depression among men compared with women observed in other studies might be due to unreliable data among men. The relatively high response rates through intensified outreach efforts and high reliability through sophisticated quality control in NHANES allow the current study to achieve more reliable measurements among men compared to previous studies, making it possible to correct the underestimation and detect a strong association among men (NCHS, 2010).

Limitations and strengths

There are several notable limitations to this study. With a cross-sectional design, we were not able to ascertain the directionality of the association, or to clarify whether low social contact caused depression or the depressed individuals had isolated themselves after a string of depressive episodes. Being observational, the current study was unable to rule out the possibility that the association detected was a reflection of genetic covariance without a direct causal relationship (Yuh et al., 2008). The literature on social determinants of disease has been augmented in recent years suggesting that a person’s structural position within a network, the middle or edge, is associated with the development of depressive symptoms (Rosenquist et al., 2011). However, the information on structural position in a social network was not available from NHANES, potentially limiting the ability of the current study to identify the most influential supportive effects from various components of the social networks. The variables we used to measure social contacts were limited to the amount of contacts and networks; more broadly defined psychosocial factors, such as closeness, quality of emotional supports from spouses (or partners) and friends, should be included for a better understanding of the associations. Different religions have different views on an individual’s control over his/her destiny, and may have different impacts upon their behaviours and mental well-being (Pritchard & Baldwin, 2000). Clearly, the identification of specific factors contributing to this intrinsic effect of religious participation requires further investigations.

The current study has strengths as well. Compared with previous studies from clinical or psychiatric backgrounds, the participants in NHANES were selected from the community-dwelling general populations. The conclusions are relatively generalizable.

Conclusion

The findings of the current study are of great importance for both public health and clinical practices. Although the digital technology has been changing almost every aspect of modern society, the current study provides additional evidence that traditional social contacts, through religious activity and marriage, remain strongly associated with a reduced risk of depression. The interventions based on family, neighbourhood, church or community might have a greater impact and relevancy in the treatment or prevention of depression, in particular among men. In clinical practice, health care professionals should be trained and encouraged to identify individuals and families at risk of social isolation, and assist them to acquire and maintain the social contacts necessary for mental well-being (Schulz et al., 2008). Digital social networking is one of the biggest growing industries, creating a new platform to make social contacts. There is an urgent need to explore how to maximize the potential of digital social networking to strengthen social bonds while minimizing its negative effects.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

APA (American Psychiatric Association). (2000). Diagnostic and statistical manual of mental disorders, 4th edn. Arlington, VA: American Psychiatric Publishing.

Barefoot

J. C.

Brummett

B. H.

Clapp-Channing

N. E.

Siegler

I. C.

Vitaliano

P. P.

Williams

R. B.

Mark

D. B.

(2000). Moderators of the effect of social support on depressive symptoms in cardiac patients. American Journal of Cardiology, 86, 438–442.

Barefoot

J. C.

Brummett

B. H.

Helms

M. J.

Mark

D. B.

Siegler

I. C.

Williams

R. B.

(2000). Depressive symptoms and survival of patients with coronary artery disease. Psychosomatic Medicine, 62, 790–795.

Berkman

L. F.

Seeman

T. E.

Albert

Blazer

Kahn

Mohs

Rowe

(1993). High, usual and impaired functioning in community-dwelling older men and women: Findings from the MacArthur Foundation Research Network on Successful Aging. Journal of Clinical Epidemiology, 46, 1129–1140.

Bloom

Cohen

R. A.

Freeman

(2009). Summary health statistics for US children: National Health Interview Survey, 2008. Vital Health Statistics, 10, 1–81.

Braam

A. W.

Hein

Deeg

D. J.

Twisk

J. W.

Beekman

A. T.

Van

T. W.

(2004). Religious involvement and 6-year course of depressive symptoms in older Dutch citizens: Results from the Longitudinal Aging Study Amsterdam. Journal of Aging and Health, 16, 467–489.

Brummett

B. H.

Babyak

M. A.

Barefoot

J. C.

Bosworth

H. B.

Clapp-Channing

N. E.

Siegler

I. C.

Mark

D. B.

(1998). Social support and hostility as predictors of depressive symptoms in cardiac patients one month after hospitalization: A prospective study. Psychosomatic Medicine, 60, 707–713.

CDCP (Centers for Disease Control and Prevention). (2009). National health and nutrition examination survey: About the national health and nutrition examination survey. Atlanta, GA: CDCP.

Cheng

S. T.

Lee

C. K.

Chow

P. K.

(2010). Social support and psychological well-being of nursing home residents in Hong Kong. International Psychogeriatric, 22, 1185–1190.

10.

Cornwell

(2003). The dynamic properties of social support: Decay, growth, and staticity, and their effects on adolescent depression. Social Forces, 81, 953–978.

11.

Davies

Avison

McAlpine

D. D.

(1997). Significant life experiences and depression among single and married mothers. Journal of Marriage and Family, 59, 294–308.

12.

Dew

R. E.

Daniel

S. S.

Goldston

D. B.

McCall

W. V.

Kuchibhatla

Schleifer

Koenig

H. G.

(2010). A prospective study of religion/spirituality and depressive symptoms among adolescent psychiatric patients. Journal of Affective Disorders, 120, 149–157.

13.

Elliott

(2001). Gender differences in causes of depression. Women Health, 33, 163–177.

14.

Frasure-Smith

Lespérance

Gravel

Masson

Juneau

Talajic

Bourassa

M. G.

(2000). Social support, depression, and mortality during the first year after myocardial infarction. Circulation, 101, 1919–1924.

15.

Heaney

C. A.

Israel

(2002). Social networks and social support. In Glanz

Rimer

B. K.

Viswanath

(eds), Health behavior and health education: Theory, research, and practice (3rd edn, pp. 185 -209). San Francisco, CA: Jossey-Bass.

16.

HHESD (Housing and Household Economic Statistics Division) US Census Bureau. (2009). How the census bureau measures poverty (official measure). Washington, DC: HHESD.

17.

Holahan

C. J.

Moos

R. H.

Holahan

C. K.

Brennan

P. L.

(1995). Social support, coping, and depressive symptoms in a late-middle-aged sample of patients reporting cardiac illness. Health Psychology, 14, 152–163.

18.

Holahan

C. J.

Moos

R. H.

Holahan

C. K.

Brennan

P. L.

(1997). Social context, coping strategies, and depressive symptoms: An expanded model with cardiac patients. Journal of Personality and Social Psychology, 72, 918–928.

19.

Inaba

Thoits

P. A.

Ueno

Gove

W. R.

Evenson

R. J.

Sloan

(2005). Depression in the United States and Japan: Gender, marital status, and SES patterns. Social Science and Medicine. 61, 2280–2292.

20.

Jang

S. N.

Kawachi

Chang

Boo

Shin

H. G.

Lee

Cho

S. I.

(2009). Marital status, gender, and depression: Analysis of the baseline survey of the Korean Longitudinal Study of Ageing (KLoSA). Social Science and Medicine, 69, 1608–1615.

21.

Kendler

K. S.

Myers

Prescott

C. A.

(2005). Sex differences in the relationship between social support and risk for major depression: A longitudinal study of opposite-sex twin pairs. American Journal of Psychiatry, 162, 250–256.

22.

Kilbourne

Cummings

S. M.

Levine

R. S.

(2009). The influence of religiosity on depression among low-income people with diabetes. Health Social Work, 34, 137–147.

23.

Koenig

H. G.

(2007). Religion and depression in older medical inpatients. American Journal of Geriatric Psychiatry, 15, 282–291.

24.

Kroenke

Spitzer

R. L.

Williams

J. B.

(2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Intern Medicine, 16, 606–613.

25.

Kroenke

Spitzer

R. L.

Williams

J. B.

Lowe

(2010). The Patient Health Questionnaire somatic, anxiety, and depressive symptom scales: A systematic review. General Hospital Psychiatry, 32, 345–359.

26.

Kubzansky

L. D.

Berkman

Glass

Seeman

T. E.

(1998). Is educational attainment associated with shared determinants of health in the elderly? Findings from the MacArthur Studies of Successful Aging. Psychosomatic Medicine, 60, 578–585.

27.

Lett

H. S.

Blumenthal

J. A.

Babyak

M. A.

Catellier

D. J.

Carney

R. M.

Berkman

L. F.

Schneiderman

(2009). Dimensions of social support and depression in patients at increased psychosocial risk recovering from myocardial infarction. International Journal of Behavioral Medicine, 16, 248–258.

28.

Maselko

Gilman

S. E.

Buka

(2009). Religious service attendance and spiritual well-being are differentially associated with risk of major depression. Psychological Medicine, 39, 1009–1017.

29.

Miller

W. R.

Thoresen

C. E.

(2003). Spirituality, religion, and health: An emerging research field. American Psychology, 58, 24–35.

30.

NCHS (National Center for Health Statistics). (2010). NHANES: 50 years of contributions to public health. Available from: http://www.cdc.gov/nchs/video/nhanes50th_contributions/nhanes_contributions.htm.

31.

Pargament

K. I.

Koenig

H. G.

Tarakeshwar

Hahn

(2004). Religious coping methods as predictors of psychological, physical and spiritual outcomes among medically ill elderly patients: A two-year longitudinal study. Journal of Health Psychology, 9, 713–730.

32.

Pearlin

L. I.

Johnson

J. S.

(1977). Marital status, life-strains and depression. American Sociological Review, 42, 704–715.

33.

Pritchard

Baldwin

(2000). Effects of age and gender on elderly suicide rates in Catholic and Orthodox countries: An inadvertent neglect? International Journal of Geriatric Psychiatry, 15, 904–910.

34.

Rosenquist

J. N.

Fowler

J. H.

Christakis

N. A.

(2011). Social network determinants of depression. Molecular Psychiatry, 16, 273–281.

35.

Schraedley

P. K.

Gotlib

I. H.

Hayward

(1999). Gender differences in correlates of depressive symptoms in adolescents. Journal of Adolescent Health, 25, 98–108.

36.

Schulz

Pischke

C. R.

Weidner

Daubenmier

Elliot-Eller

Scherwitz

Ornish

(2008). Social support group attendance is related to blood pressure, health behaviours, and quality of life in the Multicenter Lifestyle Demonstration Project. Psychology, Health and Medicine, 13, 423–437.

37.

Seeman

T. E.

Kaplan

G. A.

Knudsen

Cohen

Guralnik

(1987). Social network ties and mortality among the elderly in the Alameda County Study. American Journal of Epidemiology, 126, 714–723.

38.

Seeman

T. E.

Lusignolo

T. M.

Albert

Berkman

(2001). Social relationships, social support, and patterns of cognitive aging in healthy, high-functioning older adults: MacArthur studies of successful aging. Health Psychology, 20, 243–255.

39.

Skarupski

K. A.

Fitchett

Evans

D. A.

Mendes

Leon

C. F.

(2010). Daily spiritual experiences in a biracial, community-based population of older adults. Aging and Mental Health, 14, 779–789.

40.

Stice

Ragan

Randall

(2004). Prospective relations between social support and depression: Differential direction of effects for parent and peer support? Journal of Abnormal Psychology, 113, 155–159.

41.

Strawbridge

W. J.

Shema

S. J.

Cohen

R. D.

Kaplan

G. A.

(2001). Religious attendance increases survival by improving and maintaining good health behaviors, mental health, and social relationships. Annals of Behavioral Medicine, 23, 68–74.

42.

Spitzer

R. L.

Kroenke

Williams

J. B.

(1999). Validation and utility of a self-report version of PRIME-MD: The PHQ primary care study. Primary care evaluation of mental disorders. Patient health questionnaire. Journal of American Medical Association, 282, 1737–1744.

43.

Vollmann

Scharloo

Salewski

Dienst

Schonauer

Renner

(2010). Illness representations of depression and perceptions of the helpfulness of social support: Comparing depressed and never-depressed persons. Journal of Affective Disorders, 125, 213–220.

44.

WHO (World Health Organization). (2003). The world health report, 2003: Shaping the future. Geneva: WHO.

45.

Yuh

Neiderhiser

J. M.

Spotts

E. L.

Pedersen

N. L.

Lichtenstein

Hansson

Reiss

(2008). The role of temperament and social support in depressive symptoms: A twin study of mid-aged women. Journal of Affective Disorders, 106, 99–105.

46.

Zhang

McKeown

R. E.

Hussey

J. R.

Thompson

S. J.

Woods

J. R.

(2005). Gender differences in risk factors for attempted suicide among young adults: Findings from the Third National Health and Nutrition Examination Survey. Annals of Epidemiology, 15, 167–174.

47.

Zhang

McKeown

R. E.

Hussey

J. R.

Thompson

S. J.

Woods

J. R.

Ainsworth

B. E.

(2005). Low HDL cholesterol is associated with suicide attempt among young healthy women: The Third National Health and Nutrition Examination Survey. Journal of Affective Disorders, 89, 25–33.

Social contacts and depression in middle and advanced adulthood: Findings from a US national survey,2005–2008

Abstract

Aim:

Methods:

Results:

Conclusions:

Keywords

Background

Methods

Study population

Variable definitions and measurements

Depression

Social contacts

Major confounders and effect modifiers

Statistical analysis

Results

Discussion

Limitations and strengths

Conclusion

Footnotes

Funding

References