Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jan 31.
Published in final edited form as: Prev Med. 2014 Dec 15;0:61–66. doi: 10.1016/j.ypmed.2014.12.008

Social Engagement and Chronic Disease Risk Behaviors: The Multi-Ethnic Study of Atherosclerosis

Laura J Samuel a,1, Cheryl R Dennison Himmelfarb a, Moyses Szklo b, Teresa E Seeman c, Sandra E Echeverria d, Ana V Diez Roux e,2
PMCID: PMC4329061  NIHMSID: NIHMS650612  PMID: 25524614

Abstract

Objective

Although engagement in social networks is important to health, multiple different dimensions exist. This study identifies which dimensions are associated with chronic disease risk behaviors.

Method

Cross-sectional data on social support, loneliness, and neighborhood social cohesion from 5381 participants, aged 45–84 from the Multi-Ethnic Study of Atherosclerosis was used.

Results

After adjusting for individual characteristics and all social engagement variables, social support was associated with lower smoking prevalence (PR=0.88, 95% CI: 0.82, 0.94), higher probability of having quit (PR=1.03, 95% CI: 1.01, 1.06) and a slightly higher probability of achieving physical activity recommendations (PR=1.03, 95% CI: 1.01, 1.06). Neighborhood social cohesion was associated with very slightly higher probability of achieving recommended (PR=1.03, 95% CI: 1.01, 1.05) or any regular (PR=1.0, 95% CI: 1.01, 1.04) physical activity, and a higher probability of consuming at least five daily fruit and vegetable servings (PR=1.05, 95% CI: 1.01, 1.09).

Conclusion

Both social support and neighborhood social cohesion, a less commonly considered aspect of social engagement, appear to be important for chronic disease prevention interventions and likely act via separate pathways.

Keywords: social engagement, social support, neighborhood social cohesion, physical activity, smoking

Introduction1

Risk behaviors, including smoking, lack of physical activity and poor diet, contribute to chronic disease, including cardiovascular disease, burden (Mokdad, Marks, Stroup, & Gerberding, 2004). Social engagement, meaning the degree of an individual’s involvement in social networks, may reduce risk behaviors by enhancing self-efficacy, reducing distress and facilitating access to health-related information (Berkman & Krishna, 2014). Social engagement can be conceptualized along multiple dimensions, each capturing a resource gained from social networks (Cohen & Wills, 1985). For example, emotional social support is the love, care and trust in social networks (House, 1981). Another dimension gaining attention is loneliness, representing perceived social and emotional isolation (Hawkley, Browne, & Cacioppo, 2005). Finally, neighborhood social cohesion captures solidarity with community networks (Kawachi & Berkman, 2000).

Emotional social support (Poortinga, 2006a; Delva et al., 2006; Holahan et al., 2011; Vaananen, Kouvonen, Kivimaki, Pentti, & Vahtera, 2008) and neighborhood social cohesion (Carpiano, 2007; Kandula, Wen, Jacobs, & Lauderdale, 2009; Li, Horner, & Delva, 2012) are generally associated with lower, and loneliness with higher (Lauder, Mummery, Jones, & Caperchione, 2006; Shankar, McMunn, Banks, & Steptoe, 2011), smoking rates. However, social support (Yun, Kang, Lim, Oh, & Son, 2010) and social cohesion (Chuang & Chuang, 2008; Li et al., 2012) may be associated with higher smoking rates in groups with high rates of smoking. Emotional social support (Weyers et al., 2010) and neighborhood social cohesion (Cleland et al., 2010; Cradock, Kawachi, Colditz, Gortmaker, & Buka, 2009; Pabayo, Belsky, Gauvin, & Curtis, 2010; Shelton et al., 2011; Utter, Denny, Robinson, Ameratunga, & Milfont, 2011; Echeverria, Diez-Roux, Shea, Borrell, & Jackson, 2008) are often associated with greater, and loneliness with less (Hawkley, Thisted, & Cacioppo, 2009; Shankar et al., 2011), physical activity, although associations are inconsistent for social support (Debnam, Holt, Clark, Roth, & Southward, 2012; Poortinga, 2006b) loneliness (Lauder et al., 2006) and social cohesion (Ball et al., 2010; Veitch et al., 2012) in similar large, diverse samples. Emotional social support is also associated with greater fruit and vegetable intake (Poortinga, 2006a; Debnam et al., 2012).

Many studies are limited to one dimension of social engagement, precluding their comparison. Also, these variables should, theoretically, have synergistic interactions, so prior results may underestimate the total potential effect of social engagement on behaviors (Uchino, 2004). There is some evidence of synergistic interactions between social support and loneliness as they relate to health (O’Donovan & Hughes, 2007; Pressman et al., 2005). The presence of countervailing or interacting influences of different types of social engagement may also account for prior conflicting findings. The purpose of this paper was to examine and contrast associations of several related, but distinct, measures of social engagement with behaviors and test for hypothesized synergistic interactions between them.

Methods

Sample

The Multi-Ethnic Study of Atherosclerosis (MESA) is a multi-ethnic cohort study investigating the prevalence and progression of subclinical cardiovascular disease, described elsewhere (Bild et al., 2002). Briefly, a 6814 participants aged 45 to 84 without clinical cardiovascular disease were recruited from six U.S. geographical areas: Baltimore City and Baltimore County, Maryland; Chicago, Illinois; Forsyth County, North Carolina; Los Angeles County, California; New York City, New York; and St. Paul, Minnesota. Each site employed slightly different sampling procedures. However, all sites used random sampling strategies to recruit from available community lists and attempted to recruit equal numbers of men and women from at least two a priori categorized racial/ethnic groups (White, Black, Hispanic, and Chinese) to facilitate racial/ethnic comparison of risk factors.

Data collection

Data for these analyses were obtained during the baseline in-clinic examination, which occurred between July 2000 and July 2002, except for loneliness, which was measured in the fourth in-clinic examination, carried out between July 2005 and July 2007. Loneliness is included in the current analyses as there is evidence that loneliness is relatively stable during adulthood (Boomsma, Willemsen, Dolan, Hawkley, & Cacioppo, 2005). Participants with complete data for analyses of physical activity (n=5378), fruit and vegetable intake (n=4966) current smoking status (n=3408) and smoking cessation amongst all who ever smoked (n=2627) were included.

Outcome variables

Three behaviors, each capturing slightly different aspects of chronic disease risk, were dichotomized, using clinically relevant cut points. Smoking status was derived by asking “Have you smoked cigarettes during the last 30 days?” and “Have you smoked at least 100 cigarettes in your lifetime?”. To evaluate the progression from never smoking, to smoking, to cessation, current smokers were compared to never smokers, and former smokers were compared to current smokers. Participants who achieved CDC-recommended levels of physical activity (≥500 Metabolic equivalent (MET) minutes of moderate to vigorous physical activity weekly) (U.S. Department of Health and Human Services, 2008) were compared to those who reported less. Separately, participants who reported any regular moderate to vigorous physical activity were compared to those who reported none. MET minutes of leisure-time moderate and vigorous physical activity per week were calculated by multiplying the minutes spent in each activity by the MET level for the activity, defined by prior research (Ainsworth et al., 2000), using data from the MESA Typical Week Physical Activity Survey, which was adapted from a previously validated survey (Whitt, Levin, Ainsworth, & Dubose, 2003). The survey asks participants if they performed various activities in a “typical week in the past month”, and records the level of effort and amount of time for each activity. Fruit and vegetable intake was calculated as the average daily servings of previously itemized fruit, fruit juice, and vegetable foods (Nettleton et al., 2006) using responses to the MESA 120-item food frequency questionnaire, which is adapted from a previously validated questionnaire (Block, Woods, Potosky, & Clifford, 1990) and assesses typical diet over the past year. In analyses, five daily servings of fruits and vegetables was used as a cut point, which is roughly equivalent to the minimal suggested intake (U.S. Department of Agriculture & U.S. Department of Health and Human Services, 2010), but since most participants did not meet this recommendation, a separate model also compared those who consumed at least two daily servings to those who consumed less than two, comparable to prior work (Poortinga, 2006a).

Main Independent Variables

Social support was measured using the six-item emotional social support index (Mitchell et al., 2003), which asks about having someone available to listen, or provide advice, or show affection (Cronbach’s α in this sample = 0.88). Loneliness was measured with an instrument derived from the revised University of California at Los Angeles Loneliness Scale (Russell, Peplau, & Cutrona, 1980) (α = 0.79), asking participants how often they lack companionship, feel left out, or isolated from others. Neighborhood social cohesion was assessed with the instrument from the Project on Human Development in Chicago Neighborhoods (Sampson, Raudenbush, & Earls, 1997), asking participants if the neighborhood is close-knit and whether neighbors help each other, get along, can be trusted, and share the same values (α = 0.70). Because hypotheses focused on how individual-level perceptions of social engagement influence health behaviors, perceived neighborhood social cohesion was examined as an individual-level variable. Scores from each instrument were standardized (i.e. z scores) prior to analyses.

Covariates

Demographic characteristics, socioeconomic factors and self-reported health were considered potential confounders and adjusted for in analyses. This included age, sex, marital status (married, widowed, divorced/separated, never married), race/ethnicity (White, Black, Hispanic, Chinese), family annual income (< $16,000, $16,000–$29,999, $30,000–$49,999, $50,000–$74,999, $>75,000), level of education achieved (< High School, High School/GED, Some college/technical school/Associate’s degree, Bachelor’s degree, Graduate degree) and study site (categorized as above). Self-reported health (poor/fair, good, very good, excellent) was also included since health declines may lead to both social isolation and behavior changes.

Statistical Analysis

Prevalence ratios (PR) of behaviors associated with standardized social engagement variables (i.e. z scores) were modeled with Poisson regression, using robust standard errors (Zou, 2004; Wacholder, 1986) in Stata 10 (StataCorp, 2007). Models were built in a step-wise fashion, testing unadjusted associations, then adjusting for covariates prior to adding all social engagement variables. Interaction terms between each pair of social engagement variables were then tested and retained in the model if they were both statistically significant (p<0.05) and improved model fit, based on Akaike Information Criterion (Akaike, 1974). Stratified analyses were used to further examine statistically significant (p<0.05) interactions. Correlations between social engagement variables were also examined.

Results

Table 1 describes characteristics of the sample, comparing individuals with low and high levels of each social engagement variable, split at the median value. Overall, socially engaged participants tended to be slightly older, male, White, married, and have higher incomes and better health. Social engagement variables were weakly and moderately correlated (see Table 1). There was also no evidence of collinearity in adjusted analyses (i.e. variance inflation factor≥10, tolerance ≤0.1). Also, loneliness, which was measured at the fourth examination, was correlated with social support, measured at both the first and fourth examination (Spearman ρ=−0.3879 and −0.5106, respectively). Linear associations between the logarithmic prevalence ratio and standardized social integration variables were confirmed using lowess plots, which is a nonparametric method used to visualize the relationship between variables (Cleveland & Mcgill, 1985).

Table 1.

Characteristics of participants and subgroup mean values of social engagement variables, Multi-Ethnic Study of Atherosclerosis (MESA) (n=5381)

Sample Characteristics Mean Social Support (SD)a Mean Loneliness (SD)a Mean Neighborhood Social Cohesion (SD)a
Overall Sample 24.24 (5.22) 4.00 (1.41) 17.57 (2.88)
Mean (SD)
Scale Range 6–30 3–9 5–25
Mean Age (SD) 61.35 (9.96)
Age Categories (%)
45–54 1648 (30.6) 23.88 (5.25) 4.19 (1.53) 17.42 (2.82)
55–64 1547 (28.8) 24.22 (5.27) 3.96 (1.36) 17.59 (2.92)
65–74 1576 (29.3) 24.53 (5.12) 3.88 (1.35) 17.70 (2.86)
75–84 610 (11.3) 24.52 (5.20) 3.87 (1.33) 17.56 (2.95)
Sex (%)
Female 2844 (52.8) 23.84 (5.19) 4.12 (1.47) 17.55 (2.90)
Male 2537 (47.2) 24.69 (5.21) 3.86 (1.33) 17.59 (2.86)
Race/Ethnicity (%)
White 2193 (40.8) 24.26 (5.14) 3.97 (1.40) 17.97 (2.79)
Chinese 648 (12.0) 23.74 (5.06) 3.80 (1.25) 17.11 (2.42)
Black 1370 (25.5) 24.43 (5.06) 3.99 (1.37) 17.70 (3.01)
Hispanic 1170 (21.7) 24.26 (5.60) 4.16 (1.56) 16.91 (2.97)
Education (%)
<High School 835 (15.5) 24.18 (5.78) 4.13 (1.54) 16.94 (3.01)
High School/GED 955 (17.8) 24.33 (5.09) 3.99 (1.42) 17.46 (2.90)
Some College 1540 (28.6) 24.14 (5.20) 3.96 (1.39) 17.52 (2.87)
Bachelor’s 994 (18.5) 24.16 (5.09) 3.97 (1.39) 17.81 (2.84)
Graduate School 1057 (19.6) 24.42 (5.01) 3.98 (1.36) 18.00 (2.71)
Income (%)
< $16,000 883 (16.4) 23.09 (5.83) 4.25 (1.63) 16.86 (2.95)
$16,000–$29,999 952 (17.7) 23.49 (5.46) 4.12 (1.44) 17.17 (2.80)
$30,000–$49,999 1266 (23.5) 23.91 (5.34) 4.00 (1.40) 17.53 (2.95)
$50,000–$74,999 944 (17.5) 24.71 (4.79) 3.91 (1.33) 17.85 (2.82)
$>75,000 1336 (24.8) 25.52 (4.43) 3.80 (1.26) 18.15 (2.71)
Marital Status (%)
Married 3394 (63.1) 25.36 (4.59) 3.80 (1.27) 17.74 (2.82)
Widowed 630 (11.7) 23.27 (5.54) 4.18 (1.53) 17.54 (3.03)
Divorced/ Separated 908 (16.9) 22.06 (5.54) 4.39 (1.59) 17.14 (2.94)
Never Married 449 (8.3) 21.55 (5.87) 4.45 (1.60) 17.14 (2.84)
Physical Health (%)
Poor/Fair 442 (8.2) 23.31 (5.46) 4.51 (1.58) 16.48 (3.15)
Good 2155 (40.1) 24.11 (5.31) 4.04 (1.46) 17.33 (2.86)
Very Good 1867 (34.7) 24.23 (5.13) 3.92 (1.35) 17.87 (2.71)
Excellent 917 (17.0) 25.02 (4.96) 3.80 (1.28) 18.04 (2.92)
Open in a new tab
a

Values represent scores for social engagement variables prior to standardization. Bold print identifies statistically significant subgroup differences based on ANOVA.

Unadjusted and adjusted prevalence ratios of risk behaviors in relation to a standard deviation increase of social engagement variables are shown in Table 2. In unadjusted models (Model 1), each standard deviation increase in either social support or neighborhood social cohesion was associated with a lower prevalence of smoking, and an increase in loneliness was associated with higher prevalence of smoking. After adjusting for other individual characteristics (Model 2), only associations with social support and loneliness remained statistically significant. After further adjustment for social support, loneliness and neighborhood social cohesion (Model 2), only the relationship with social support remained statistically significant.

Table 2.

Prevalence ratios of selected behaviors associated with a standard deviation increase in social engagement variables, Multi-Ethnic Study of Atherosclerosis, 2000–2002

Model 1a
PR (95% CI)		 Model 2b
PR (95% CI)		 Model 3c
PR (95% CI)
Current Smoking Status (n=3408)
Social Support 0.83 (0.78, 0.88) 0.87 (0.81, 0.92) 0.88 (0.82, 0.94)
Loneliness 1.15 (1.09, 1.22) 1.08 (1.02, 1.14) 1.03 (0.97, 1.10)
Neighborhood Social Cohesion 0.92 (0.85, 0.98) 0.97 (0.90, 1.03) 0.99 (0.93, 1.06)
Former Smoking (n=2627)
Social Support 1.07 (1.04, 1.09) 1.05 (1.02, 1.07) 1.03 (1.01, 1.06)
Loneliness 0.93 (0.91, 0.96) 0.96 (0.94, 0.99) 0.98 (0.95, 1.00)
Neighborhood Social Cohesion 1.03 (1.01, 1.05) 1.02 (1.00, 1.04) 1.01 (0.99, 1.03)
Recommended Level of Physical Activity (n=5378)
Social Support 1.05 (1.03, 1.08) 1.05 (1.02, 1.07) 1.03 (1.01, 1.06)
Loneliness 0.95 (0.93, 0.97) 0.96 (0.94, 0.98) 0.97 (0.95, 1.00)
Neighborhood Social Cohesion 1.05 (1.03, 1.07) 1.03 (1.01, 1.06) 1.03 (1.01, 1.05)
Any Regular Physical Activity (n=5378)
Social Support 1.03 (1.01, 1.05) 1.03 (1.01, 1.05) 1.02 (1.00, 1.04)
Loneliness 0.97 (0.96, 0.99) 0.98 (0.97, 1.00) 0.99 (0.97, 1.01)
Neighborhood Social Cohesion 1.04 (1.03, 1.06) 1.03 (1.01, 1.04) 1.02 (1.01, 1.04)
≥5 Daily Fruit and Vegetable Servings (n=4966)
Social Support 1.02 (0.98, 1.06) 1.02 (0.98, 1.06) 1.01 (0.97, 1.06)
Loneliness 0.99 (0.96, 1.03) 0.99 (0.96, 1.03) 1.00 (0.96, 1.05)
Neighborhood Social Cohesion 1.05 (1.01, 1.09) 1.05 (1.01, 1.09) 1.05 (1.01, 1.09)
≥2 Daily Fruit and Vegetable Servings (n=4966)
Social Support 1.01 (1.00, 1.02) 1.01 (1.00, 1.02) 1.01 (1.00, 1.02)
Loneliness 0.99 (0.98, 1.01) 1.00 (0.99, 1.01) 1.00 (0.99, 1.02)
Neighborhood Social Cohesion 1.01 (1.00, 1.02) 1.01 (1.00, 1.02) 1.01 (1.00, 1.02)
Open in a new tab
a

Unadjusted associations for each standard deviation increase in social engagement variable. Social support SD= 5.17 on scale ranging from 6–30. Loneliness SD=1.40 on scale ranging from 3–9. Neighborhood social cohesion SD=2.86 on scale ranging from 5–25.

b

Adjusted for age, sex, race/ethnicity, income, education, and study site, marital status, and physical health

c

Model added standardized scores for social support, loneliness and neighborhood social cohesion

In unadjusted models (Table 2, Model 1), higher social support and neighborhood social cohesion were both associated with a higher probability, and loneliness with a lower probability, of having quit amongst all participants who had ever smoked. After accounting for other individual characteristics (Model 2), only social support and loneliness remained associated with having quit and after accounting for all social engagement variables (Model 3), only social support was associated with having quit.

In unadjusted analyses, (Table 2, Model 1) higher social support and neighborhood social cohesion were associated with slightly higher probability, and loneliness with slightly lower probability, of achieving the recommended level of physical activity. These associations remained after adjusting for individual characteristics (Model 2), but after accounting for all social engagement variables (Model 3), only social support and neighborhood social cohesion remained statistically significant. A similar pattern of results was obtained when examining the probability of engaging in any regular physical activity, except that neighborhood social cohesion alone was statistically significant in the fully adjusted model (Model 3).

In unadjusted analyses, (Table 2, Model 1) neighborhood social cohesion was associated with a slightly higher probability of consuming at least five daily fruit and vegetable servings, and this association remained after adjusting for individual characteristics and other social engagement variables (Model 3). None of the social engagement variables was associated with the probability of consuming at least two daily fruit and vegetable servings.

Finally, this study tested interactions between dimensions of social engagement. There was evidence for interaction between social support and loneliness in the case of current smoking (p for interaction=0.028). In the fully adjusted model, the association of social support with current smoking was stronger in persons reporting no degree of loneliness (n=1823) than in those who reported some loneliness (n=1585): (PR=0.82, 95% CI: 0.74, 0.92 and PR=0.91, 95% CI: 0.84, 0.99 respectively). However, no other interactions were found between social engagement variables for risk behavior outcomes.

Discussion

This study is among the first to compare several dimensions of social engagement in relation to chronic disease risk behaviors and consider interactions among them. After adjustment for all three dimensions of social engagement, both social support and neighborhood social cohesion, a less commonly considered variable, were relevant to chronic disease risk behaviors. Social support was more strongly related to smoking prevalence in non-lonely individuals, but no other evidence was found for hypothesized synergistic interactions.

Although social support is a frequent focus of health researchers, few studies have compared its predictive utility relative to other measures of social engagement. This study adds to the literature by suggesting that while only social support is associated with smoking behaviors, both social support and neighborhood social cohesion are independently associated with physical activity and neighborhood social cohesion is relevant for dietary behaviors. In another Chicago study of middle aged and older adults, loneliness, and not social support, was associated with physical activity (Hawkley et al., 2009). Together, these results suggest that social support is important for smoking, but other aspects of social engagement should be considered for physical activity and diet.

There are several potential reasons for these results. For smoking, although there was no evidence of collinearity between loneliness and social support, the moderate correlation between social support and loneliness, combined with relatively weak associations, may have limited precision when estimating the independent associations of both variables. It is also possible that social support and loneliness act via separate intersecting pathways, as has been theorized elsewhere (Uchino, 2004), since the present study found interactions in their relationships with smoking behaviors.

Although these results suggest that social support is related to smoking behaviors, the evidence for physical activity and fruit and vegetable intake is not as strong in this study and others, reviewed earlier. Although these results do not fully elucidate prior inconsistent results for neighborhood cohesion and physical activity found even in longitudinal data (Cradock et al., 2009; Pabayo et al., 2010; Veitch et al., 2012), they add to the literature by hinting that perceived neighborhood cohesion may be more relevant than either social support or loneliness to physical activity, though the underlying association may be relatively weak. Also, the present study, among the first to examine associations between social engagement and fruit and vegetable intake, suggests that neighborhood social cohesion is associated with only a slightly higher likelihood of consuming the recommended intake of fruits and vegetables.

Lack of robust associations between social engagement and physical activity and diet may occur because numerous other factors, such as time, finances, cultural and normative factors, personal preferences and the physical environment, constrain physical activity and diet. Alternatively, we characterized neighborhood cohesion using individual perceptions because we hypothesized that perceptions influence behaviors. There may be other processes through which actual neighborhood-level cohesion (characterized as neighborhood aggregated measures) affect behaviors, such as communication of social norms (Ahern, Galea, Hubbard, & Syme, 2009) or associations of cohesion with physical environment features (Cohen, Inagami, & Finch, 2008). Weak associations may also be due, in part, to use of dichotomous outcomes. However, dichotomous outcomes ensures that the findings are clinically relevant by anchoring outcomes to recommended behavioral thresholds.

Finally, this study found very little evidence of hypothesized synergistic interactions between social engagement variables. Consistent with other studies (O’Donovan & Hughes, 2007; Pressman et al., 2005), social support and loneliness interacted. Perhaps either social support buffers loneliness-induced stress (Cohen & Wills, 1985; Hawkley & Cacioppo, 2003), or lonely individuals are less able to access social support in their networks (Uchino, 2004). Failure to find other interactions may be due to weaker associations for physical activity and fruit and vegetable intake. Alternatively, results may reflect truly independent pathways between social various dimensions of social engagement and risk behaviors. For example, social cohesion may influence behaviors by enforcing social norms for behaviors or increasing tangible support for particular behaviors (McNeill, Kreuter, & Subramanian, 2006), and this may not be affected by social support or loneliness.

Regardless of the reasons for these results, they suggest that behavioral interventions and other chronic disease prevention strategies might do well to consider both social support and neighborhood social cohesion. Behavior change researchers have called for more behavioral interventions that target social contexts, which are thought to be superior to those that simply target individuals because they address the setting in which behaviors are performed (Emmons, 2000). Few interventions have actually targeted social variables, other than the well-known Enhancing Recovery in Coronary Heart Disease trial (ENRICHD Investigators, 2000). Alternatively, health communication may be tailored to an individual’s social engagement (Kreuter & Wray, 2003), so that specific content in health messages vary according to an individual’s level of social support, for example. Despite the idea that such tailoring will improve the relevance of interventions, and some evidence of increased effectiveness (Noar, Benac, & Harris, 2007), very few socially tailored interventions have been tested. Results from this study add to the literature by suggesting that such interventions consider not only social support, but also neighborhood social cohesion.

It should be noted that associations between social engagement and behaviors represent just one pathway leading from social engagement to chronic disease. Psychosocial pathways, such as stress and depression, and physiologic pathways, including allostatic load and inflammatory burden, have also been hypothesized (Berkman & Krishna, 2014). Thus, the total association between social engagement and chronic disease is likely greater than the associations found in this study and further work is needed to elucidate those pathways.

Study Limitations and Strengths

Cross-sectional data prevent evaluating temporality and are amenable to selection/survival bias. Also, self-report may induce bias for physical activity (Prince et al., 2008), although self-reported smoking is generally valid (Patrick et al., 1994), and the food frequency questionnaire in this study demonstrated criterion validity for carbohydrate and fat intake by comparison with plasma lipids (Nettleton, Rock, Wang, Jenny, & Jacobs, 2009). Reliance on self-reported data for both the predictor and outcome may also have inflated estimates, due to same source bias. Due to the large sample, these limitations may have biased the results, resulting in statistically significant findings despite lack of true associations and statistically significant interaction terms may have been found by chance, due to the multiple comparisons tested. This study conceptualized social engagement as universally beneficial. However, engagement may also consist of negative aspects, including conflict, demands, and social role strain (Berkman & Krishna, 2014). Greater attention should be given in future work to the potential countervailing adverse effects of social engagement. Also, other social factors, such as social norms or social influence, were not measured in this study, but may influence associations found (Berkman & Krishna, 2014). This study is strengthened by addressing theoretically-driven hypotheses. Also, this study is strengthened by use of a large multi-ethnic sample, use of several distinctly different measures of social engagement, and clinically relevant outcomes.

Conclusions

In conclusion, our study found that social support was associated with smoking prevalence and smoking cessation and was more strongly associated with lower prevalence of smoking amongst non-lonely individuals, suggesting an interacting pathway between these two types of social engagement. However, neighborhood social cohesion, a less commonly studied dimension of social engagement, may be more relevant for physical activity and diet and may act via separate pathways.

Highlights.

  • Social engagement has been inconsistently associated with health behaviors

  • This study compared three dimensions of social engagement and tested interactions

  • Only social support was associated with smoking behaviors

  • However, neighborhood social cohesion was also relevant to health behaviors

  • Only one interaction was found, suggesting largely independent associations

Acknowledgments

This research was supported by grant 2R01 HL071759 (PI: Diez Roux) and by contracts N01-HC-95159 through N01-HC-95169 from the National Heart, Lung, and Blood Institute and by grants UL1-RR-024156 and UL1-RR-025005 from NCRR and grant F31NR012087 from the National Institute of Nursing Research. The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org. The authors also thank Dr. Wendy Post, who reviewed this manuscript, for her helpful comments.

Footnotes

1

Abbreviations: MESA: Multi-Ethnic Study of Atherosclerosis; MET: Metabolic equivalent

Conflict of Interest Statement

The authors declare that there are no conflicts of interest

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Laura J. Samuel, Email: lsamuel@jhmi.edu.

Cheryl R. Dennison Himmelfarb, Email: cdennis4@jhu.edu.

Moyses Szklo, Email: mszklo@jhsph.edu.

Teresa E. Seeman, Email: tseeman@mednet.ucla.edu.

Sandra E. Echeverria, Email: sandra.echeverria@rutgers.edu.

Reference List

  1. Ahern J, Galea S, Hubbard A, Syme SL. Neighborhood smoking norms modify the relation between collective efficacy and smoking behavior. Drug and Alcohol Dependence. 2009;100(1–2):138–145. doi: 10.1016/j.drugalcdep.2008.09.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, et al. Compendium of Physical Activities: an update of activity codes and MET intensities. Medicine and Science in Sports and Exercise. 2000;32(9):S498–S516. doi: 10.1097/00005768-200009001-00009. Retrieved from ISI:000089257400009. [DOI] [PubMed] [Google Scholar]
  3. Akaike H. New Look at Statistical-Model Identification. Ieee Transactions on Automatic Control. 1974;AC19(6):716–723. Retrieved from ISI:A1974U921700011. [Google Scholar]
  4. Ball K, Cleland VJ, Timperio AF, Salmon J, Giles-Corti B, Crawford DA. Love thy neighbour? Associations of social capital and crime with physical activity amongst women. Social Science & Medicine. 2010;71(4):807–814. doi: 10.1016/j.socscimed.2010.04.041. [DOI] [PubMed] [Google Scholar]
  5. Berkman LF, Krishna A. Social Network Epidemiology. In: Berkman LF, Kawachi I, Glymour MM, editors. Social Epidemiology. 2. Oxford: Oxford University Press; 2014. pp. 234–289. [Google Scholar]
  6. Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, et al. Multi-ethnic study of atherosclerosis: objectives and design. American Journal of Epidemiology. 2002;156(9):871–881. doi: 10.1093/aje/kwf113. [DOI] [PubMed] [Google Scholar]
  7. Block G, Woods M, Potosky A, Clifford C. Validation of a self-administered diet history questionnaire using multiple diet records. J Clin Epidemiol. 1990;43(12):1327–1335. doi: 10.1016/0895-4356(90)90099-B. Retrieved from PM:2254769. [DOI] [PubMed] [Google Scholar]
  8. Boomsma DI, Willemsen G, Dolan CV, Hawkley LC, Cacioppo JT. Genetic and environmental contributions to loneliness in adults: the Netherlands twin register study. Behavior Genetics. 2005;35(6):745–752. doi: 10.1007/s10519-005-6040-8. [DOI] [PubMed] [Google Scholar]
  9. Carpiano RM. Neighborhood social capital and adult health: an empirical test of a Bourdieu-based model. Health & Place. 2007;13(3):639–655. doi: 10.1016/j.healthplace.2006.09.001. [DOI] [PubMed] [Google Scholar]
  10. Chuang YC, Chuang KY. Gender differences in relationships between social capital and individual smoking and drinking behavior in Taiwan. Social Science & Medicine. 2008;67(8):1321–1330. doi: 10.1016/j.socscimed.2008.06.033. [DOI] [PubMed] [Google Scholar]
  11. Cleland V, Ball K, Hume C, Timperio A, King AC, Crawford D. Individual, social and environmental correlates of physical activity among women living in socioeconomically disadvantaged neighbourhoods. Social Science & Medicine. 2010;70(12):2011–2018. doi: 10.1016/j.socscimed.2010.02.028. [DOI] [PubMed] [Google Scholar]
  12. Cleveland WS, Mcgill R. Graphical Perception and Graphical Methods for Analyzing Scientific-Data. Science. 1985;229(4716):828–833. doi: 10.1126/science.229.4716.828. Retrieved from ISI:A1985APH2400009. [DOI] [PubMed] [Google Scholar]
  13. Cohen DA, Inagami S, Finch B. The built environment and collective efficacy. Health & Place. 2008;14(2):198–208. doi: 10.1016/j.healthplace.2007.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cohen S, Wills TA. Stress, social support, and the buffering hypothesis. Psychological Bulletin. 1985;98(2):310–357. doi: 10.1037/0033-2909.98.2.310. [DOI] [PubMed] [Google Scholar]
  15. Cradock AL, Kawachi I, Colditz GA, Gortmaker SL, Buka SL. Neighborhood social cohesion and youth participation in physical activity in Chicago. Social Science & Medicine. 2009;68(3):427–435. doi: 10.1016/j.socscimed.2008.10.028. [DOI] [PubMed] [Google Scholar]
  16. Debnam K, Holt CL, Clark EM, Roth DL, Southward P. Relationship between religious social support and general social support with health behaviors in a national sample of African Americans. Journal of Behavioral Medicine. 2012;35(2):179–189. doi: 10.1007/s10865-011-9338-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Delva J, Tellez M, Finlayson TL, Gretebeck KA, Siefert K, Williams DR, et al. Correlates of cigarette smoking among low-income African American women. Ethnicity & Disease. 2006;16(2):527–533. Retrieved from http://scholar.harvard.edu/files/davidrwilliams/files/2006-orrelates_of_cigarette-williams.pdf. [PubMed] [Google Scholar]
  18. Echeverria S, Diez-Roux AV, Shea S, Borrell LN, Jackson S. Associations of neighborhood problems and neighborhood social cohesion with mental health and health behaviors: the Multi-Ethnic Study of Atherosclerosis. Health & Place. 2008;14(4):853–865. doi: 10.1016/j.healthplace.2008.01.004. [DOI] [PubMed] [Google Scholar]
  19. Emmons KM. Health Behaviors in a Social Context. In: Berkman LF, Kawachi I, editors. Social Epidemiology. Oxford: Oxford University Press; 2000. pp. 242–266. [Google Scholar]
  20. ENRICHD Investigators. Enhancing recovery in coronary heart disease patients (ENRICHD): study design and methods. American Heart Journal. 2000;139(1 Pt 1):1–9. doi: 10.1016/s0002-8703(00)90301-6. Retrieved from http://www.ahjonline.com/article/S0002-8703%2800%2970001-9/fulltext. [DOI] [PubMed] [Google Scholar]
  21. Hawkley LC, Browne MW, Cacioppo JT. How can I connect with thee? Let me count the ways. Psychological Science. 2005;16(10):798–804. doi: 10.1111/j.1467-9280.2005.01617.x. [DOI] [PubMed] [Google Scholar]
  22. Hawkley LC, Cacioppo JT. Loneliness and pathways to disease. Brain, Behavior, and Immunity. 2003;17(Suppl 1):S98–105. doi: 10.1016/S0889-1591(02)00073-9. [DOI] [PubMed] [Google Scholar]
  23. Hawkley LC, Thisted RA, Cacioppo JT. Loneliness predicts reduced physical activity: cross-sectional & longitudinal analyses. Health Psychology. 2009;28(3):354–363. doi: 10.1037/a0014400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Holahan CJ, North RJ, Holahan CK, Hayes RB, Powers DA, Ockene JK. Social influences on smoking in middle-aged and older women. Psychology of Addictive Behaviors. 2011;26(3):519–526. doi: 10.1037/a0025843. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. House JS. Work Stress and Social Support. Reading, Massachusetts: Addison-Wesley; 1981. [Google Scholar]
  26. Kandula NR, Wen M, Jacobs EA, Lauderdale DS. Association between neighborhood context and smoking prevalence among Asian Americans. American Journal of Public Health. 2009;99(5):885–892. doi: 10.2105/AJPH.2007.131854. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kawachi I, Berkman LF. Social cohesion, social capital and health. In: Berkman LF, Kawachi I, editors. Social Epidemiology. Oxford: Oxford University Press; 2000. pp. 174–190. [Google Scholar]
  28. Kreuter MW, Wray RJ. Tailored and targeted health communication: strategies for enhancing information relevance. American Journal of Health Behavior. 2003;27(Suppl 3):S227–S232. doi: 10.5993/ajhb.27.1.s3.6. Retrieved from http://png.publisher.ingentaconnect.com/content/png/ajhb/2003/00000027/A00300s3/art00006. [DOI] [PubMed] [Google Scholar]
  29. Lauder W, Mummery K, Jones M, Caperchione C. A comparison of health behaviours in lonely and non-lonely populations. Psychology, Health & Medicine. 2006;11(2):233–245. doi: 10.1080/13548500500266607. [DOI] [PubMed] [Google Scholar]
  30. Li S, Horner P, Delva J. Social capital and cigarette smoking among Latinos in the United States. Substance Abuse & Rehabilitation. 2012;2012(3 Supplement 1):83–92. doi: 10.2147/SAR.S31164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. McNeill LH, Kreuter MW, Subramanian SV. Social environment and physical activity: a review of concepts and evidence. Social Science & Medicine. 2006;63(4):1011–1022. doi: 10.1016/j.socscimed.2006.03.012. [DOI] [PubMed] [Google Scholar]
  32. Mitchell PH, Powell L, Blumenthal J, Norten J, Ironson G, Pitula CR, et al. A short social support measure for patients recovering from myocardial infarction: the ENRICHD Social Support Inventory. Journal of Cardiopulmonary Rehabilitation. 2003;23(6):398–403. doi: 10.1097/00008483-200311000-00001. Retrieved from http://journals.lww.com/jcrjournal/Citation/2003/11000/A_Short_Social_Support_Measure_for_Patients.1.aspx. [DOI] [PubMed] [Google Scholar]
  33. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. Journal of the American Medical Association. 2004;291(10):1238–1245. doi: 10.1001/jama.291.10.1238. [DOI] [PubMed] [Google Scholar]
  34. Nettleton JA, Rock CL, Wang Y, Jenny NS, Jacobs DR. Associations between dietary macronutrient intake and plasma lipids demonstrate criterion performance of the Multi-Ethnic Study of Atherosclerosis (MESA) food-frequency questionnaire. British Journal of Nutrition. 2009;102(8):1220–1227. doi: 10.1017/S0007114509382161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nettleton JA, Steffen LM, Mayer-Davis EJ, Jenny NS, Jiang R, Herrington DM, et al. Dietary patterns are associated with biochemical markers of inflammation and endothelial activation in the Multi-Ethnic Study of Atherosclerosis (MESA) American Journal of Clinical Nutrition. 2006;83(6):1369–1379. doi: 10.1093/ajcn/83.6.1369. Retrieved from http://ajcn.nutrition.org/content/83/6/1369.long. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Noar SM, Benac CN, Harris MS. Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions. Psychological Bulletin. 2007;133(4):673–693. doi: 10.1037/0033-2909.133.4.673. [DOI] [PubMed] [Google Scholar]
  37. O’Donovan A, Hughes B. Social support and loneliness in college students: effects on pulse pressure reactivity to acute stress. International Journal of Adolescent Medicine & Health. 2007;19(4):523–528. doi: 10.1515/IJAMH.2007.19.4.523. [DOI] [PubMed] [Google Scholar]
  38. Pabayo R, Belsky J, Gauvin L, Curtis S. Do area characteristics predict change in moderate-to-vigorous physical activity from ages 11 to 15 years? Social Science & Medicine. 2010;72(3):430–438. doi: 10.1016/j.socscimed.2010.09.039. [DOI] [PubMed] [Google Scholar]
  39. Patrick DL, Cheadle A, Thompson DC, Diehr P, Koepsell T, Kinne S. The validity of self-reported smoking: a review and meta-analysis. American Journal of Public Health. 1994;84(7):1086–1093. doi: 10.2105/AJPH.84.7.1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Poortinga W. Do health behaviors mediate the association between social capital and health? Preventive Medicine. 2006a;43(6):488–493. doi: 10.1016/j.ypmed.2006.06.004. [DOI] [PubMed] [Google Scholar]
  41. Poortinga W. Perceptions of the environment, physical activity, and obesity. Social Science & Medicine. 2006b;63(11):2835–2846. doi: 10.1016/j.socscimed.2006.07.018. [DOI] [PubMed] [Google Scholar]
  42. Pressman SD, Cohen S, Miller GE, Barkin A, Rabin BS, Treanor JJ. Loneliness, social network size, and immune response to influenza vaccination in college freshmen. Health Psychology. 2005;24(3):297–306. doi: 10.1037/0278-6133.24.3.297. [DOI] [PubMed] [Google Scholar]
  43. Prince SA, Adamo KB, Hamel ME, Hardt J, Gorber SC, Tremblay M. A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. International Journal of Behavioral Nutrition & Physical Activity. 2008;5:56. doi: 10.1186/1479-5868-5-56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Russell D, Peplau LA, Cutrona CE. The revised UCLA Loneliness Scale: concurrent and discriminant validity evidence. Journal of Personality & Social Psychology. 1980;39(3):472–480. doi: 10.1037/0022-3514.39.3.472. [DOI] [PubMed] [Google Scholar]
  45. Sampson RJ, Raudenbush SW, Earls F. Neighborhoods and violent crime: a multilevel study of collective efficacy. Science. 1997;277(5328):918–924. doi: 10.1126/science.277.5328.918. [DOI] [PubMed] [Google Scholar]
  46. Shankar A, McMunn A, Banks J, Steptoe A. Loneliness, social isolation, and behavioral and biological health indicators in older adults. Health Psychology. 2011;30(4):377–385. doi: 10.1037/a0022826. [DOI] [PubMed] [Google Scholar]
  47. Shelton RC, McNeill LH, Puleo E, Wolin KY, Emmons KM, Bennett GG. The association between social factors and physical activity among low-income adults living in public housing. American Journal of Public Health. 2011;101(11):2102–2110. doi: 10.2105/AJPH.2010.196030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. StataCorp. Stata Statistical Software: Release 10. College Station, TX: StataCorp LP; 2007. [Google Scholar]
  49. U.S. Department of Agriculture, & U.S. Department of Health and Human Services. Dietary Guidelines for Americans, 2010. 7. Washington, DC: U.S. Government Printing Office; 2010. Retrieved from: http://www.cnpp.usda.gov/Publications/DietaryGuidelines/2010/PolicyDoc/PolicyDoc.pdf. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. U.S. Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans. Washington, DC: US Department of Health and Human Services; 2008. Retrieved from: http://www.health.gov/paguidelines/pdf/paguide.pdf. [Google Scholar]
  51. Uchino BN. Social Support & Physical Health: Understanding the Health Consequences of Relationships. New Haven, Connecticut: Yale University Press; 2004. [Google Scholar]
  52. Utter J, Denny S, Robinson E, Ameratunga S, Milfont TL. Social and physical contexts of schools and neighborhoods: associations with physical activity among young people in New Zealand. American Journal of Public Health. 2011;101(9):1690–1695. doi: 10.2105/AJPH.2011.300171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Vaananen A, Kouvonen A, Kivimaki M, Pentti J, Vahtera J. Social support, network heterogeneity, and smoking behavior in women: the 10-town study. American Journal of Health Promotion. 2008;22(4):246–255. doi: 10.4278/0701094R1.1. [DOI] [PubMed] [Google Scholar]
  54. Veitch J, van Stralen MM, Chinapaw MJ, te Velde SJ, Crawford D, Salmon J, et al. The neighborhood social environment and body mass index among youth: a mediation analysis. International Journal of Behavioral Nutrition & Physical Activity. 2012;9:31. doi: 10.1186/1479-5868-9-31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wacholder S. Binomial regression in GLIM: estimating risk ratios and risk differences. Am J Epidemiol. 1986;123(1):174–184. doi: 10.1093/oxfordjournals.aje.a114212. Retrieved from PM:3509965. [DOI] [PubMed] [Google Scholar]
  56. Weyers S, Dragano N, Mobus S, Beck EM, Stang A, Mohlenkamp S, et al. Poor social relations and adverse health behaviour: stronger associations in low socioeconomic groups? International Journal of Public Health. 2010;55(1):17–23. doi: 10.1007/s00038-009-0070-6. [DOI] [PubMed] [Google Scholar]
  57. Whitt MC, Levin S, Ainsworth BE, Dubose KD. Evaluation of a two-part survey item to assess moderate physical activity: the Cross-Cultural Activity Participation Study. J Womens Health (Larchmt) 2003;12(3):203–212. doi: 10.1089/154099903321667537. Retrieved from PM:12804350. [DOI] [PubMed] [Google Scholar]
  58. Yun EH, Kang YH, Lim MK, Oh JK, Son JM. The role of social support and social networks in smoking behavior among middle and older aged people in rural areas of South Korea: a cross-sectional study. BMC Public Health. 2010;10:78. doi: 10.1186/1471-2458-10-78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zou G. A modified poisson regression approach to prospective studies with binary data. American Journal of Epidemiology. 2004;159(7):702–706. doi: 10.1093/aje/kwh090. [DOI] [PubMed] [Google Scholar]

RESOURCES