Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 1.
Published in final edited form as: J Behav Med. 2020 Jan 1;43(2):155–165. doi: 10.1007/s10865-019-00131-3

Association of Social Network Factors with Weight Status and Weight Loss Intentions among Hispanic Adults

Mark L Wieland 1, Jane W Njeru 1, Janet M Okamoto 2, Paul J Novotny 3, Margaret K Breen-Lyles 2, Miriam Goodson 4, Graciela D Porraz Capetillo 4,5, Luz E Molina 4,5, Irene G Sia 1
PMCID: PMC7071972  NIHMSID: NIHMS1547663  PMID: 31894451

Abstract

Hispanic adults have the highest obesity prevalence in the United States, but little is known about weight-related social network influences. A community-based sample of 610 Hispanic participants completed height/weight and a survey. The proportion of overweight or obese (OW/OB) network members was higher for OW/OB respondents compared to normal weight respondents. Participants with high weight loss intentions reported more positive social norms for weight control, social support, and social cohesion. If most or all of OW/OB participant’s social contacts were trying to lose weight, the odds that they were likely to try to lose weight was four times higher than other participants. The relationship between weight loss intentions and number of social contacts trying to lose weight was strongly mediated by social norms for weight control and social support. These results suggest that social contacts and functional network characteristics may impact weight status and weight control intentions among Hispanic adults.

Keywords: Social network, Hispanic, Obesity, Community-Based Participatory Research

INTRODUCTION

In the midst of a well-documented obesity epidemic in the United States, Hispanic adults are 1.2 times more likely to be obese than non-Hispanic whites, representing the highest age-adjusted obesity prevalence in the country (National, Center, for, Health, & Statistics, 2017). Dietary behaviors and physical activity are associated with obesity (Hall et al., 2011; Seo & Li, 2010), and these behaviors are less healthful for Hispanic Americans (Neighbors et al., 2008; Rehm et al., 2016). The mechanisms of these disparities include a complex web of social, environmental, and behavioral factors (Williams et al., 2011).

Social network characteristics are highly associated with obesity, which has been shown to cluster according to social ties (Christakis & Fowler, 2007). Obesogenic behaviors likewise tend to cluster by social groups (Fletcher et al., 2011). Accordingly, social network structures have been shown to be significant contributors to the obesity epidemic (Zhang et al., 2018) (Hammond, 2010). Indeed, simulation models suggest that traditional weight loss interventions frequently fail because they lack consideration of the participant’s surrounding social networks (Bahr et al., 2009).

The mechanisms of social network effects on obesity can be conceived as both structural and functional (Shelton et al., 2019). Structural network factors (e.g., size, composition, etc.) may influence obesity-related behaviors through mirroring the behaviors of others (Befort et al., 2008; Pachuck et al., 2011), self-comparison (Shakya et al., 2015) and direct influence of peers to adopt similar behaviors (Rancourt et al., 2015). Functional network factors may shape obesity-related health behaviors through social norms (Leahey et al., 2015; Leahey et al., 2011), social support (Marcoux et al., 1990), and social cohesion (Hystad & Carpiano, 2012), among other constructs (Powell et al., 2015).

Social norms are series of beliefs about socially acceptable behavior (Cialdini & Trost, 1998). In a study of social norms for obesity among a sample of young adults, Leahey and colleagues found that injunctive norms (beliefs about what is socially acceptable in general) did not mediate the relationship between weight status of participants and their social network (Leahey et al., 2011). However, among participants who were overweight or obese, subjective norms (beliefs about what is socially acceptable among close contacts) strongly mediated the relationship between participant’s weight loss intentions and number of social contacts trying to lose weight, suggesting that norms exhibited within a social network may influence obesogenic behaviors more than societal norms in general (Leahey et al., 2011). The association between social norms and weight status or weight loss intentions among Hispanic populations and their networks has not been previously investigated.

Depending on the context, members of social networks may or may not engage in social support (Heany & Israel, 2008), which is a functionally positive relational interaction that can manifest in a variety of ways, i.e., emotional support (love, empathy, etc.), instrumental support (direct assistance), informational support (advice, information, etc.), and appraisal support (feedback, praise, etc.) (House et al., 1988). Social support is an important mechanism for improved efficacy of dietary and physical activity interventions (Greaves et al., 2011), as well as for weight loss interventions carried out within social networks (Marcoux et al., 1990) (Wing & Jeffery, 1999).

Social cohesion is an established network-level metric (Valente, 2010) that has been used to assess network intervention implementation progress (Feinberg et al., 2005; Gesell et al., 2013). Higher social cohesion, i.e., close relationships among community/group members with strong mutual trust and reciprocity (Kawachi, 2008), has been associated with a range of positive health outcomes (Inoue et al., 2013; Subramanian et al., 2002; Wen et al., 2005).

Among Hispanic populations in the United States, less is known about social influences on the obesity epidemic. In a study of 107 Hispanic women enrolled in a lifestyle intervention program, participants and their social ties shared similarities in weight control behaviors and weight change (Marquez et al., 2018). Marquez and colleagues demonstrated minimal association between structural network characteristics of size and composition (i.e., proportion of network that was Latino, Spanish speaking, etc.) and health promoting behaviors among a Hispanic population (Marquez et al., 2014), but in a different study, they found a significant association between functional network factors (social support for physical activity) and weight loss (Marquez et al., 2016).

Social network data may be used to inform social network interventions, which involve purposeful utilization of existing social networks to promote positive behavior change and health outcomes (Valente, 2012). One difficulty in designing interventions to promote weight loss is that the reasons for sub-optimal behaviors among Hispanic populations are multiple and complex (Malmusi et al., 2010). Community-based participatory research (CBPR) is a means to collaboratively investigate health topics, whereby community members and academics partner in an equitable relationship through all phases of research and intervention development (Israel et al., 1998). This approach is well suited to intervention work that addresses the interplay between health behaviors and the social determinants of health such that it empowers communities, promotes understanding of culturally pertinent issues, and targets the multi-faceted barriers to health (Wallerstein & Duran, 2010). Rochester Healthy Community Partnership (RHCP) is a CBPR partnership in Southeast Minnesota with a mission to promote health and wellbeing in the Rochester (MN) population through CBPR, education, and civic engagement to promote health equity (Rochester Healthy Community Partnership, 2019). RHCP partners previously developed and tested a nutrition and physical activity intervention with Hispanic families through a CBPR approach that demonstrated improvements in dietary quality (Wieland et al., 2018). Because more than 80% of participants in that study were overweight or obese, RHCP partners turned their attention to adaptation of the existing intervention for weight loss within social networks. The social network analysis presented here was conducted by community and academic partners in order to inform development of a social network intervention.

The aims and hypotheses of the study focus on the application of social network constructs to a large community sample of Hispanic adults. Aims were to 1) assess whether weight status clusters within social networks among Hispanic adults, and to 2) assess structural and functional social network characteristics associated with weight loss intentions among overweight and obese Hispanic adults. The hypotheses were as follows: 1) there is an association between weight status of Hispanic study participants and their social networks; 2) social norms for obesity mediate the relationship between weight status of participants and weight status of their social networks; 3) there is an association between weight loss intentions of overweight or obese study participants and weight loss intentions of members of their social networks; 4) social norms for weight loss, social support for healthy behaviors, and social cohesion mediate the relationship between weight loss intentions of participants and weight loss intentions of their social network.

METHODS

Setting and Participants

Recruitment was conducted throughout a medium size city in Southeast Minnesota. Eligibility criteria were 1) age 18 or greater, 2) residence in the metro area, and 3) self-identification as Hispanic or Latino. Study participants were recruited by bilingual study team members (who were also RHCP community partners) through multiple approaches, including the hosting of community events, collaboration in existing community events, clinics that serve majority Hispanic populations, collaboration with existing institutions (e.g., churches with large proportion of Hispanic congregants), previous study participants, word-of-mouth, and referrals from community-based organizations and individuals.

Data Collection

Data were collected by bilingual study team members at various community locations (similar to recruitment venues). Surveys were conducted as a face-to-face structured interview. Height and weight were measured on site with portable equipment. Written informed consent was provided by each participant, and all study procedures were approved by the Mayo Clinic Institutional Review Board.

Demographic and general health measures

Study participants reported on the following demographic data: age, gender, ethnicity, country of birth, annual household income, education level, employment status, years lived in the US, primary language spoken at home, and level of English language proficiency on a 5-point Likert scale (Flores et al., 2005).

Structural network measures

Participants were asked to name up to 10 members of their social network, adults with whom they discuss important issues in their life (i.e., discussion network). Names and phone numbers were collected for each named alter to verify identities only (to clarify between individuals with the same names), not for the purposes of contacting alters. Respondents were instructed to report their perception of the weight status for each named alter as underweight, normal weight, overweight, or very overweight (obese). All respondents and named alters were assigned a numeric study ID, such that an anonymized dataset was used for analysis. Network size, or the number of people a respondent had a reported tie with, was calculated for each participant.

Functional network measures

Social norms for obesity and obesity-related behaviors were measured with an existing instrument on a 5-point Likert scale of acceptability (Leahey et al., 2011): How socially acceptable it is to be overweight? How socially acceptable is it to eat unhealthy foods? How acceptable is it to be physically inactive?

Social cohesion was measured using an existing instrument that has been used in network intervention studies (Gesell et al., 2013). The 3-item measure reflects the dimension of cohesion related to feeling a sense of belonging on a 7-point Likert scale of agreement: I feel a sense of belonging to my community (however defined by the participant); I feel that I am a member of my community; I see myself as part of my community (Chin et al., 1999).

Social support for eating a healthy diet was assessed by adaptation of a 5-item instrument on a 5-point Likert scale (Sallis et al., 1987). It asked participants how often over the last 30 days family or friends have encouraged them to eat healthy foods, discussed the benefits of eating healthy foods, reminded them to choose healthy foods, shared ideas on how to eat healthy foods, and eaten healthy foods with them. Response options were: never, once in a while, sometimes, often, or always.

For obese and overweight participants only, the following constructs were adapted from Leahey, et al. (Leahey et al., 2011): Weight loss intentions were assessed with the following item on a 5-point Likert scale of likeliness: How likely are you to try to lose weight within the next 3-months? Number of social contacts trying to lose weight was assessed on a 5-point Likert scale from 0 (nobody) to 4 (all). Social norms for weight loss were assessed by a 3-item instrument on a 5-point Likert scale from 0 (never/strongly disapprove) to 4 (always/strongly approve): How frequently do your family and/or friends encourage you to lose weight? How frequently do your family and/or friends offer you weight loss information or tools to help with weight loss? To what extent would people closest to you approve or disapprove if you were to lose weight within the next 3-months?

Most survey measures had existing Spanish language translations with good validity evidence. For those items that did not have Spanish language versions in the literature, we edited the English-language version of each item with community partners, followed by forward-translation, panel discussion, backward translation, a pre-test, a cognitive briefing and a consensus on the final version by a core group of community members (World Health Organization, 2007). We have previously described this process of adapting the World Health Organization translation procedure for use with survey instruments in a CBPR framework (Formea et al., 2014).

In our sample, the social norms for obesity instrument demonstrated fair internal consistency (Cronbach alpha 0.63) with individual item loading ranging from 0.70 to 0.78. The social cohesion instrument demonstrated excellent internal consistency (Cronbach alpha 0.96) with individual item loading ranging from 0.94 to 0.97. The social support instrument demonstrated good internal consistency (Cronbach alpha 0.89) with individual item loading ranging from 0.78 to 0.83. The social norms for weight loss instrument demonstrated good internal consistency (Cronbach alpha 0.78) with individual item loading ranging from 0.71 to 0.89. On bivariate analysis, social support was correlated with social norms for weight loss (correlation of 0.38, p<0.0001), social cohesion (correlation of 0.18, p<0.0001), and social norms for obesity (correlation of −0.11, p=0.005). There were no correlations between any of the other three functional network characteristics.

Biometric measures

Weight was measured to the nearest 0.1 kg using a portable scale (Seca 880 Digital Floor Scale). Height was measured to the nearest 0.1 cm using a stadiometer. Body mass index (BMI) was calculated as weight (kg)/height squared (m2).

Data Analysis

Descriptive statistics were used to report the survey and biometric data. Logistic models were used to describe the basic associations of participant and network demographics with binary characteristics such as being overweight or weight loss intentions. Estimation of causal mediation analysis effects used the method described by VanderWeele (VanderWeele, 2014). Associations were assessed using logistic models with no adjustments for other covariates. Analyses were performed using SAS version 9.2 software package and, in particular, used the SAS CAUSALMED procedure for mediation analysis, which combines several logistic regression models to determine mediation effects.

Network data was analyzed to identify nodes and ties and to visualize the network. The study assessed egocentric, or personal networks, of participants. In addition, network members were linked across individuals through matching names and phone numbers. This hybrid approach of compiling egocentric network data to form a larger sociocentric network, was used in order to visualize this real-world community network. Ties between nodes were all considered as directed ties. Where names and phone numbers of named network members did not match, these were considered as different individuals in the analysis, so that any error would be an under-estimate of position and importance in the overall network.

Edgelists of egos and alters were created, and ego attribute data based on survey responses was compiled and imported into statistical and network analytic software. Stata (StataCorp, 2015), UciNet (Borgatti, 2002), Gephi (Bastian et al., 2009), and R (Handcock, 2003; Team, 2017) were used to calculate network metrics (Stata, UciNet, Gephi), conduct more complex network analysis (UciNet, R), and to visualize social networks (Gephi). Though a hybrid approach was used to build and visualize a larger community network, the measures used in the analysis were egocentric in nature (e.g. percentage of personal network that is overweight).

For network members for whom there was no survey data, the weight status variable was determined by the person identifying them as a social tie. If there were survey data and a BMI measure for a network member, then BMI from their survey was used to identify them as overweight/obese. If the network member was named by multiple survey respondents, consensus on overweight status was determined by the average of all reports. Those reported as overweight or very overweight were considered as overweight for the analysis. An average of overweight network members was calculated for each survey respondent that reported a network tie. The average was calculated based on the number of overweight personal network members divided by the total number of reported personal network members. Network size was calculated by taking a count of how many network members a respondent named.

RESULTS

Participant and network characteristics

A total of 610 Hispanic adults completed the survey and biometric measurements. Mean age was 40 (±13) years, 57% (342) were female, and 81% (457) of participants were overweight or obese (Table 1). There were 1,998 named nodes in the network with 2,354 ties between these nodes. Among named alters, only 32% were perceived to be overweight or obese. Full demographics, weight status, and social network measures of the study population are shown in Table 1. A graphical depiction of the social network according to weight status is shown in Figure 1.

Table 1.

Participant Demographics and Survey Responses (N=610)

Measurement
Age, mean (SD) 40 (13)
Female gender, n (%) 342 (57%)
Formal schooling completed, n (%)
 8 grades or less 208 (35%)
 Some high school 96 (16%)
 High school or GED completed 198 (33%)
 College or advanced degree 100 (17%)
Annual family income, n (%)
 $0 to $9,999 116 (20%)
 $10,000 to $19,999 109 (19%)
 $20,000 to $29,000 117 (20%)
 $30,000 to $49,999 143 (24%)
 $50,000 or higher 100 (17%)
Health insurance in last 12 months, n (%)
 Yes 324 (54%)
 No 281 (46%)
Country of Birth, n (%)
 United States 93 (15%)
 Mexico 421 (70%)
 Other Latin America 89 (15%)
English language proficiency, n (%)
 Not at all 80 (13%)
 Not very well 213 (35%)
 Well 148 (24%)
 Very well 166 (27%)
Body Mass Index, mean (SD) 29.8 (5.8)
 Normal weight, n (%) 100 (19%)
 Overweight, n (%) 208 (37%)
 Obese, n (%) 249 (44%)
Social support for healthy diet, mean (SD)1 3.4 (0.9)
Social cohesion, mean (SD) 2 3.4 (1.4)
Social norms for obesogenic behaviors, mean (SD) 1 2.9 (0.9)
Social norms for weight loss, mean (SD)1 2.1 (1)
How likely are you to try to lose weight in the next 3 months?, n (%)
 Very unlikely or unlikely 98 (18%)
 Neither likely nor unlikely 55 (10%)
 Very likely or likely 371 (70%)
Open in a new tab
1

Reported as a 5-point Likert scale, with 1 being the lowest response option and 5 the highest

2

Reported as a 7-point Likert scale, with 1 being the lowest response option and 7 the highest

Figure 1.

Figure 1.

Open in a new tab

Social network diagram of participants and their network ties according to weight status

Association between weight status of participants and weight status of their social networks

The mean percentage of overweight and obese network members was significantly higher among overweight and obese participants than for normal weight respondents (33.7% vs. 28.7%, P=<0.001). This percentage was also higher for obese respondents than for overweight respondents (36.5% vs 30.5%, p=0.04). Overall findings and results by age and gender are shown in Figure 2.

Figure 2.

Figure 2.

Open in a new tab

Association between participants being overweight or obese and number of overweight network members

Mediation analysis for association between weight status of participants and weight status of their social networks

Consistent with above, there was a positive association between participant BMI and perceived weight status of their social network members. The estimated direct effect odds ratio for the association between being overweight or obese and the percent of their social network being overweight was 2.3 (P=0.007).

Overweight and obese participants had a smaller mean (SD) network size than normal weight participants [4.4 (2.0) vs. 4.9 (2.1), P=0.02]. Network size did not mediate the relationship between weight status of participants and the weight status of their social networks.

There was no correlation between participant BMI and social norms for obesity (Pearson correlation 0.03, p=0.50). Likewise, there was no correlation between overweight social network members and social norms for obesity (Pearson correlation 0.06, p=0.17). This resulted in social norms for obesity only mediating 0.1% of the total effect of the weight status of their social network members. This percentage was not significantly different than zero (p=0.94). Similarly, social support and social cohesion did not mediate the relationship between weight status of participants and their social networks because neither variable was significantly associated with being overweight.

Association between weight loss intentions among overweight/obese participants (n=457) and social network influences

If most or all of overweight/obese participant’s social contacts were trying to lose weight (29.3% of respondents), the odds of stating they were likely or very likely to try to lose weight was four times higher compared to other participants (OR 3.99, 95% CI 2.2–7.2). Overall findings and results by age and gender are shown in Figure 3.

Figure 3.

Figure 3.

Open in a new tab

Association between weight loss intentions and network members trying to lose weight among participants who were overweight or obese

Participants who stated that they were likely or very likely to lose weight reported more positive (re-enforcing) social norms for weight loss than participants who were not likely to lose weight [mean (SD) 2.4 (0.8) vs. 1.5 (0.8), P=<0.0001].

Participants who stated that they were likely or very likely to lose weight reported more positive social support for eating a healthy diet than participants who were not likely to lose weight [mean (SD) 3.2 (1.1) vs. 2.6 (1.2), P=<0.0001].

Participants who stated that they were likely or very likely to lose weight reported more social cohesion than participants who were not likely to lose weight [mean (SD) 3.3 (1.4) vs. 3.0 (1.3), P=0.008].

There was no significant difference in obesogenic social norms between participants who stated that they were likely or very likely to lose weight and those who were not likely to lose weight [mean (SD) 2.9 (3.05) vs. 2.8 (3.02), P=0.25].

Association between social contacts trying to lose weight among overweight/obese participants and social network influences

Participants who stated that most or all of their social contacts were trying to lose weight reported more positive (re-enforcing) social norms for weight loss than participants whose social contacts were not trying to lose weight [mean (SD) 2.8 (0.8) vs. 2 (0.9), P=<0.0001].

Participants who stated that most or all of their social contacts were trying to lose weight reported more positive social support for eating a healthy diet than participants whose social contacts were not trying to lose weight [mean (SD) 3.5 (1.0) vs. 2.8 (1.1), P=<0.0001].

Participants who stated that most or all of their social contacts were trying to lose weight reported more positive social cohesion than participants whose social contacts were not trying to lose weight [mean (SD) 3.4 (1.4) vs. 3.2 (1.3), P=0.04].

There was no significant difference in obesogenic social norms between participants who stated that most or all of their social contacts were trying to lose weight and those who were not trying to lose weight [mean (SD) 2.9 (3.15) vs. 2.8 (3.00), P=0.76].

Mediation analysis for association between weight loss intentions among overweight/obese participants and number of social contacts trying to lose weight

Consistent with above, there was a positive correlation between weight loss intentions and number of social contacts trying to lose weight. The estimated direct effect odds ratio for the association between intention to lose weight and having most or all of their social contacts trying to lose weight was 4.60 (P=0.004).

There was a significant association between social norms for weight loss and both intention to lose weight (P<0.0001) and having most or all of their social contacts trying to lose weight (P<0.0001). This resulted in social norms for weight loss explaining 81% of the association between weight loss intentions and social contacts trying to lose weight. This proportion was significantly different from zero (p<0.0001).

There was a significant association between social support and both intention to lose weight (P<0.0001) and having most or all of their social contacts trying to lose weight (P<0.0001). This resulted in social support explaining 31% of the association between weight loss intentions and social contacts trying to lose weight. This proportion was significantly different from zero (p<0.0001).

There was a significant association between social cohesion and intention to lose weight (P=0.02) but not having most or all of their social contacts trying to lose weight (P=0.07). Social cohesion explained 6% of the association between weight loss intentions and social contacts trying to lose weight, but this effect was not significant (P=0.15).

DISCUSSION

The results of this study demonstrated that overweight and obesity cluster by social networks within a large community-based sample of Hispanic adults in the Midwest United States. According to the 2007–2011 American Community Survey, the achieved sample size represented more than 30% of Hispanic adults residing in the city. These findings extend the results of previous studies (Christakis & Fowler, 2007) to Hispanic populations, where overweight and obesity prevalence is very high. The discordance between the percentage overweight or obese among measured respondents (81%) compared to weight status assigned by respondents to members of their social network (32% overweight/obese) reflects a well-documented underreporting bias (Connor Gorber et al., 2007), which would skew results towards the null, thereby strengthening the basis of the conclusion that overweight and obesity clusters by social networks within this population.

While the “direction” of this association was consistent across age ranges and genders, the association was stronger among women and younger participants. In sub-group analysis, we found that men were less likely to consider themselves overweight or obese compared with women. This gap between perceived and actual BMI may also be reflected in male participant’s perception of weight status for their social network members, thereby blunting the actual association. The subgroup analysis also found that younger participants reported higher social norms for obesity, which may act as a stronger mechanistic foundation for the association between participant and network weight status.

Among participants who were overweight or obese, there was a strong association between weight loss intentions and social network factors: If most or all of their social contacts were trying to lose weight, participants were 4 times more likely to report high weight loss intentions than participants with fewer contacts trying to lose weight. This finding is consistent with previous studies that have consistently demonstrated the importance of social ties for promotion of health behavior change in general (Christakis & Fowler, 2008; Rosenquist et al., 2010), and for weight control among a Hispanic group in particular (Marquez et al., 2018).

The findings that overweight and obese participants with a higher proportion of social contacts trying to lose weight had higher weight loss intentions, and that this association was potentially mediated by social norms for weight loss (subjective norms), replicated the findings by Leahey and colleagues among young adults who were predominantly white (Leahey et al., 2011). The current study extends those findings to a community-based sample of Hispanic adults. This is important because weight loss intentions are associated with weight loss (Schifter & Ajzen, 1985), and because social norms for weight loss have been shown to be associated with the degree of actual weight loss among participants in a weight control program (Leahey et al., 2015).

The current study also adds to the literature by demonstrating a significant contribution of other functional network characteristics (social support and social cohesion) on weight loss intentions within Hispanic populations. Previous research has demonstrated an association between social support for physical activity and weight loss among Hispanic participants in a weight loss program (Marquez et al., 2016). The current study adds to the foundation of this association by demonstrating that social support for healthy eating may play a mediating role in the relationship between participant’s weight loss intentions and the number of their social contacts trying to lose weight. Likewise, social cohesion was strongly associated with weight loss intentions and number of social contacts trying to lose weight, which is important because of the strong association between group social cohesion and health promoting behaviors (Beal et al., 2003). Network size, while slightly smaller for overweight/obese networks, did not mediate the relationship between weight status of participants and the weight status of their social networks.

This is consistent with previous literature that demonstrated very little association between structural network factors and health behaviors among Hispanic networks (Marquez et al., 2014).

Taken as a whole, these findings suggest that future weight loss interventions among Hispanic populations may be more successful if social network factors are addressed. In particular, promotion of positive subjective social norms for weight control within networks of predominantly overweight and obese individuals may be beneficial. Likewise, intervention components that facilitate positive social support and social cohesion within networks may compound efficacy of weight loss interventions.

Strengths of the study include a large, community-based sample of Hispanic adults who were recruited through diverse mechanisms by community partners, and the assessment of a wide range of network constructs. However, the study has significant limitations. Potentially important structural network characteristics (e.g., modeling) were not assessed, and reciprocation of network ties were not delineated. Negative network effects (e.g., weight stigma in the context of social norms) were also not assessed. Importantly, the cross-sectional study design precludes the ability to assign causality to the associations discovered. While BMI was calculated for study respondents, data for named alters relied on perceived weight status. Weight loss intentions were measured rather than actual changes in weight status, and while there is evidence that weight loss intentions may predict actual weight loss (Schifter & Ajzen, 1985), this literature is mixed (Rancourt et al., 2017). Likewise, participant perceptions of weight loss intention within their social networks may represent a “false consensus” effect, whereby participants perceive themselves as behaving in more similar ways to their alters than is true (Suls et al., 1988). However, it is not clear that perception is less important than reality as it relates to associated behavioral decisions by the ego or health outcomes (Duncan et al., 2011; Steffens et al., 1996). The social cohesion instrument used for this study defines “community” through the lens of the participant, the contextual nature of which may extend beyond network-related constructs. Finally, findings may not be generalizable to other cities or to more rural areas.

CONCLUSION

These results indicate that social contacts and functional network characteristics are associated with weight status and weight control intentions among Hispanic adults, thereby highlighting the importance of addressing social influence in programs aimed at the reduction of obesity-related health disparities among these high risk groups.

Acknowledgments:

The authors thank all RHCP partners who contributed to the organization, implementation, and dissemination of this work.

Funding Sources: This publication was supported by NIH Grant No. R01 HL 111407 from the National Heart, Lung, and Blood Institute and by CTSA Grant No. UL1 TR000135 from the National Center for Advancing Translational Science (NCATS), and by the Mayo Clinic Office of Health Disparities Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. The funding bodies had no role in study design; in the collection, analysis, and interpretation of data; writing of the manuscript; and in the decision to submit the manuscript for publication.

Footnotes

Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.

Conflicts of Interest: The authors declared no conflict of interest.

Human Rights: All procedures performed in this study were approved by the Mayo Clinic Institutional Review Board, and are in accordance with the 1964 Helsinki Declaration and its later amendments.

Informed Consent: Informed consent was obtained from all individual participants included in the study.

Welfare of Animals: This article does not contain any studies with animals performed by any of the authors.

REFERENCES

  1. Bahr DB, Browning RC, Wyatt HR, & Hill JO (2009). Exploiting social networks to mitigate the obesity epidemic. Obesity, 17(4), 723–728. doi: 10.1038/oby.2008.615 [DOI] [PubMed] [Google Scholar]
  2. Bastian M, Heymann S, & Jacomy M (2009). Gephi: an open source software for exploring and manipulating networks. Paper presented at the International AAAI Conference on Weblogs and Social Media. [Google Scholar]
  3. Beal DJ, Cohen RR, Burke MJ, & McLendon CL (2003). Cohesion and performance in groups: a meta-analytic clarification of construct relations. J Appl Psychol, 88(6), 989–1004. doi: 10.1037/0021-9010.88.6.989 [DOI] [PubMed] [Google Scholar]
  4. Befort CA, Thomas JL, Daley CM, Rhode PC, & Ahluwalia JS (2008). Perceptions and beliefs about body size, weight, and weight loss among obese African American women: a qualitative inquiry. Health Educ Behav, 35(3), 410–426. doi: 10.1177/1090198106290398 [DOI] [PubMed] [Google Scholar]
  5. Borgatti SP, Everett MG, Freeman LC. (2002). Ucinet for Windows: Software for Social Network Analysis. In. Cambridge, MA: Analytic Technologies. [Google Scholar]
  6. Chin WW, Salisbury DW, Pearson AW, & MJ. S (1999). Perceived cohesion in small groups: adapting and testing the perceived cohesion scale in a small group setting. Small Group Research(30), 751–766. [Google Scholar]
  7. Christakis NA, & Fowler JH (2007). The spread of obesity in a large social network over 32 years. The New England journal of medicine, 357(4), 370–379. doi: 10.1056/NEJMsa066082 [DOI] [PubMed] [Google Scholar]
  8. Christakis NA, & Fowler JH (2008). The collective dynamics of smoking in a large social network. N Engl J Med, 358(21), 2249–2258. doi: 10.1056/NEJMsa0706154 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cialdini RB, & Trost MR (1998). Social influence: Social norms, conformity, and compliance In Gilbert DT, Fiske ST, & Kindzey G (Eds.), The handbook of social psychology (Vol. 2). New York: McGraw-Hill. [Google Scholar]
  10. Connor Gorber S, Tremblay M, Moher D, & Gorber B (2007). A comparison of direct vs. self-report measures for assessing height, weight and body mass index: a systematic review. Obes Rev, 8(4), 307–326. doi: 10.1111/j.1467-789X.2007.00347.x [DOI] [PubMed] [Google Scholar]
  11. Duncan DT, Wolin KY, Scharoun-Lee M, Ding EL, Warner ET, & Bennett GG (2011). Does perception equal reality? Weight misperception in relation to weight-related attitudes and behaviors among overweight and obese US adults. Int J Behav Nutr Phys Act, 8, 20. doi: 10.1186/1479-5868-8-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Feinberg ME, Riggs NR, & Greenberg MT (2005). Social networks and community prevention coalitions. The journal of primary prevention, 26(4), 279–298. doi: 10.1007/s10935-005-5390-4 [DOI] [PubMed] [Google Scholar]
  13. Fletcher A, Bonell C, & Sorhaindo A (2011). You are what your friends eat: systematic review of social network analyses of young people’s eating behaviours and bodyweight. Journal of epidemiology and community health, 65(6), 548–555. doi: 10.1136/jech.2010.113936 [DOI] [PubMed] [Google Scholar]
  14. Flores G, Abreu M, & Tomany-Korman SC (2005). Limited english proficiency, primary language at home, and disparities in children’s health care: how language barriers are measured matters. Public Health Rep, 120(4), 418–430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Formea CM, Mohamed AA, Hassan A, Osman A, Weis JA, Sia IG, & Wieland ML (2014). Lessons learned: cultural and linguistic enhancement of surveys through community-based participatory research. Progress in community health partnerships : research, education, and action, 8(3), 331–336. doi: 10.1353/cpr.2014.0037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gesell SB, Barkin SL, & Valente TW (2013). Social network diagnostics: a tool for monitoring group interventions. Implementation science : IS, 8, 116. doi: 10.1186/1748-5908-8-116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Greaves CJ, Sheppard KE, Abraham C, Hardeman W, Roden M, Evans PH, … Group IS (2011). Systematic review of reviews of intervention components associated with increased effectiveness in dietary and physical activity interventions. BMC Public Health, 11, 119. doi: 10.1186/1471-2458-11-119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL, & Swinburn BA (2011). Quantification of the effect of energy imbalance on bodyweight. Lancet, 378(9793), 826–837. doi: 10.1016/S0140-6736(11)60812-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hammond RA (2010). Social influence and obesity. Current opinion in endocrinology, diabetes, and obesity, 17(5), 467–471. doi: 10.1097/MED.0b013e32833d4687 [DOI] [PubMed] [Google Scholar]
  20. Handcock MS, Hunter DR, Butts CT, Goodreau SM, Morris M. (2003). statnet: Software tools for the Statistical Modeling of Network Data. Retrieved from http://statnetproject.org [DOI] [PMC free article] [PubMed]
  21. Heany CA, & Israel BA (2008). Social networks and social support In Glanz K, Rimer BK, & Viswanath K (Eds.), Health behavior and health education (pp. 189–208). San Francisco: Jossey-Bass. [Google Scholar]
  22. House JS, Umberson D, & Landis KR (1988). Structures and processes of social support. Annual Review of Sociology, 14, 293–318. [Google Scholar]
  23. Hystad P, & Carpiano RM (2012). Sense of community-belonging and health-behaviour change in Canada. Journal of epidemiology and community health, 66(3), 277–283. doi: 10.1136/jech.2009.103556 [DOI] [PubMed] [Google Scholar]
  24. Inoue S, Yorifuji T, Takao S, Doi H, & Kawachi I (2013). Social cohesion and mortality: a survival analysis of older adults in Japan. American journal of public health, 103(12), e60–66. doi: 10.2105/AJPH.2013.301311 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Israel BA, Schulz AJ, Parker EA, & Becker AB (1998). Review of community-based research: assessing partnership approaches to improve public health. Annu Rev Public Health, 19, 173–202. [DOI] [PubMed] [Google Scholar]
  26. Kawachi IS, S.V.; Daniel, K (2008). Social Capital and Health. New York, NY: Springer. [Google Scholar]
  27. Leahey TM, Doyle CY, Xu X, Bihuniak J, & Wing RR (2015). Social networks and social norms are associated with obesity treatment outcomes. Obesity (Silver Spring), 23(8), 1550–1554. doi: 10.1002/oby.21074 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Leahey TM, Gokee LaRose J, Fava JL, & Wing RR (2011). Social influences are associated with BMI and weight loss intentions in young adults. Obesity, 19(6), 1157–1162. doi: 10.1038/oby.2010.301 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Malmusi D, Borrell C, & Benach J (2010). Migration-related health inequalities: showing the complex interactions between gender, social class and place of origin. Soc Sci Med, 71(9), 1610–1619. doi:S0277-9536(10)00629-5 [pii] 10.1016/j.socscimed.2010.07.043 [DOI] [PubMed] [Google Scholar]
  30. Marcoux BC, Trenkner LL, & Rosenstock IM (1990). Social networks and social support in weight loss. Patient Educ Couns, 15(3), 229–238. [Google Scholar]
  31. Marquez B, Anderson A, Wing RR, West DS, Newton RL, Meacham M, … Look ARG (2016). The relationship of social support with treatment adherence and weight loss in Latinos with type 2 diabetes. Obesity (Silver Spring), 24(3), 568–575. doi: 10.1002/oby.21382 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Marquez B, Elder JP, Arredondo EM, Madanat H, Ji M, & Ayala GX (2014). Social network characteristics associated with health promoting behaviors among Latinos. Health Psychol, 33(6), 544–553. doi: 10.1037/hea0000092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Marquez B, Norman GJ, Fowler JH, Gans KM, & Marcus BH (2018). Weight and weight control behaviors of Latinas and their social ties. Health Psychol, 37(4), 318–325. doi: 10.1037/hea0000597 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. National, Center, for, Health, & Statistics. (2017). Health, United States, 2016. Hyattsville, MD: U.S. Department of Health and Human Services. [Google Scholar]
  35. Neighbors CJ, Marquez DX, & Marcus BH (2008). Leisure-time physical activity disparities among Hispanic subgroups in the United States. Am J Public Health, 98(8), 1460–1464. doi: 10.2105/AJPH.2006.096982 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Pachucki MA, Jacques PF, & Christakis NA (2011). Social network concordance in food choice among spouses, friends, and siblings. Am J Public Health, 101(11), 2170–2177. doi: 10.2105/AJPH.2011.300282 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Powell K, Wilcox J, Clonan A, Bissell P, Preston L, Peacock M, & Holdsworth M (2015). The role of social networks in the development of overweight and obesity among adults: a scoping review. BMC Public Health, 15, 996. doi: 10.1186/s12889-015-2314-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rancourt D, Leahey TM, LaRose JG, & Crowther JH (2015). Effects of weight-focused social comparisons on diet and activity outcomes in overweight and obese young women. Obesity (Silver Spring), 23(1), 85–89. doi: 10.1002/oby.20953 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rancourt D, Thurston IB, Sonneville KR, Milliren CE, & Richmond TK (2017). Longitudinal impact of weight misperception and intent to change weight on body mass index of adolescents and young adults with overweight or obesity. Eat Behav, 27, 7–13. doi: 10.1016/j.eatbeh.2017.08.002 [DOI] [PubMed] [Google Scholar]
  40. Rehm CD, Penalvo JL, Afshin A, & Mozaffarian D (2016). Dietary Intake Among US Adults, 1999–2012. JAMA, 315(23), 2542–2553. doi: 10.1001/jama.2016.7491 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rochester Healthy Community Partnership. Retrieved from www.rochesterhealthy.org
  42. Rosenquist JN, Murabito J, Fowler JH, & Christakis NA (2010). The spread of alcohol consumption behavior in a large social network. Ann Intern Med, 152(7), 426–433, W141. doi: 10.7326/0003-4819-152-7-201004060-00007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sallis JF, Grossman RM, Pinski RB, Patterson TL, & Nader PR (1987). The development of scales to measure social support for diet and exercise behaviors. Prev Med, 16(6), 825–836. [DOI] [PubMed] [Google Scholar]
  44. Schifter DE, & Ajzen I (1985). Intention, perceived control, and weight loss: an application of the theory of planned behavior. J Pers Soc Psychol, 49(3), 843–851. [DOI] [PubMed] [Google Scholar]
  45. Seo DC, & Li K (2010). Leisure-time physical activity dose-response effects on obesity among US adults: results from the 1999–2006 National Health and Nutrition Examination Survey. Journal of epidemiology and community health, 64(5), 426–431. doi: 10.1136/jech.2009.089680 [DOI] [PubMed] [Google Scholar]
  46. Shakya HB, Christakis NA, & Fowler JH (2015). Self-comparisons as motivators for healthy behavior. Obesity. doi: 10.1002/oby.21201 [DOI] [PubMed] [Google Scholar]
  47. Shelton RC, Lee M, Brotzman LE, Crookes DM, Jandorf L, Erwin D, & Gage-Bouchard EA (2019). Use of social network analysis in the development, dissemination, implementation, and sustainability of health behavior interventions for adults: A systematic review. Soc Sci Med, 220, 81–101. doi: 10.1016/j.socscimed.2018.10.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. StataCorp. (2015). Stata Statistical Software: Release 14. In. College Station, TX: StataCorpLP.Ed. [Google Scholar]
  49. Steffens DC, Hays JC, George LK, Krishnan KR, & Blazer DG (1996). Sociodemographic and clinical correlates of number of previous depressive episodes in the depressed elderly. J Affect Disord, 39(2), 99–106. doi: 10.1016/0165-0327(96)00019-5 [DOI] [PubMed] [Google Scholar]
  50. Subramanian SV, Kim DJ, & Kawachi I (2002). Social trust and self-rated health in US communities: a multilevel analysis. Journal of urban health : bulletin of the New York Academy of Medicine, 79(4 Suppl 1), S21–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Suls J, Wan CK, & Sanders GS (1988). False consensus and false uniqueness in estimating the prevalence of health-protective behaviors. J Appl Soc Psychol, 18, 66–79. [Google Scholar]
  52. Team RC (2017). R: A language and environment for statistical computing. Retrieved from https://www.R-project.org/
  53. Valente TW (2010). Social networks and health: models, methods, and applications. New York: Oxford University Press. [Google Scholar]
  54. Valente TW (2012). Network interventions. Science, 337(6090), 49–53. doi: 10.1126/science.1217330 [DOI] [PubMed] [Google Scholar]
  55. VanderWeele TJ (2014). A unification of mediation and interaction: a 4-way decomposition. Epidemiology, 25(5), 749–761. doi: 10.1097/EDE.0000000000000121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wallerstein N, & Duran B (2010). Community-based participatory research contributions to intervention research: the intersection of science and practice to improve health equity. Am J Public Health, 100 Suppl 1, S40–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wen M, Cagney KA, & Christakis NA (2005). Effect of specific aspects of community social environment on the mortality of individuals diagnosed with serious illness. Social science & medicine, 61(6), 1119–1134. doi: 10.1016/j.socscimed.2005.01.026 [DOI] [PubMed] [Google Scholar]
  58. Wieland ML, Hanza MMM, Weis JA, Meiers SJ, Patten CA, Clark MM, … Sia IG (2018). Healthy Immigrant Families: Randomized Controlled Trial of a Family-Based Nutrition and Physical Activity Intervention. Am J Health Promot, 32(2), 473–484. doi: 10.1177/0890117117733342 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Williams JE, Kabukuru A, Mayo R, & Griffin SF (2011). Commentary: A social-ecological perspective on obesity among Latinos. Ethn Dis, 21(4), 467–472. [PubMed] [Google Scholar]
  60. Wing RR, & Jeffery RW (1999). Benefits of recruiting participants with friends and increasing social support for weight loss and maintenance. J Consult Clin Psychol, 67(1), 132–138. [DOI] [PubMed] [Google Scholar]
  61. World Health Organization. (2007). World Health Organization process of translation and adaptation of instruments. Retrieved from http://www.who.int/substance_abuse/research_tools/translation/
  62. Zhang S, de la Haye K, Ji M, & An R (2018). Applications of social network analysis to obesity: a systematic review. Obes Rev, 19(7), 976–988. doi: 10.1111/obr.12684 [DOI] [PubMed] [Google Scholar]

RESOURCES