Activating Communal Coping Related to Diabetes Risk in Mexican-Heritage Families

Abstract

We investigate how interpersonal ties influence communication about type 2 diabetes risk and encouragement to maintain or adopt a healthy lifestyle between family members of Mexican heritage, after a family history-based risk assessment intervention. Results suggest that individuals are more likely to initiate risk communication with another family member if they are close to, already seek advice from, or discuss health with him/her. Risk communication precedes encouragement, which is initiated by the older generation of the family. Understanding the role of interpersonal relationships in Mexican-heritage families can help identify who best to target in future health behavior interventions.

Primary risk factors for type 2 diabetes (T2D) include age, gestational diabetes, impaired glucose metabolism, family history, lack of physical activity, and obesity.^1–3 In 2017, an estimated 15.6% of adults living in the United States who were of Mexican heritage had diabetes.⁴ More than half of Mexican American adults are not sufficiently active,⁵ and about 70% are overweight or obese.⁵ Engaging in physical activity and adopting a low calorie and low fat diet can help prevent T2D or at least delay the onset of this condition.^6,7 Thus, how to more effectively promote risk-reducing behaviors in Mexican Americans has been an important research focus.

This paper situates individual health behaviors in the context of the family network, seeking to characterize the interpersonal dynamics relevant to communicating risk of T2D and encouraging others to maintain or adopt a healthy lifestyle. To activate the communication pathway, we use a family health history-based intervention, providing household members with personalized feedback about their familial risk for T2D. Family history is a powerful predictor of T2D—individuals at moderate or high familial risk of T2D are 2.3 and 5.5 times more likely to develop T2D compared to those at average risk.³ Because familial risk of complex disease is relevant to the entire household, the provision of family history risk feedback can potentially activate a communal coping process, whereby family members consider the management of shared familial risk of T2D as a joint responsibility.^{8, 9} One or more family members may appraise and communicate about familial T2D risk and encourage others to adopt healthy behaviors. More specifically, the communal coping model predicts that appraisal and communication about disease risk precede encouragement and the development of strategies to engage in healthy behaviors and disease screening.⁹

This process of communal coping as it relates to T2D risk is particularly salient in Mexican-heritage families where familismo and respeto are central cultural values.¹⁰ Familismo refers to the importance that Latinos place on family closeness.¹¹ Research shows that Mexican American families are often characterized by more shared daily activities,¹¹ as well as strong family ties.¹² In addition to the primacy of family relationships, respeto further emphasizes a sense of obligation to and respect for the older generation of the family.^13,14 As such, Mexican Americans are found to be more likely to seek older family member’s advice for important decisions.^13,14 Jointly, familismo and respeto represent interpersonal processes that may be characterized by highly correlated interpersonal closeness, support exchange, and advice-seeking ties. Family support, family cohesion and close family relationships, which are common in Mexican-heritage individuals, have been associated with better health outcomes and risk-reducing behaviors,^15–17 and respeto has been associated with decreased substance use in young Hispanic adults.¹⁸

While some have examined risk communication and encouragement in Mexican Americans from a network perspective,^19,20 the characteristics of individuals involved in these communication and encouragement ties remain unknown. Given the centrality of familismo and respeto in this population, it is important to explore the characteristics of communicators and encouragers in Mexican-heritage families, as well as the properties of the ties connecting them. Employing family members who are influential in motivating risk-reducing behaviors in interventions may be particularly effective for achieving sustainable behavior change.

In a sample of multigenerational Mexican-heritage families, we apply a family health history-based intervention that provides risk feedback generated by the Centers for Disease Control and Prevention (CDC)’s Family Healthware™.²¹ Using social network methodologies, we explore individual- and interpersonal-level predictors of new ties, post-intervention, that represent communication about T2D risk and encouragement of another family member to exercise regularly, maintain a healthy weight, and eat a healthy diet. Examining these predictors will help researchers and clinicians better understand the interpersonal processes that contribute to communication about disease risk and encouragement to change behaviors and help identify who best to target in future interventions.

METHODS

Procedure

The current study drew data from Project [removed for blind review], which was a randomized intervention study engaging Mexican-heritage households in Harris County, Texas.²² Project [removed for blind review] recruited a total of 497 participants from 162 multigenerational Mexican American households from the Mano a Mano cohort.²² Participants included at least three adult members within each household, with two who were biologically related across two generations (i.e., parent and adult child) and two who were socially related (i.e., partnered or married). The baseline interview was conducted by a pair of bilingual interviewers during in-home visits in 2008. Within two weeks of completing the baseline interview, participants received in the mail an individual feedback packet generated by Family Healthware™. Two follow-up telephone interviews were conducted three months and ten months post intervention, respectively. The interviewers were blinded to household feedback condition. Individual demographic and health information, family networks, and family health history of T2D and comorbid conditions were collected using structured questionnaires at baseline and follow-up interviews. The Institutional Review Boards of [removed for blind review] and [removed for blind review] approved all study materials. Written and verbal consent were obtained prior to study participation.

Intervention

Developed by CDC, Family Healthware™ is a web-based tool that assesses familial risk for six common complex diseases and provides a personalized prevention plan.²¹ In this study, we input the family health information participants provided at baseline interview into Family Healthware™ to obtain a pedigree and supplementary feedback including personalized risk assessment and tailored behavioral recommendations about lifestyle management and preventive screening for type 2 diabetes and co-morbid conditions.

Randomization

While all participants received their pedigree, some participants received supplementary feedback with different elements. This was determined at the household-level by a 2×2 factorial design. We first contrasted a family-centered approach where all household members received supplementary feedback with an individual-centered approach where one household member received it. Second, the element of the supplementary feedback could be either predisposing—a personalized risk assessment based on family history, or enabling—a risk assessment coupled with personalized behavioral recommendations. Thus, all participating households were randomized into one of the following four conditions: (1) all family members received supplemental risk assessments and behavioral recommendations, (2) all family members received supplemental risk assessments but no behavioral recommendations, (3) one family member received supplemental risk assessment and behavioral recommendations, and (4) one family member received supplemental risk assessment but no behavioral recommendations.

Measures

Network ties with respect to diabetes risk communication, encouragement to adopt a healthy lifestyle, interpersonal closeness, advice seeking, and general health discussions were measured as part of the family network assessment, where participants enumerated a roster of living first and second degree relatives and other important persons who played a significant role in their life during the past year, such as friends and family members. The list of enumerated other important persons was limited to at most twenty. Using this roster, participants indicated with whom they discussed risk for diabetes and whom they had encouraged to a) do regular physical activity, b) maintain a healthy weight, and c) eat more fruits and vegetables. The responses could be everyone, no one, or specific name(s) from the roster. We measured these ties at baseline and reassessed risk communication ties at three-month post intervention and reassessed encouragement ties at ten-month post intervention. This was done to establish temporal precedence of risk communication and behavioral encouragement, consistent with the communal coping model. Participants also indicated to whom they felt close, from whom they sought advice, and with whom they discussed health and/or health concerns, all of which were measured at baseline.

Additional predictors included baseline individual demographic and health information including age (in years), gender (female = 1, male = 0), education (high school or more = 1, less than high school = 0), birth place (born in Mexico = 1, born in United States = 0), obese (defined as a body mass index of 30 or greater based on self-reported height and weight), and whether the participant had ever been diagnosed with T2D (based on self-reports; yes = 1, no = 0). We used the participants’ reported kinship relation with network members to determine if the nominated network member was in the younger generation (i.e., children, nieces/nephews, grandchildren), older generation (i.e., parents, aunts/uncles, grandparents), or the same generation (i.e., siblings, cousins, friends/coworkers; reference category) relative to the participant.

Analysis Plan

After deleting missing values in study variables, the final sample consisted of 447 individuals from 161 Mexican-heritage households, who had enumerated 4,827 uniquely identified network members and reported a total of 7,020 family network ties for analysis. The descriptive analysis calculated the proportion of risk communication ties and encouragement ties with regards to exercise, weight control and diet by feedback conditions at baseline and follow-up interviews. We performed equality of proportions tests to determine if there was a significant increase in these ties following receipt of supplementary feedback.

In explanatory analysis, we examined how individual- and interpersonal-level predictors were associated with risk communication and encouragement ties at baseline and at follow-up interviews. The first two models estimated risk communication ties at baseline, and at three-month post intervention conditional on no such communication tie being present at baseline (i.e., new communication ties). The models included feedback conditions (all received risk assessment and recommendations as reference category), closeness ties, advice seeking ties, health discussion ties, generation of the network member relative to participant, and participant attributes as predictors. Similarly, we modeled encouragement ties, for each lifestyle domain, at baseline and at ten-month post intervention conditional on no such tie being present at baseline (i.e., new encouragement ties). For illustrative purposes, in this set of models we used the feedback condition where one member of the household received risk assessment without recommendations as the reference category to provide a more straightforward interpretation of the intervention effect. The models estimating new encouragement ties included, additionally, all risk communication ties at three-month post intervention as a predictor, to examine whether risk communication ties mediate the effect of feedback conditions on new encouragement ties.

Generalized estimation equations (GEE) with exchangeable correlations were used to estimate these models to account for clustering in the data. A logit link was used to accommodate dichotomized outcome variables. All models control for dependency arising from families potentially having different propensities for communicating risk or encouraging healthful behaviors, by including family-level density of the risk communication and encouragement networks as a covariate.

RESULTS

As shown in Table 1, the average age of the participants was 41 years. Most of the participants were born in Mexico and just over half (55%) were female. At the time of the study, 46% were obese, and 16% were diagnosed with T2D. More than half of the network members enumerated were from the older generation relative to the participant, 14% were from the younger generation and the remaining 34% were of the same generation as the participant. At baseline, the proportion of ties reported was 29% for closeness, 16% for advice seeking, and 17% for discussing health and/or health concerns. The proportion of risk communication ties increased from 19% at baseline to 22% at three-month post intervention. Across lifestyle domains, the proportion of encouragement ties was very similar, at slightly above a quarter at baseline and increased to more than one-third at ten-month follow-up interview. When stratified by feedback conditions (Table 2), the increase in risk communication ties and encouragement ties was statistically significant across feedback conditions The only exception was that there was no change in risk communication ties from baseline to three-month post intervention for households where one member received risk assessment and recommendations.

Table 1.

Descriptive Statistics of Study Variables

	BL	3m	10m
Participant Attributes (n = 447)
Age (in years)	Mean = 41.22
Female	55%
Born in Mexico	70%
High School or Above	41%
Obese	46%
Have Diabetes	16%
Generation of Family Member (n = 4,827)
Younger	14%
Older	52%
Same	34%
Network Ties (n = 7,020)
Closeness	29%
Advice Seeking	16%
Health Discussion	17%
Risk Communication	19%	22%
Encouragement
Exercise	26%		33%
Weight Control	27%		33%
Diet	27%		35%

Note: BL = baseline, 3m = 3-month follow up, 10m = 10-month follow up.

Table 2.

Proportion of Risk Communication and Encouragement Ties by Feedback Conditions at Baseline and Follow-Up Interviews

	n	Risk Communication		Encouragement
				Exercise		Weight Control		Diet
Feedback Condition		BL	3m	BL	10m	BL	10m	BL	10m
All received risk assessment and recommendation	1,737	18%	22%	25%	33%	27%	34%	30%	38%
		(p<.001)		(p<.001)		(p<.001)		(p<.001)
All received risk assessment, no recommendation	1,796	23%	28%	26%	33%	28%	33%	29%	36%
		(p=.001)		(p<.001)		(p<.001)		(p<.001)
One received risk assessment and recommendation	1,695	22%	20%	27%	32%	27%	33%	26%	33%
		(p=.339)		(p<.001)		(p<.001)		(p<.001)
One received risk assessment, no recommendation	1,792	14%	19%	23%	32%	25%	33%	26%	33%
		(p<.001)		(p<.001)		(p<.001)		(p<.001)
Total	7,020	19%	22%	26%	33%	27%	33%	27%	35%
		(p<.001)		(p<.001)		(p<.001)		(p<.001)

Note: BL = baseline, 3m = 3-month follow up, 10m = 10-month follow up.

Patterns of Baseline and New Risk Communication

Table 3 presents GEE models examining predictors of risk communication and encouragement ties at baseline and at follow-up interviews. There was no difference in baseline risk communication ties across households randomized into different intervention conditions. T2D risk communicators tended to be female (Odds Ratio [OR] = 2.08; 95% Confidence Interval [CI]: 1.74–2.50), older (OR = 1.04; 95% CI: 1.03–1.05), born in Mexico (OR = 1.49; 95% CI: 1.17–1.89), have a high school or higher education (OR = 1.47; 95% CI: 1.19–1.81), and have a T2D diagnosis (OR = 2.00; 95% CI: 1.56–2.57). Generational difference between the participant and enumerated family member was not associated with risk communication at baseline. Participants were more likely to discuss T2D risk with a family member to whom they felt close (OR = 2.68; 95% CI, 2.18–3.29) and from whom they sought advice (OR = 2.08; 95% CI: 1.64–2.63). Having a general health discussion tie with a family member was associated with high likelihood of discussing T2D risk with him/her at baseline (OR = 12.58; 95% CI: 10.00–15.82).

Table 3.

Predictors of Risk Communication and Encouragement Ties at Baseline and Following Feedback

	Risk Communication		Encouragement
			Exercise		Weight Control		Diet
Feedback Condition	BL	3m	BL	10m	BL	10m	BL	10m
1. All received risk assessment and recommendation	Ref.	Ref.	0.73 [0.52−1.03]	1.83 [1.05−3.18]	0.76 [0.54−1.08]	1.44 [0.83−2.50]	0.74 [0.50−1.10]	2.14 [1.23−3.72]
2. All received risk assessment, no recommendation	0.85 [0.61−1.17]	1.67 [1.11−2.50]	0.97 [0.70−1.35]	1.88 [1.10−3.22]	0.81 [0.57−1.13]	1.20 [0.70−2.07]	0.71 [0.48−1.05]	1.79 [1.03−3.09]
3. One received risk assessment and recommendation	0.98 [0.70−1.37]	1.79 [1.16−2.78]	0.68 [0.48−0.95]	0.60 [.32−1.13]	0.61 [0.43−0.88]	0.65 [0.36−1.18]	0.48 [0.32−0.72]	0.95 [0.54−1.69]
4. One received risk assessment, no recommendation	1.08 [0.77−1.53]	1.86 [1.22−2.84]	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Participant Attributes
Age	1.04 [1.03−1.05]	1.05 [1.04−1.06]	1.01 [1.00−1.01]	0.99 [0.98−1.00]	1.00 [0.99−1.00]	1.00 [0.99−1.01]	0.99 [0.99−1.00]	1.02 [1.01−1.03]
Female	2.08 [1.74−2.50]	0.90 [0.74−1.08]	1.03 [0.90−1.19]	0.91 [0.74−1.13]	1.26 [1.09−1.45]	0.85 [0.68−1.06]	1.07 [0.94−1.22]	1.11 [0.90−1.37]
Born in Mexico	1.49 [1.17−1.89]	1.97 [1.48−2.62]	1.82 [1.47−2.25]	1.34 [0.95−1.88]	1.95 [1.57−2.42]	1.42 [1.00–2.02]	1.59 [1.30–1.95]	1.40 [1.00−1.95]
High School and Above	1.47 [1.19−1.81]	1.79 [1.40−2.28]	1.27 [1.06−1.51]	1.02 [0.76−1.36]	1.24 [1.09−1.45]	1.26 [0.93−1.70]	0.81 [0.68−0.96]	1.51 [1.13−2.03]
Obese	0.88 [0.72−1.07]	1.06 [0.85−1.31]	0.97 [0.82−1.14]	0.88 [0.67−1.14]	1.04 [0.88−1.23]	0.82 [0.63−1.08]	1.09 [0.93−1.27]	0.86 [0.66−1.12]
Has Diabetes	2.00 [1.56−2.57]	1.00 [1.00−1.00]	0.93 [0.74−1.16]	1.00 [1.00−1.01]	1.38 [1.11−1.71]	1.00 [1.00−1.01	1.62 [1.32−1.98]	0.99 [0.99−1.00]
Generation of Family Member
Younger	0.95 [0.75−1.20]	1.04 [0.82−1.33]	2.26 [1.89−2.71]	4.76 [3.69−6.15]	1.92 [1.59−2.31]	4.84 [3.75−6.26]	2.96 [2.50−3.51]	8.09 [6.28−10.43]
Older	0.82 [0.68−1.01]	0.83 [0.67−1.01]	0.64 [0.55−0.75]	0.40 [0.31−0.52]	0.74 [0.63−0.87]	0.46 [0.35−0.60]	0.77 [0.67−0.89]	0.43 [0.33−0.56]
Same	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Close at BL	2.68 [2.18−3.29]	1.47 [1.17−1.85]	2.20 [1.87−2.59]	1.14 [0.89−1.45]	2.85 [2.42−3.37]	1.54 [1.20−1.97]	2.86 [2.46−3.34]	1.45 [1.14−1.85]
Advice at BL	2.08 [1.64−2.63]	2.27 [1.70−3.03]	2.05 [1.68−2.50]	1.27 [0.92−1.73]	1.54 [1.26−1.89]	1.25 [0.89−1.74]	1.27 [1.05−1.54]	0.75 [0.54−1.04]
Discuss Health/ Concerns at BL	12.58 [10.00−15.82]	3.31 [2.47−4.45]	2.95 [2.43−3.57]	2.06 [1.55−2.75]	3.69 [3.02−4.50]	1.02 [0.73−1.42]	3.22 [2.67−3.88]	1.27 [0.92−1.74]
Risk Communication at 3m	--	--	--	2.08 [1.59−2.71]		2.06 [1.57−2.72]	--	1.76 [1.36−2.29]
Density (normalized)	3.62 [3.18−4.12]	3.92 [3.33−4.51]	3.06 [2.69−3.48]	1.86 [1.53−2.28]	3.54 [3.08−4.07]	2.10 [1.70−2.58]	2.65 [2.28−3.08]	2.20 [1.78−2.73]
n	7,020	5,684	7,020	5,227	7,020	5,148	7,020	5,094

Note: BL = baseline, 3m = 3-month follow up, 10m = 10-month follow up. Generalized estimation equations with exchangeable correlations and a logit link. Odds ratios and 95% confidence intervals (in brackets) reported. Statistically significant effects (α = .05) are bolded.

The next model estimated new risk communication ties at three-month follow-up, conditional on such ties being absent at baseline. Relative to the feedback condition where all members received risk assessment and behavioral recommendation, the other three feedback conditions were all significantly associated with a sizable increase in the chance of the participant initiating new communication ties at three-month follow-up (ORs = 1.67 ([95% CI: 1.11–2.50], 1.79 [95% CI: 1.16–2.78], and 1.86 [95% CI: 1.22–2.84], respectively). Older participants had a slightly higher chance of initiating new risk communication ties (OR = 1.05; 95%CI: 1.04–1.06). Being born in Mexico and having a high school or higher education were both positively associated with new risk communication ties (ORs = 1.97 [95% CI: 1.48–2.62] and 1.79 [95% CI: 1.40–2.28], respectively). The participants’ gender and health status and generational difference between the participant and the network member did not have an effect on new risk communication ties. Similar to the pattern observed at baseline, an established communication channel appeared to be particularly important for more focused risk communication about T2D. There was a greater chance for a participant to initiate risk communication with a family member post intervention, if the participant was close to (OR = 1.47, 95% CI:1.17–1.85), already sought advice from (OR = 2.27; 95% CI: 1.70–3.03) or discussed health and/or health concerns with him/her OR = 3.31; 95% CI: 2.47–4.45).

Patterns of Baseline and New Encouragement

To aid the interpretation of the intervention effect, models estimating encouragement ties at baseline and ten-month follow-up used the feedback condition where one member receives feedback, no recommendation as the reference category. Relative to this condition, participants in households where one received risk assessment and recommendation were less likely to have, at baseline, encouragement ties to exercise (OR = 0.68; 95% CI: 0.48–0.95), maintain a healthy weight (OR = 0.61; 95% CI: 0.43–0.88) and eat a healthy diet (OR = 0.48; 95% CI: 0.32–0.72). At ten-month post intervention, the conditions where all members received supplementary feedback were associated with a significant increase in new encouragement ties about exercise (risk assessment only: OR = 1.83, 95% CI: 1.05–3.18; risk assessment and recommendations: OR = 1.88, 95% CI: 1.10–3.22) and diet (risk assessment only: OR = 2.14, 95% CI: 1.23–3.72; risk assessment and recommendations: OR = 1.79, 95% CI: 1.03–3,09), but not weight control.

The participants’ birth place and education were significantly associated with all three types of encouragement ties at baseline. At ten-month follow-up, the effect of birth place was observed for new encouragement ties for weight control (OR = 1.42, 95% CI: 1.00–2.02) and diet (OR = 1.40, 95% CI: 1.00–1.95), whereas the effect of education was only observed for new ties encouraging eating a healthy diet (OR = 1.51, 95% CI: 1.13–2.03). Contrary to the pattern observed for risk communication ties, generational differences between the participant and the family member had a significant, sizable effect on existing and new encouragement ties initiated post intervention. At baseline, participants were more likely to encourage a family member of the younger generation to exercise (OR = 2.26; 95% CI: 1.89–2.71), maintain a healthy weight (OR = 1.92; 95% CI: 1.59–2.31), and eat a healthy diet (OR = 2.9; 95% CI: 2.50–3.51). They were, at the same time, less likely to do so for someone from the older generation (exercise: OR = 0.64; 95% CI: 0.55–0.75; weight control: OR = 0.74, 95% CI: 0.63–0.87; diet: OR = 0.77, 95% CI: 0.67–0.89). Such a generational effect was even more prominent at the 10-month follow-up, with participants being four to eight times more likely to encourage the younger generation to adopt a healthy lifestyle, and about 60% less likely to do so for the older generation.

Notably, risk communication ties at three-month follow up was a significant predictor of new encouragement ties about exercise (OR = 2.08; 95% CI: 1.59–2.71), weight control (OR = 2.06; 95% CI: 1.57–2.72) and diet (OR = 1.76; 95% CI: 1.36–2.29), evident of a mediation effect. Closeness, advice seeking and general health discussion ties affect new encouragement ties at 10-month post intervention differently. Baseline closeness ties were associated with more new encouragement ties about weight control (OR = 1.54; 95% CI: 1.20–1.97) and diet (OR = 1.45; 95% CI: 1.14–1.85), but did not affect encouragement to exercise. Advice seeking ties had no effect on any of the new encouragement ties examined. General health discussion ties were associated with greater chance of the participant encouraging family members to exercise regularly (OR = 2.06; 95% CI: 1.55–2.75), but had no effect on weight control or diet.

DISCUSSION

Overall, we found significant, sizable increases in new risk communication and encouragement ties after an intervention that provided personalized risk feedback to Mexican American family members. More importantly, the activation of risk communication and encouragement ties requires different feedback conditions. Providing less information (i.e., risk assessment and no behavioral recommendations) to only one family member is optimal for initiating new risk communication ties, whereas providing more information (i.e., risk assessments and behavioral recommendations) to all household members is more effective in generating new encouragement ties. However, discussions of T2D risk are important to initiating encouragement of health promoting behaviors. These results suggest that interventions aimed at promoting healthier lifestyle in Mexican-heritage families may benefit from a two pronged approach: First, a family member can be recruited to act as a genomics health educator with the goal of engaging members in discussions about risk. In so doing, family members develop common appraisals of their shared risk of disease.²³ After which, an education program aimed at promoting risk-reducing behaviors can be provided to the entire household, with the goal of activating support and encouragement of such behaviors within the family system.

Consistent with the communal coping model, receiving risk feedback activates risk communication ties at the three-month follow-up, which precede behavioral encouragement ties at the ten-month follow-up. Here, two interpersonal factors merit attention. First, feeling close to someone was associated with a higher chance of the participant’s communicating T2D risk to others and encouraging positive behavioral modifications. Previous studies focusing on families affected by hereditary nonpolyposis cancer have found that interpersonal closeness facilitates communication about cancer risk and encouragement of screening and testing.^24,25 Therefore, close interpersonal relationships can be important to health communication in multiple disease contexts and across different populations. Similarly, advice seeking and general health discussion ties are prerequisite for more focused risk communication and encouragement in this sample of Mexican-heritage individuals, suggesting the need to consider the broader relational context within which specific conversations about T2D, weight, diet and exercise occur. Cultural beliefs of familismo, characterized by close interpersonal relationships and those with whom one discuss health and health concerns, may play important roles in identifying those involved in risk communication and behavioral encouragement processes within Mexican-heritage families.

Second, we have found that participants are more likely to encourage physical activity, weight control and healthy eating among family members who are of the younger generation. Moreover, the older generation are found to be the new encouragers, initiating new conversations about maintaining or adopting a healthy lifestyle after the intervention. Previous research has shown that risk feedback based on family history can change individuals’ perceived susceptibility to disease,^8,26 potentially leading to behavioral changes. Our results suggest that providing personalized risk feedback can have broader impact beyond the individual who have received risk feedback. Future family health history-based interventions should promote encouragement of risk-reducing behaviors among the younger generation by the older generation of the family—an important social resource in the family system and reflective of respeto within the Mexican-heritage community.¹⁹

At the same time, our results show that younger generation family members are more likely to be recipients of encouragement, rather than advocates. Interestingly, the literature is not consistent regarding how generation is related to specific communication and encouragement behaviors in families. Adult children have been identified as encouragers for healthy eating and weight control,¹⁸ whereas parents are found to be more effective at encouraging screening for blood sugar, blood pressure and blood cholesterol.²⁷ One possible explanation is that individuals give encouragement on activities that are most salient to their own generation. We also found that for the purpose of initiating new encouragement ties, an intervention targeting all members of the household is most effective, resonating with the literature on increasing the number of encouragers in a network to achieve more desirable social influence.^19,27,28 This finding suggest the need to better engage the younger generation in family-based health interventions and call for more research exploring how the exchange of information and resources is affected by generational structure in Mexican-heritage families and other at-risk populations.

At the individual-level, education is positively associated with initiating new risk communication and encouragement ties, complementing previous research showing that Hispanic adults with higher levels of education are more receptive to interpersonal health messages related to cardiovascular disease risk.²⁹ Therefore, education is important to both sending and receiving health information in Mexican-heritage families. Contrary to prior studies, which mostly suggest that health communication tends to be initiated by women,^24,25 we do not observe any gender effect on new risk communication or behavioral encouragement ties. This opens up the possibility of recruiting both male and female family members as health educators and sources of support for risk-reducing behaviors in Mexican-heritage families. Future research is needed to fully understand the perspectives of men in this population.

The primary limitation of our study results from a rather selective sample. We recruited adult participants of Mexican heritage in a metropolitan area. The participants are also largely homogenous with regards to socioeconomic status. This means that our findings may not be generalizable to other populations living in other geographic areas or with different social and cultural backgrounds and acculturation levels. Relatedly, the families under study mostly have family members living in Texas. Since geographic proximity affects interpersonal contact, our results may be biased towards more communication and encouragement ties.

CONCLUSION

Our findings suggest that an intervention to increase awareness and communication about family history risk may have resulted in more encouragement of younger-generation family members and that interventions targeting multiple family members are more effective at increasing encouragement to adopt healthy behaviors than ones which target only one family member. Communication about risk and encouragement of younger-generation family members is significant since healthy behaviors are best learned at a young age and the earlier health promotion behaviors are adopted, the greater the long-term impact in terms of reduced morbidity and mortality. Our findings highlight the importance to consider not only individual-level factors, but the broader family environment when trying to address the complex and multi-faceted nature of behavioral change, especially among at-risk populations.