Determinants of Perceived Neighborhood Social Cohesion in a New York/New Jersey Based Latino Sample

Background: Perceived neighborhood social cohesion (NSC) describes relational characteristics such as the degree of trust and shared values among neighborhood residents. Interventions that promote NSC present an opportunity for improving health since NSC has been associated with positive health outcomes. However, few studies have examined its predictors. Furthermore, though differences in perception of NSC by acculturation proxies (i.e. nativity status and English language proficiency [ELP]) have not been studied, social ties studies suggest important differences, with those who are US-born having more extensive and heterogenous networks than immigrants. We assessed demographic, socioeconomic, and other individual-level neighborhood factors as determinants of perceived NSC followed by an exploration of differences by nativity status and ELP.

Methods: Using data from 186 healthy Latino adults who lived in New York or New Jersey and enrolled in the Latino Sleep and Health Study between 2016 - 2019, three linear regression models were used to assess predictors of NSC. NSC was measured using a validated 5-item scale, where increasing scores indicate increasing cohesion. Model 1 included demographic variables (age, gender), model 2 added objective and subjective indicators of socioeconomic status [SES] (household income, education, subjective social status), and model 3 added neighborhood years of residence and perceived neighborhood size in blocks. Analyses for models 1-3 were repeated stratified by nativity status and ELP (poor/fair and good/excellent).

Results: Participants were on average 37.36 years old (SD=1.02), 65.50% female, and 67.74% were at least college graduates. Most were immigrants (58.06%) and reported good/excellent ELP (76.22%). Overall, men reported an NSC score of 1.07 less than women (SE= 0.43, p=0.01). Having a household income $50,000 compared to $15,000 (b= 1.73, SE= 0.73, p=0.02) was positively associated with NSC. For immigrants, identifying as male vs. female (b= -1.61, SE= 0.60, p=0.01) and having a household income $50,000 compared to $15,000 (b= 2.74, SE= 0.97, p=0.01) were stronger predictors of NSC than in the overall sample. Subjective social status (SSS) was only statistically significant among US-born adults, increasing NSC by 0.49 (SE=0.24, p=0.05) for every 1-point increase in SSS score. Interestingly, for US-born participants, increasing neighborhood size was associated with a 0.04 increase in NSC (SE=0.02, p=0.02), whereas for immigrants as neighborhood size increased, NSC decreased (b= -0.03, SE=0.01, p=0.04). For those with good/excellent ELP, identifying as male versus female was associated with lower NSC score (b= -1.40, SE= 0.51, p=0.01), but not among the those with poor/fair ELP.

Conclusion: Female gender was the most consistent predictor of NSC across models. We also identified heterogeneity of findings by nativity status, where the importance of both objective measures of SES (household income for immigrants) and subjective measures of SES (SSS for US-born) were highlighted. Findings also indicated opposing effects of neighborhood size on NSC for immigrants (negative) vs. US-born (positive), which represent a novel contribution. These findings are consistent with prior research documenting differences in network size by nativity status, and imply existing social ties may be used to define spatial neighborhood boundaries.