Social networks and disparities in health behaviors and breast cancer outcomes in immigrant women

摘要 乳腺癌（BC）是亚太地区最常见的癌症和癌症死亡原因 岛民（API）和西班牙裔妇女是美国增长最快的两个少数族裔群体， 移民.一旦确诊，移民API妇女有类似或更糟，移民西班牙裔妇女 具有更好的，BC生存比种族相似，美国出生的妇女，后者尽管在诊断后期。 调查人员经常推测，社交网络在BC结果中发挥重要作用， 移民人口，通过有益的健康行为，由于生活在社区， 被描述为“种族或移民飞地”或通过与减少可用性有关的不良行为， 资源在个人层面上，在大部分非西班牙裔白色人群中， 网络已被证明具有更好的健康行为和BC存活率。然而，没有研究表明 探讨了社交网络对健康行为和BC的多层次（邻里和个人）影响 结果在不同的API和西班牙裔移民妇女。因此，在一个大的（N= 6，184），汇集北方 加州队列，包括683名西班牙裔（41%移民）、851名API（59%移民）和4，650名（8%移民）NHW 来自Pathways、癌症后生活流行病学和亚洲CHI研究的女性，被诊断为 在BC第I-IV阶段，我们建议：1）评估移民身份与社交网络之间的关联， 邻域级别（即，移民和族裔飞地以及社会文化机构）和个人层面 (i.e.,自我报告的规模，多样性，质量和社会关系类型），总体和种族/民族; 2）评估 移民身份、多层次社会网络特征与行为因素的关系 已知影响BC存活率（西方饮食，大豆和乳制品消费，MET-体力活动小时， 酒精摄入量和频率、当前和包年吸烟、肥胖和腰围），总体 和种族/民族;和3）评估移民身份，多层次的社会网络之间的关联 特征和BC结局，包括复发、BC特异性死亡率和总死亡率，总体和按 种族/民族，评估社会网络特征和健康行为的介导程度 协会.我们将使用荟萃分析技术来评估研究异质性;线性，逻辑， 广义线性混合效应（多水平）模型，用于分析目标1和2中的关联;以及多水平 考克斯比例风险回归评价目标3中的关联。独特的融合细节 关于移民身份、种族/民族、多层次社交网络、非常详细的行为测量的数据，以及 业连结果，提供了第一次机会，审查多层次的影响，社交网络对健康 行为和BC结果在不同的移民妇女与BC，调整了广泛的协变量。 这项工作对于查明不列颠哥伦比亚不同种族移民妇女中存在差异的根本因素至关重要， 特别是那些尽管处于社会经济劣势，但仍有利于BC结果的因素。

ABSTRACT Breast cancer (BC) is the most commonly diagnosed cancer and cause of cancer mortality in Asian/Pacific Islander (API) and Hispanic women, the two most rapidly growing minority ethnic groups in the US due to immigration. Once diagnosed, immigrant API women have similar or worse, and immigrant Hispanic women have better, BC survival than ethnically similar, US-born women, the latter despite later stage at diagnosis. Investigators have frequently speculated that social networks play an important role in BC outcomes among immigrant populations, through beneficial health behaviors that are maintained due to living in neighborhoods characterized as “ethnic or immigrant enclaves” or through adverse behaviors related to reduced availability of resources. At the individual level, in largely non-Hispanic white populations, women with large supportive social networks have been shown to have better health behaviors and BC survival. However, no research has explored the multilevel (neighborhood and individual) influence of social networks on health behaviors and BC outcomes in diverse API and Hispanic immigrant women. Therefore, in a large (N=6,184), pooled northern California cohort of 683 Hispanic (41% immigrant), 851 API (59% immigrant), and 4,650 (8% immigrant) NHW women from the Pathways, the Life After Cancer Epidemiology, and the Asian CHI studies, diagnosed with stage I-IV BC, we propose to: 1) Evaluate associations between immigrant status and social networks at the neighborhood level (i.e., immigrant and ethnic enclaves and sociocultural institutions) and at the individual level (i.e., self-reported size, diversity, quality, and type of social ties), overall and by race/ethnicity; 2) Evaluate associations between immigrant status and multilevel social network characteristics with behavioral factors known to influence BC survival (Western diet, soy and dairy consumption, MET-hours of physical activity, amount and frequency of alcohol intake, current and pack years of smoking, obesity, and waist size), overall and by race/ethnicity; and 3) Evaluate associations between immigrant status, multilevel social network characteristics, and BC outcomes including recurrence, BC-specific mortality, and total mortality, overall and by race/ethnicity, evaluating the degree to which social network characteristics and health behaviors mediate associations. We will use meta-analytic techniques to evaluate study heterogeneity; linear, logistic, and generalized linear mixed effects (multilevel) models to analyze associations in Aims 1 and 2; and multi-level Cox proportional hazards regression to evaluate associations in Aim 3. The unique convergence of detailed data on immigrant status, race/ethnicity, multilevel social networks, highly detailed behavioral measures, and BC outcomes, provides the first opportunity to examine the multilevel influence of social networks on health behaviors and BC outcomes in diverse immigrant women with BC, adjusted for an extensive set of covariates. This work is central to identifying factors underlying BC disparities in ethnically diverse immigrant women, particularly those factors underlying favorable BC outcomes despite socioeconomic disadvantage.