TY - JOUR
T1 - Localizing syndemics
T2 - A comparative study of hunger, stigma, suffering, and crime exposure in three Haitian communities
AU - Brewis, Alexandra
AU - Wutich, Amber
AU - Galvin, Michael
AU - Lachaud, James
N1 - Funding Information:
The data discussed herein were collected in association with the USAID Haiti Justice Sector Strengthening Project (JSSP), in partnership with USAID contractors Chemonics International and sub-contractors Diagnostics and Development Group (DDG). The results herein, however, reflect the authors' analyses and interpretation only. We thank everyone involved for acute attention to the quality and integrity of sampling, data collection, and data entry processes, most especially Christelle Safi, Helga Klein, and Donald Vertus (Chemonics International), Luckny Zephyr [technical director], Isnel Pierreval, Nelson Sylvestre, Rosalvo Dort, Mireille Guerrier, and Shirley Augustin (DDG). We are also grateful to the 55 students from State University of Haiti for their significant and important efforts during data collection.
Publisher Copyright:
© 2020 Elsevier Ltd
PY - 2022/2
Y1 - 2022/2
N2 - Theoretically, disease syndemics are hyper-localized in the forms they take, but little empirical data show how localization manifests. We present a comparison across three sites in Haiti, from data collected in June–august 2017 testing for localizations of risks across three communities: rural farming, border town, and in a high gang-activity urban zone. First, we modeled survey responses collected from heads of 4055 geographically-sampled households via linear regression, considering additive and interaction effects of food insecurity, crime exposure, and discrimination on depression and anxiety levels. Exposure to food insecurity, crime exposure, and discrimination were each associated with more depression and anxiety symptoms. For those living in the urban zone, there was weak evidence of possible interactional risks between the three vulnerabilities, suggesting little meaningful localized syndemic patterning. Second, we conducted thematic and word-based semantic network analysis to identify if people themselves cognitively connected vulnerabilities of hunger/poverty, crime, and suffering/discrimination using 7321 text blocks from 95 semi-structured interviews/focus groups. Network visualization suggested people commonly connect these domains. While the patterns were localized, crime concerns were central to all networks. The domain connections expressed through people's own words were more complexly inter-related than was evident from the modeled survey data, and suggested counter-intuitive influences. The quantitative approach to modeling syndemic interactions suggests no apparent practical benefits to layering or combining local anticrime, anti-hunger, and anti-discrimination programming. However, the qualitative network analysis suggests that programming could none-the-less leverage the perceived connections across domains for more meaningful and effective interventions. For the broader study of syndemics, incorporating novel qualitative approaches clarifies that constituent processes are not just potentially localizing suffering, but are also extremely important in how people cognitively understand and organize their everyday lives.
AB - Theoretically, disease syndemics are hyper-localized in the forms they take, but little empirical data show how localization manifests. We present a comparison across three sites in Haiti, from data collected in June–august 2017 testing for localizations of risks across three communities: rural farming, border town, and in a high gang-activity urban zone. First, we modeled survey responses collected from heads of 4055 geographically-sampled households via linear regression, considering additive and interaction effects of food insecurity, crime exposure, and discrimination on depression and anxiety levels. Exposure to food insecurity, crime exposure, and discrimination were each associated with more depression and anxiety symptoms. For those living in the urban zone, there was weak evidence of possible interactional risks between the three vulnerabilities, suggesting little meaningful localized syndemic patterning. Second, we conducted thematic and word-based semantic network analysis to identify if people themselves cognitively connected vulnerabilities of hunger/poverty, crime, and suffering/discrimination using 7321 text blocks from 95 semi-structured interviews/focus groups. Network visualization suggested people commonly connect these domains. While the patterns were localized, crime concerns were central to all networks. The domain connections expressed through people's own words were more complexly inter-related than was evident from the modeled survey data, and suggested counter-intuitive influences. The quantitative approach to modeling syndemic interactions suggests no apparent practical benefits to layering or combining local anticrime, anti-hunger, and anti-discrimination programming. However, the qualitative network analysis suggests that programming could none-the-less leverage the perceived connections across domains for more meaningful and effective interventions. For the broader study of syndemics, incorporating novel qualitative approaches clarifies that constituent processes are not just potentially localizing suffering, but are also extremely important in how people cognitively understand and organize their everyday lives.
KW - Anxiety
KW - Crime
KW - Depression
KW - Discrimination
KW - Food insecurity
KW - Haiti
KW - Suffering
KW - Syndemics
UR - http://www.scopus.com/inward/record.url?scp=85085338589&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85085338589&partnerID=8YFLogxK
U2 - 10.1016/j.socscimed.2020.113031
DO - 10.1016/j.socscimed.2020.113031
M3 - Article
C2 - 32466849
AN - SCOPUS:85085338589
SN - 0277-9536
VL - 295
JO - Ethics in Science and Medicine
JF - Ethics in Science and Medicine
M1 - 113031
ER -