Awarded Collaborative Faculty Seed Grants, 2010
Analyzing Socio-Demographic and Environmental Data to Better Understand Changing Patterns of Disease
Mark Wilson, Professor of Epidemiology and Professor of Ecology and Evolutionary Biology, University of Michigan
Peter Larson, University of Michigan
Don Mathanga, University of Malawi
More than one third of Africa’s population (~300 million people) currently lives in urban environments, and this proportion is projected to grow to more than half (1.4 billion) by 2030. Underlying this continent-wide trend, however, is enormous regional and national variation, compounded by the very definition of what is actually “urban.” Globally, neither national governments nor international agencies employ a common definition of urban. For sub-Saharan Africa (SSA), urban communities have similarly been defined by population number, density within specific geographic areas, or ensembles of people with governance responsibilities. Definitions aside, health and well-being is largely determined by the environmental, socio-economic, demographic and infrastructural characteristics of human residential patterns. This pilot study will develop innovative analyses of existing socio-demographic and environmental data to better understand how this is changing patterns of disease.
Amid the many diseases afflicting people in SSA, malaria is among the most important, killing perhaps one million children each year. Most efforts to understand and reduce this enormous disease burden have been focused on rural areas, even though urbanization is placing increasingly larger numbers of Africans at risk of malaria. However, the contextual factors of changing urban malaria risk are poorly understood. We propose to investigate these social, environmental and behavioral determinants of malaria risk in and around urban areas of Malawi to better understand fine-scale variation in transmission and space-time patterns of risk.
This research will examine hypothesized social and environmental risks of disease using multivariable and multilevel statistical analysis, satellite images, spatial statistics and simulation modeling at different spatial scales. Resulting knowledge should help focus interventions to more efficiently reduce transmission. This pilot project has two specific aims:
- Characterize social and environmental associations with malaria in various demographic contexts and spatial scales. Using data from previous malaria surveillance, and specific social and
- Create a dynamic, adaptive, agent-based model for simulating malaria transmission and insecticide treated net usage in diverse urban and rural settings.
Overall this research is designed to enhance understanding of the complex interplay of multiple drivers of disease risk across the spectrum of social, economic and behavioral conditions across the urban-rural continuum. We seek to characterize the most important factors affecting disease risk in these diverse environments, and determine which are most amenable to intervention. Results should help to inform prevention policy so as to encourage the most efficient delivery of interventions to those in greatest need. Taken together, the combinations of data being employed and the methods of analysis are truly innovative. Few socio-environmental studies of disease in SSA explore how urbanization is changing the patterns of ill-health and future needs for intervention. Most efforts don’t consider complex, dynamic interactions among qualitatively different kinds of factors which often are considered the domains of different academic disciplines. This pilot work should set the stage for more in-depth studies that will require more extensive effort and external support.
The Pharmacological Correlates of Stillbirth/Birth Asphyxia in Bong County, Liberia
Carol Boyd, Director of Institute for Research on Women and Gender, Deborah J Oakley Collegiate Professor of Nursing and Research Professor in the Substance Abuse Research Center, University of Michigan
Jody R. Lori, Clinical Assistant Professor in the Office of International Affairs, and Nurse Midwifery in the School of Nursing, University of Michigan
Mary Tiah, Director of Nursing, Phebe Hospital, Bong County, Liberia
Liberia has a population of approximately 3 million and lies on the west coast of Africa; it is organized into 12 counties, one of which is Bong County. Liberia’s overall peri-natal mortality rate is approximately 38/1000 pregnancies, while at Phebe Hospital, the primary referral center for Bong County, the peri-natal mortality rate is almost four-fold higher (120/1000).
During a two-to-three week period in February 2010, a team of United States (U.S.) nurses went to Phebe Hospital and Bong County health clinics. Our goals were to expose U.S. graduate nursing students to maternal/child health in sub-Saharan Africa and to explore collaborative research possibilities with Liberia organizations. Three nursing faculty – two from U-M (Boyd and Lori) and one from Madonna (Osher) – accompanied three U-M graduate nursing students. Funding and/or instrumental support came from the U-M Institute for Research on Women and Gender, Africare Liberia and the Bong County Health Department. While at Phebe, it became clear that birth asphyxia (e.g. respiratory depression) is unusually high, as are stillbirth rates within Bong County. Boyd and Osher have experience working with opioid-dependent pregnant women in the U.S. and both nurses noted the similarities between these Phebe neonates and U.S. babies born to opioid users (e.g. users of heroin, morphine, opium, methadone, etc.). One striking difference between the Phebe and U.S. neonates is that U.S. babies go through drug withdrawal because of their prolonged pre-natal exposure; however, after resuscitation, the Phebe babies appear healthy.
Although most laboring mothers denied using any medications prior to labor, several women with a history of stillbirths said they had used a white powder just prior to labor; the medicinal remedy was to help with childbirth and was provided by the traditional healer in their villages. Using plants for medicinal purposes is an important aspect of African culture and tradition (Karou, et al 2006) and up to 80% of Africans depend on traditional medicine, using products extracted from indigenous plants and shrubs (Kirby, 1996). In Liberia, 22% of household spending for health goes to
traditional healers or mobile vendors (Liberia Ministry of Health, October 2009). Based on our observations in February, we concluded: 1) Phebe Hospital and the surrounding Bong County Health facilities have unusually high rates of birth asphyxia and stillbirths and 2) a subset of Bong County women who delivered babies with birth asphyxia acknowledged the use of traditional, medicinal remedies prior to their deliveries. OBJECTIVE: To advance our understanding of the high rates of birth asphyxia/stillbirth in Bong County, Liberia.
The aims of this exploratory study are: 1) To determine the value of using saliva testing in the labor and delivery room at Phebe Hospital in order to determine neonates’ exposure to indigenous substances (e.g. opioid alkaloids, barbituric acid, etc.). 1.a To assess the relationship between Apgar scores and exposure to indigenous substances categorized as CNS depressants or sedative/hypnotics. 2) To establish the feasibility of traveling to villages in Bong County (Liberia) to obtain substances given to pregnant women by traditional healers (and/or family members or friends). 3) To characterize the pharmacological properties of the traditional medicinal remedies that are often given to Bong County women prior to labor.
Drs. Boyd and Lori will travel to Bong County with the plans of staying four weeks at the Phebe Hospital compound. During this month, Boyd and Lori will train Phebe nursing staff to collect saliva samples using the protocols established by the manufacturer. The first 165 patients admitted to the Labor and Delivery unit (L&D) during the four-week period will be included in the study. Phebe nursing staff will be retrained to collect accurate Apgar scores on all neonatal patients. An Apgar score is determined by evaluating the newborn baby on five criteria on a scale from zero to two, then summing the five values. The resulting Apgar score ranges from zero to 10. A short questionnaire will be used to determine the medicines and medicinal remedies used during the last trimester.
Two research assistants from Cuttington University (Bong County, Liberia) will be hired to work with Drs. Boyd and Lori. These research assistants will travel to the villages and collect samples of medicinal remedies from traditional healers. Study Aim #1: To determine the value of using saliva testing in the labor and delivery room at Phebe Hospital in order to determine neonates’ exposure to indigenous substances (e.g. opioid alkaloids, barbituric acid, etc.) . Study Sub-Aim #1a: To assess the relationship between Apgar scores and exposure to indigenous substances categorized as CNS depressants or sedative/hypnotics. To meet study aim #1, Saliva data will provide categorical data indicating a mother’s drug exposure to at least three relevant drug-classes (opioid, barbiturates and sedative/hypnotics). The questionnaire will provide demographic data (including place of residence during the third-trimester), and a history of medication and medicinal remedies used during pregnancy. Apgar data (ordinal), saliva data (categorical) and questionnaire data (descriptive/categorical) will be described and/or correlated to meet the study sub-aim (#1a). Study Aims #2: To determine the feasibility of traveling to villages in Bong County (Liberia) to obtain substances given to pregnant women by traditional healers (and/or family members or friends). To meet study aim #2, if a pregnant woman delivers a baby that exhibits a low Apgar score, or if the child is stillborn or if the woman tests positive forsubstance exposure, a research assistant will re-interview the woman and ask permission to meet in her village and be taken to the healer/family member/friend. The purpose of the village visit is to get a sample of the substance the woman was given. The research assistant will collect a sample (recording all information regarding it). Study Aim #3: To determine the pharmacological properties of the traditional medicinal remedies given to Bong County women prior to labor. To meet study aim #3, Dr. Boyd will bring samples of the medicinal remedies back to the United States for testing in Dr. John Traynor’s laboratory. Catagorical and descriptive data from the laboratory tests will be correlated with previously collected data in order to determine whether a relationship exists between the pharmacological properties of the medicinal remedies used by pregnant women and birth asphyxia/stillbirth.
Sugar Daddies and Empowered Women: the Relationship between Money and Sex
Rebecca Thornton, Ph.D. University of Michigan Economics and Population Studies Center
Lasse Brune, Graduate Student, University of Michigan Economics
Background: Researchers and policy makers have associated the lack of financial resources among women as a determinate of riskier sexual behavior because of the monetary transfers they receive from men (Wojcicki, 2002; World Bank, 2005; Shelton, Cassell, and Adetunji, 2005; Dupas 2009; Wines, 2004; Halperin and Epstein, 2004; Hallman, 2004; Robinson and Yeh, 2009). Similarly, there is a hypothesized positive relationship between male wealth and unsafe sexual behavior because among other things, men with higher incomes can afford to purchase riskier sex (See for example Gertler et al. 2003 and Luke 2009).1 Evidence quantifying the effects of income transfers on sexual behavior among men and women, however, is limited, and often confounded by omitted variables that bias causal estimates.2
Preliminary Work (Kohler and Thornton 2010): In a study of a conditional cash transfer program in rural Malawi, approximately 1300 men and women were tested for HIV and then offered financial incentives of random amounts if they maintained their HIV status for approximately one year. These amounts ranged from zero to values worth approximately four month’s wage. This conditional offer of money had no impact on reported sexual behavior during the year, and no impact on HIV results. However, there were large and significant effects of receiving the money approximately one week after the second round of HIV testing was conducted, after the incentive program had ceased. Men who received money were 13 percentage points more likely to have had vaginal sex and had approximately 0.5 days more of sex. While they were 6.9 percentage points more likely to report using a condom, overall on net there was a 8.5 percentage point increase in risky sex. On the other hand, women were 7.5 percentage points less likely to have engaged in risky sexual sex — driven by abstinence rather than increased condom use.
Identifying Mechanisms for Why Money Matters: There could be several mechanisms through which money affects sexual behavior among both the men and women. One prevailing qualitative and anecdotal theory is that money is used directly by men to purchase risky sex, is used by women to substitute away from ‘selling’ risky sex. However, there are several other possible mechanisms for why money might matter. In the context of the Kohler and Thornton conditional cash transfers study, men receiving the money may have signaled to potential partners (or their wives) that they were HIV-negative. If everyone in the village knew about the incentives program, a man could use the earning of the incentive as an indication that he was not infected. Ironically, this could have resulted in an increase in risky sex. Another mechanism could have been that men who received money significantly increased their purchases on new clothes, transportation, and airtime for communications. It is possible that the money increased their level of attraction or facilitated meeting with their partners. It may 1 The relationship between income and HIV has been studied in other settings. For example, research has found a positive correlation between household assets and HIV or early adult mortality (Shelton, Cassell, and Adetunji 2005; De Walque 2006; Yamano and Jayne 2004). Alternatively, wealthier men have higher returns to safe sex.2 Several exceptions include Duflo et al. who find that Kenyan girls receiving free school uniforms were less likely to become pregnant, Baird et al. 2009 who find that direct payments of secondary school fees lead to significant declines in early marriage, teenage pregnancy, and self-reported sexual activity, and Robinson and Yeh 2006 who find that health shocks lead Kenyan prostitutes to engage in riskier sexual behavior.
Thornton, Rebecca: Sugar Daddies and Empowered Women2 also have simply made their wives happy which increased the amount of sex within their home. For women, there could have also been alternative mechanisms for why money had an impact on sexual behavior. After being given money, women may have had less leisure time and thus reduced the amount of sex that they had. They may have also reduced the amount of sex they had with their husbands, for example.
Broader Research Goal: The broader goal of the project is to first conduct a smaller, in-depth pilot; and second to apply for larger funding (likely with an R01 NIH mechanisms). This would build in collaboration with University of Michigan faculty and students as well as faculty and students from African institutions Research Design: The purpose of this pilot is to combine both qualitative and quantitative methods to begin to understand the exact mechanisms through which money matters for sexual behavior. This would involve field testing both survey instruments and methodologies in order to determine which particular methods are most effective for measuring sexual behavior and spending – both sensitive survey topics. In particular, the pilot would consist of field testing qualitative and quantitative survey instruments to measure daily expenditures and sexual behavior as well as conducting in-depth interviews with male and female respondents on the mechanisms of cash affecting sexual behavior. Funding will also cover field testing logistics for a series of randomized experiments that will measure the effect of receiving money. In addition, this pilot will allow for exploring relationships between Michigan researchers and African social scientists to determine the best collaborations for the larger proposal.
Partnerships: The pilot project would explore a partnership between faculty at the University of Michigan and researchers affiliated at IFORD in Cameroon (l’Institut de Formation et de Recherche Démographiques) with linkages with Parfait Eloundou-Enyegue, Ph.D. Cornell University Sociology and IFORD. Eloundou-Enyegue will be involved in the design and field testing, especially focusing on the qualitative component. The Michigan study team will consist of the following:
- Rebecca Thornton, Ph.D. University of Michigan Economics and Population Studies Center – Thornton is the lead PI who will be involved in the design and field testing, especially focusing on the randomized quantitative component.
- Lasse Brune, Graduate Student, University of Michigan Economics – Brune will be involved in the field testing and design of the project
Toxic Metal Exposure and Women of Reproductive Age in Ghana
Dr. Niladri Basu (University of Michigan) and Dr. Mawuli Dzodzomenyo (University of Ghana)
There is building evidence that millions of children worldwide have impaired neurodevelopmental disorders and other long-term health effects resulting from early-life exposures to complex mixtures of toxic chemicals (especially metals such as lead, arsenic, and mercury), and that such impacts affect the social well-being and economic prosperity of nations.1 In Africa, specifically, a recent position paper highlights that certain sub-groups are exposed to elevated levels of toxic chemicals and that associated health outcomes are magnified when other vulnerability factors, such as poverty and malnutrition, are considered.2 Recent data suggest that Ghanaians are exposed to elevated concentrations of several toxic metals3,4 which pose major concern during particularly susceptible time periods, such as pregnancy. Unfortunately, a focused study on predictors of multiple metals exposure to women of reproductive age has not yet been conducted in Ghana.
The long-term objective of this proposed research is to establish a longitudinal birth cohort to further build upon the long standing collaborative research with our Ghanaian colleagues and help to develop a greater understanding, within the social context of Ghana, of prenatal metal-metal interactions and associations to adverse children’s health outcomes. Specific to this proposal, in the short-term, our objective is to characterize toxic metals exposure in Ghanaian women of child bearing age in three distinct, socially vulnerable, populations: A) a small-scale gold mining community that we started to work with in the Talensi-Nabdam District outside of Bolgatanga; B) a subsistence fishing community on the Gulf of Guinea; and C) urban residents in Accra. The outcome of this work will enhance our understanding of multiple metals exposures in Ghana, help shape future research goals with respect to environmental and social health, and further a growing collaboration between U-M and Ghanaian researchers and their students.
Context of Work: Over the past two years PI-Basu has worked closely with Prof. Elisha Renne (Department of Anthropology, University of Michigan) on issues related to the health and environment of small scale miners in Ghana. With funding from ASC and OVPR, a comprehensive epidemiological study was performed during the summer of 2009 on 120 individuals associated with small-scale mining in the Talensi-Nabdam District. The initial findings from that work documented high levels of mercury exposure among some miners, and even among individuals (including some women of child bearing age) not directly associated with the mining and extraction process. To build upon these findings, in March 2010, PI-Basu spent one week in Ghana where planning meetings were held with Dr. Mawuli Dzodzomenyo (Professor, School of Public Health, University of Ghana-Legon), other Faculty (Prof. Osman Al-Hassan, Institute of African Studies, University of Ghana; Prof. Kenneth Pelig-Ba, University of Development Studies-Navrongo), and government officials (Wilson Waanab Zoogah, Simon Attebiya, Minerals Commission of Ghana). Out of these meetings a consensus was reached to collaborate with Dr. Dzodzomenyo and expand the scope of the previous work done by Drs. Renne, Basu, and Pelig-Ba to include exposure assessment at multiple small scale mining sites throughout the country, as well as other susceptible sub-groups such as coastal inhabitants that may be at risk from mercury exposure through fish consumption and urban residents that may be exposed to multiple toxic metals. The goal of this current proposal is to help initiate this larger study.
Project Activities: There are three distinct objective/activities for this proposed research. Objective one will be to establish and build further collaborations among researchers and community members at additional small-scale gold mining sites, coastal communities, and within Accra. Working with Dr. Dzodzomenyo, Drs. Basu and Cantonwine will identify key personnel to collaborate with in hopes of building the capacity to carry out longer term relevant environmental health research and training. Objective two will be to train Dr. Dzodzomenyo’s students and on-site personnel in advanced exposure assessment techniques in order to initially recruit and perform a detailed exposure assessment upon a minimum of fifty women of childbearing age at each distinct location. This will be achieved through two established programs. Every year (Apr – June), Dr. Dzodzomenyo has a group of 6-10 MPH students from the University of Ghana that visit various sites across the country to conduct environmental health research, and Profs. Renne and Basu have 4-6 students conduct focused work near Bolgatanga at a small-scale mining site. Objective three will be to work with and guide Dr. Dzodzomenyo’s students in analysis to compare and contrast exposure patterns within each location to help assist in identifying at risk populations from metal exposure and disseminate knowledge to participants, educators, lay people, and government officials. The outcomes of this project include: A) increased understanding of metals exposures in women of reproductive age in Ghana; B) increased awareness of environmental health risks in Ghana by students, faculty, laypeople, and government officials; C) increased ability of U-M and Ghana public health researchers and students to collaborate and address complex environmental health issues; and D) initial steps towards the establishment of a longitudinal birth cohort that will enable us to improve understanding of prenatal exposures to toxic chemicals (focus on metals) and associations to adverse health outcomes, especially within the social context of Ghana.
Project Rationale: In Africa, there is growing concern that many are exposed to elevated levels of toxic chemicals, and that such exposures when considered alongside other vulnerability factors, such as poverty and malnutrition, result in synergistic adverse health outcomes. Our preliminary data on small-scale miners has indicated high exposures to mercury, and steps are being taken by PI-Basu and U-M Prof. Renne to address the long-term health and social consequences associated with such exposures. Here, we propose to increase understanding of other environmental health risks in Ghana, and to advance the growing relationship between PIs Basu, Dzodzomenyo and additional Ghanaian faculty and students. This will be achieved by taking the framework used by PI-Basu and his collaborators to study small-scale gold miners near Bolgatanga, and apply to other susceptible sub-groups in Ghana, namely other small-scale gold mining communities, subsistence fishing communities along the Gulf of Guinea, and urban residents.
Partnerships: The partnership described below is founded on a larger, +20-yr collaboration among the University of Michigan, the Ghanaian Ministry of Health, the University of Ghana, the Kwame Nkrumah University of Science and Technology, and through the new learning grant proposal funded by the Bill and Melinda Gates Foundation. Dr. Mawuli Dzodzomenyo is the chair of the Department of Biological, Environmental, and Occupational Health Sciences at the University of Ghana in Accra. His research specializations, which will contribute to the proposed project, include the recruitment and training of University of Ghana MPH students to perform qualitative and quantitative health assessments, overseeing environmental quality studies in each site, and help disseminate information to government officials and the general public. He will work closely with Drs. Niladri Basu and David Cantonwine on selecting and developing the recruitment sites. Dr. Niladri Basu is an Assistant Professor in the Environmental Health Sciences Department at the University of Michigan’s School of Public Health. Dr. Basu is a specialist in the health and environmental consequences of mercury on a global level. Recently, he has worked closely with Dr. Elisha Renne and her students to address the risks associated with mercury use in a small-scale gold mining site in Ghana. Dr. David Cantonwine is a Post-Doctoral Fellow in the Department of Obstetrics and Gynaecology at the University of Michigan Medical School. Dr. Cantonwine is a prenatal environmental epidemiologist with a research focus on metal-metal interactions, genetic susceptibility, and adverse birth outcomes. He will work the Drs. Basu and Dzodzomenyo to foster new collaboration and oversee the environmental assessment at each location.