DGHI Member M. Giovanna Merli discovered demography in a roundabout way. “My undergraduate major was in China studies, with an emphasis on contemporary Chinese politics and history. I spent three years in China, and in my third year, I discovered demography. I took a class in population and family planning in China. That was when I discovered the depth and intricacies of the explanations for population outcomes such as fertility, and how policy and politics are implicated in these personal decisions in the unique Chinese context.”
That discovery prompted her to complete an M.A. and Ph.D. in Demography from the University of Pennsylvania, and an M.A. in International Relations from the Paul H. Nitze School of Advanced International Studies at the Johns Hopkins University. She joined the Sanford School of Public Policy at Duke in 2008 as an associate professor.
Merli combines her passion for demography and her love of China in her most recent work – examining the social patterns that create conditions that put individuals in China at risk of HIV infection.
Q: What led you into researching the spread of HIV in China in particular?
Merli: In 2002, UNAIDS put out a report on “China’s Titanic Peril.” They projected that 10 million people in China would be infected with HIV by 2010. But we are not seeing such a rapid rise—HIV prevalence is still estimated at very low levels in China. I began to wonder whether concerns about an explosion of HIV infections in China – the world’s most populous country – are misplaced.
I started tinkering with the idea of trying to project the spread of HIV with a simulation model that we could adjust to fit the Chinese profile of sexual behavior. It turns out that in 2002, the results of the first nationwide population-based survey of sexual behaviors ever conducted in China were released. This survey was conducted by the University of Chicago in collaboration with the People’s University of China, and provided information that allowed us to estimate the parameters that drive the simulation.
The model shows a very high sensitivity of simulated HIV prevalence to assumptions about the number of sexual partners and the patterns of sexual mixing.
Based on these simulations, and what we know about sexual behavior in China, my hunch is that the levels of sexual activity and the patterns of sexual mixing in China are not as conducive to the rapid spread of HIV as they may be in other countries or settings.
Q: Why is sexual mixing such an important factor?
Merli: When the prevalence of HIV is relatively high in a given population, the chance of being sexually active with an infected person is also quite high. But in a low-prevalence setting, the chance of being infected is highly dependent on the characteristics of your partners. If you only have sex with people who are not likely to be infected, even if you have many such partners, your chance of infection is lower. If people in a high risk group, such as injecting drug users or female sex workers, are highly sexually active, but mixing only with others who are also in this high risk group, the spread of HIV will remain concentrated in this group.
We know from previous data that Chinese men and women are more likely to report only one partner within the last 12 months than American men and women, and less likely to report having multiple partners. What we don’t know are the patterns of sexual mixing in China.
Q: How are you gathering that information?
Merli: In March 2008, I completed a survey of sexual behavior and sexual networks in Shanghai, China. We surveyed a probability sample of 2,200 individuals, representative of the Shanghai population of urban residents and rural migrants, and did a separate survey of female sex workers. We asked respondents about their own behavior and the sexual behavior of their partners, and attributes of these partnerships. We will be using the information to inform more realistically the simulations of the spread of HIV, assess the role of characteristics of sexual networks for the spread of HIV, and verify our hypothesis that the Chinese regime of sexual relations cannot sustain a generalized HIV/AIDS epidemic
Q: What other areas of global health are you involved in?
Merli: As part of my work in China, I am also involved in collaborations with statisticians at Wisconsin and public health researchers at UNC to evaluate the statistical properties of various methods of sampling female sex workers and other hard-to-reach populations.
I also study HIV/AIDS in South Africa, where the AIDS morbidity and mortality crises are tantamount to a perturbation of the age structure. Since HIV/AIDS in South Africa mostly affects individuals in the mid-adult ages, my work focuses on the consequences of this mortality and morbidity crisis for families and households.
I have determined the burden of illnesses and death that the South African elderly must endure as a consequence of their children becoming ill or dying from AIDS. The findings document massive shifts in family structure.
I would like to evaluate the consequences of increases in AIDS morbidity and mortality in South Africa with a simulation of the spread of HIV/AIDS, kin availability, and co-residence, but for this, I need much more empirical information.
Looking ahead, with the rising burden of chronic diseases in China, I am eager to extend work to a multipurpose data collection effort to explore how rapid changes in the economy and society are leading individuals to adopt behaviors that put them at risk of acquiring chronic diseases, especially cardiovascular diseases and cancers.
Q: What drew you to Duke?
Merli: I am a demographer, and there are many others at Duke whose research pertains to populations, whether they public policy scholars, economists, sociologists, biologists, or health scientists. There is also lot of interest and élan about working internationally at Duke; the Duke Global Health Institute is an example of that. And Duke is located in an interesting area when it comes to the study of population and health in international settings. I have colleagues at UNC, and companies like Family Health International and RTI are nearby. It is a very vibrant place with lots of people engaged in the study of population and public health.
I was sorry to leave [the Department of Sociology at the University of] Wisconsin, but it was exciting to be pulled into a new job with such opportunity.
For more details on Dr. Merli’s work, view the powerpoint of her presentation at the September 2008 Duke Center for AIDS Research (CFAR) scientific retreat and her Public Policy webpage at http://fds.duke.edu/db/aas/PublicPolicy/faculty/giovanna.merli
