Martina Morris

Concurrent partnerships and the spread of HIV.

Objective:To examine how concurrent partnerships amplify the rate of HIV spread, using methods that can be supported by feasible data collection. Methods:A fully stochastic simulation is used to represent a population of individuals, the sexual partnerships that they form and dissolve over time, and the spread of an infectious disease. Sequential monogamy is compared with various levels of concurrency, holding all other features of the infection process constant. Effective summary measures of concurrency are developed that can be estimated on the basis of simple local network data. Results:Concurrent partnerships exponentially increase the number of infected individuals and the growth rate of the epidemic during its initial phase. For example, when one-half of the partnerships in a population are concurrent, the size of the epidemic after 5 years is 10 times as large as under sequential monogamy. The primary cause of this amplification is the growth in the number of people connected in the network at any point in time: the size of the largest ‘component’. Concurrency increases the size of this component, and the result is that the infectious agent is no longer trapped in a monogamous partnership after transmission occurs, but can spread immediately beyond this partnership to infect others. The summary measure of concurrency developed here does a good job in predicting the size of the amplification effect, and may therefore be a useful and practical tool for evaluation and intervention at the beginning of an epidemic. Conclusion:Concurrent partnerships may be as important as multiple partners or cofactor infections in amplifying the spread of HIV. The public health implications are that data must be collected properly to measure the levels of concurrency in a population, and that messages promoting ‘one partner at a time’ are as important as messages promoting fewer partners.

Birds of a feather, or friend of a friend? using exponential random graph models to investigate adolescent social networks*

In this article, we use newly developed statistical methods to examine the generative processes that give rise to widespread patterns in friendship networks. The methods incorporate both traditional demographic measures on individuals (age, sex, and race) and network measures for structural processes operating on individual, dyadic, and triadic levels. We apply the methods to adolescent friendship networks in 59 U.S. schools from the National Longitudinal Survey of Adolescent Health (Add Health). We model friendship formation as a selection process constrained by individuals’ sociality (propensity to make friends), selective mixing in dyads (friendships within race, grade, or sex categories are differentially likely relative to cross-category friendships), and closure in triads (a friend’s friends are more likely to become friends), given local population composition. Blacks are generally the most cohesive racial category, although when whites are in the minority, they display stronger selective mixing than do blacks when blacks are in the minority. Hispanics exhibit disassortative selective mixing under certain circumstances; in other cases, they exhibit assortative mixing but lack the higher-order cohesion common in other groups. Grade levels are always highly cohesive, while females form triangles more than males. We conclude with a discussion of how network analysis may contribute to our understanding of sociodemographic structure and the processes that create it.

Measures of concurrency in networks and the spread of infectious disease

An investigation is made into the impact of concurrent partnerships on epidemic spread. Starting from a definition of concurrency on the level of individuals, the authors define ways to quantify concurrency on the population level. An index of concurrency based on graph theoretical considerations is introduced, and the way in which it is related to the degree distribution of the contact graph is demonstrated. Then the spread of an infectious disease on a dynamic partnership network is investigated. The model is based on a stochastic process of pair formation and separation and a process of disease transmission within partnerships of susceptible and infected individuals. Using Monte Carlo simulation, the spread of the epidemic is compared for contact patterns ranging from serial monogamy to situations where individuals can have many partners simultaneously. It is found that for a fixed mean number of partners per individual the distribution of these partnerships over the population has a major influence on the speed of the epidemic in its initial phase and consequently in the number of individuals who are infected after a certain time period.

Concurrent partnerships and transmission dynamics in networks

Abstract Network structure plays a critical role in channeling person-to-person transmission of infectious disease agents like HIV. Stochastic simulations are used here to examine the effect of two structural attributes: concurrent partnerships (degree) and mixing patterns (degree-based homophily bias). Four alternative partnership scenarios are compared: (1) sequential monogamy, and concurrent partnerships with (2) random (3) assortative and (4) disassortative mixing by degree. Concurrent partnerships are found to dramatically increase the speed and pervasiveness of epidemic spread, and the implications for HIV transmission are discussed.

Concurrent Partnerships and HIV Prevalence Disparities by Race: Linking Science and Public Health Practice

Concurrent sexual partnerships may help to explain the disproportionately high prevalence of HIV and other sexually transmitted infections among African Americans. The persistence of such disparities would also require strong assortative mixing by race. We examined descriptive evidence from 4 nationally representative US surveys and found consistent support for both elements of this hypothesis. Using a data-driven network simulation model, we found that the levels of concurrency and assortative mixing observed produced a 2.6-fold racial disparity in the epidemic potential among young African American adults.

Relative distribution methods in the social sciences

and Motivation.- The Relative Distribution.- Location, Scale and Shape Decomposition.- Application: White Mens Earnings 1967-1997.- Summary Measures.- Application: Earnings by Race and Sex: 1967-1997.- Adjustment for Covariates.- Application: Comparing Wage Mobility in Two Eras.- Inference for the Relative Distribution.- Inference for Summary Measures.- The Relative Distribution for Discrete Data.- Application: Changes in the Distribution of Hours Worked.- Quantile Regression.

Women's Gains or Men's Losses? A Closer Look at the Shrinking Gender Gap in Earnings

The recent closing of the gender wage gap is often attributed to increases in womens human capital. This explanation neglects the effect of growing inequality in mens earnings. The authors develop a decomposition that allows them to test how distributional changes in mens and womens earnings combine to yield changes in womens economic status. Using Current Population Survey data from 1967 to 1987, the authors find that the striking polarization in white mens earnings has played a critical role in generating womens relative economic gains, though more for white women than for black women. For both groups, the results predict a future slowing of womens relative progress.

The prevalence of trichomoniasis in young adults in the United States

Background and Objectives: The prevalence of trichomoniasis in the general population of the United States is unknown. This study provides the first population-based prevalence estimates of trichomoniasis among young adults in the United States. Methods: The National Longitudinal Study of Adolescent Health (Add Health) is an ongoing prospective cohort study. In a cross-sectional analysis of Wave III of Add Health (N = 12,449), we determined the prevalence of trichomoniasis using a polymerase chain reaction assay. Results: The estimated overall prevalence of trichomoniasis in U.S. young adults was 2.3% (95% confidence interval [CI], 1.8–2.7%). The prevalence was slightly higher among women (2.8%; 95% CI, 2.2–3.6%) than men (1.7%; 95% CI, 1.3–2.2%). The prevalence increased with age and varied by region, with the south having the highest prevalence (2.8%; 95% CI, 2.2–3.5%). The prevalence was highest among black women (10.5%; 95% CI, 8.3–13.3%) and lowest among white women (1.1%; 95% CI, 0.8–1.6%). Among men, the prevalence was highest among Native Americans (4.1%; 95% CI, 0.4–29.3%) and blacks (3.3%; 95% CI, 2.2–4.9%), and lowest among white men (1.3%; 95% CI, 0.9–1.8%). Conclusions: Trichomoniasis is moderately prevalent among the general U.S. population of young adults and disturbingly high among certain racial/ethnic groups.

Epidemiology and Social Networks

The spread of a disease through human populations can be significantly altered by patterned networks of social contact. Largely in response to AIDS, the way in which social networks channel disease has recently become the focus of a sustained modeling effort in epidemiology. The challenge has been to develop a general framework capable of representing both simple and arbitrarily complicated mixing structures, and of solving the matching problem in a non-equilibrium multi-group population. This work is reviewed here. A recent contribution from network analysis links log-linear models of contact structure to diffusion equations for transmission. This framework is described in detail, and some applications to the spread of AIDS are described. The results show that careful analysis of structured mixing can reveal significant features of an epidemic that would otherwise be missed.

Timing is everything: international variations in historical sexual partnership concurrency and HIV prevalence.

Background Higher prevalence of concurrent partnerships is one hypothesis for the severity of the HIV epidemic in the countries of Southern Africa. But measures of the prevalence of concurrency alone do not adequately capture the impact concurrency will have on transmission dynamics. The importance of overlap duration and coital exposure are examined here. Methodology/Principal Findings We conducted a comparison of data from three studies of sexual behavior carried out in the early 1990s in Uganda, Thailand and the US. Using cumulative concurrency measures, the three countries appeared somewhat similar. Over 50% of both Thai and Ugandan men reported a concurrency within the last three partnerships and over 20% reported a concurrency in the last year, the corresponding rates among US men were nearly 20% for Blacks and Hispanics, and about 10% for other racial/ethnic groups. Concurrency measures that were more sensitive to overlap duration, however, showed large differences. The point prevalence of concurrency on the day of interview was over 10% among Ugandan men compared to 1% for Thai men. Ugandan concurrencies were much longer duration – a median of about two years – than either the Thai (1 day) or US concurrencies (4–9 months across all groups), and involved 5–10 times more coital risk exposure with the less frequent partner. In the US, Blacks and Hispanics reported higher prevalence, longer duration and greater coital exposure than Whites, but were lower than Ugandans on nearly every measure. Together, the differences in the prevalence, duration and coital exposure of concurrent partnerships observed align with the HIV prevalence differentials seen in these populations at the time the data were collected. Conclusions/Significance There were substantial variations in the patterns of concurrent partnerships within and between populations. More long-term overlapping partnerships, with regular coital exposure, were found in populations with greater HIV epidemic severity.