An eight-state union history model with a transient phase after sexual debut and marriage

To understand how diseases spread through a population, or how families grow and change over generations, scientists must first map the invisible pathways of human connection. Two moments in a person's life are particularly critical for these maps: the age when they first become sexually active, and the age when they first enter a marriage or long-term partnership. These are not just personal milestones; they are the gateways that determine how long a person is exposed to sexually transmitted infections and how long they remain at risk of having children. For decades, researchers have tried to estimate these patterns by looking at survey answers, asking people when they started their sexual lives and when they got married. However, a significant challenge has always been that these answers are often reported separately. A person might say they had their first sexual experience at one age and got married at another, but the time between those two events is often treated as a simple average. This approach misses the complex reality that the risk of getting married changes drastically in the first few years after becoming sexually active, and that the way a marriage ends—whether through death, divorce, or simply staying together—shapes the future risks for that individual.

A team of researchers has now built a new, more complete picture of these life events by treating a person's entire relationship history as a single, flowing story rather than a collection of separate facts. Using data from national surveys in four countries across sub-Saharan Africa—Rwanda, Malawi, Sierra Leone, and São Tomé and Príncipe—they constructed a model that tracks women through eight different stages of life, from never having had sex, to becoming sexually active, to marrying, to potentially divorcing or losing a spouse, and finally to entering a new relationship. By fitting all these transitions together at once, they discovered that the time between a woman's first sexual experience and her first marriage is not a steady, predictable stretch of time. Instead, it is a period with two very different speeds.

The researchers found that immediately after a woman becomes sexually active, the likelihood of her getting married is much higher than it is in the years that follow. In Rwanda, for instance, the rate of marriage in the first year after a woman's sexual debut is more than five times higher than the rate in later years. In Malawi, this initial spike is about twice as high, while in Sierra Leone it is nearly double. This intense period of high marriage rates does not last forever; it fades away over time. On average, this "transient" phase of high activity lasts about three and a half years in Malawi, and between one and two and a half years in the other countries studied. After this initial burst, the rate of marriage settles into a much slower, flatter pattern that continues for the rest of a woman's life. This discovery is crucial because it means that the time a person spends unmarried but sexually active is not a uniform block of time. It is a mixture of a short, intense period right after debut and a much longer, slower period that follows. If models used to predict disease spread or population growth ignore this difference and just use an average, they will misplace the timing of when people are most at risk.

The study also revealed how marriages end and what happens afterward, showing that the story does not stop at a first wedding. In Rwanda, the proportion of women who are widowed grows steadily over time, reaching nearly 23 percent by 27 years after marriage. In contrast, in São Tomé and Príncipe, the proportion of widowed women remains very low, at less than one percent even after 29 years. However, in São Tomé, the rate of women entering a second or later union is remarkably high, with more than half of all first unions being replaced by a new one within 29 years. In Rwanda and Malawi, the proportion of divorced women stays relatively flat over the years, hovering around 13 percent. This happens because the rate at which divorced women find new partners is almost exactly the same as the rate at which new divorces occur. Because of this rapid replacement, a simple snapshot of a population at one moment in time would make it look like fewer marriages are ending than actually are. The cross-sectional view hides the churn of people leaving one marriage and quickly entering another.

The researchers achieved this clarity by creating a model that accounts for the fact that people often report their age at first sex and first marriage as being the same year, or even the same day, even if they happened slightly apart. Their model includes a special "transient" phase that captures this immediate rush into marriage right after sexual debut, as well as a separate group of women who never marry at all. This approach allowed them to fit the data from all four countries simultaneously, producing a consistent set of rules that explain how people move from one state to another. The results show that the patterns of marriage, divorce, and widowhood are distinct for each country, influenced by local history, conflict, and mortality rates. For example, Rwanda experienced a mortality crisis in the mid-1990s, which is reflected in the higher rates of widowhood seen in the data, while Sierra Leone's history of conflict also left its mark on the timing of these events.

The value of this work lies in its ability to provide a complete and consistent set of rates for every stage of a person's union history. Previous methods often estimated the age of first sex, the age of first marriage, and the duration of marriage separately, which could lead to contradictions when these pieces were put together. By treating the whole sequence as one connected chain, the researchers ensure that the proportions of people in every state—married, divorced, widowed, or single—add up correctly. This precision is vital for scientists who build models to predict the spread of infections like HIV or to forecast future population growth. If the timing of when people enter and leave relationships is misunderstood, the predictions for how fast a disease might spread or how many children will be born will be off. The study confirms that the early years after sexual debut are a unique and critical window, and that the way marriages dissolve and reform is a dynamic process that cannot be captured by simple averages. By mapping these complex transitions with such detail, the researchers have provided a clearer, more realistic foundation for understanding human behavior and its consequences for public health and demography.