Our investigation shows that a large decline in estimated population immunity was observed before a 2017 increase in cases of monkeypox and, based on this, postulate a relationship between decreased immunity to smallpox and resurgence of monkeypox in Nigeria. The potential role of declining population immunity in the resurgence of monkeypox has been raised in earlier studies (4,6,7,20). Epidemiologic evidence suggests previous smallpox vaccination provides at least partial protection against severe MPXV infections (13,20), further supported by immunologic studies of smallpox vaccine. Residual IgG and neutralizing antibodies were shown to persist in vaccinated persons (21–23) and have been associated with milder disease among infected patients (24). Among US monkeypox patients, those vaccinated for smallpox displayed evidence of vaccination immunity (orthopoxvirus [OPXV] IgG and memory B cells) after monkeypox exposure (24). Smallpox vaccine induces both humoral and cell-mediated response against OPXV, including MPXV, targeting a wide range of viral particles and preventing viral replication (23,25).

Our results show that the effect of a decline in individual-level immunity among vaccinated persons, as well as population growth in the postvaccination era, has substantially reduced the overall population immunity level within the past 45 years. The median age of the patients was 29 years old (2), notably higher than for previous outbreaks except from the 2017 outbreak in Central African Republic (median 27.5 years of age) and a single case in Sierra Leone in 1970 (27.5 years old) (4). This finding can be explained by the fact that children too young to get vaccinated in the 1970s have grown up and now form most of the contemporary susceptible population. The smallpox vaccination campaign officially ceased in 1980; by 2017, when the monkeypox outbreak in Nigeria occurred, the unvaccinated cohort would encompass all residents <37 years of age. This contemporary susceptible population is composed mainly of working adults who maintain wider social contact and are more likely to engage in activities that include risk of animal exposures, such as hunting, farming, or trading bush meat (26). In addition, the expanding unvaccinated population means that entire households are now susceptible to monkeypox instead of just children, which enhances the risk of human-to-human transmission. In fact, the index case in 2017 was part of a 5-member family cluster of cases (27).

Most confirmed cases were concentrated in the southern zones, which are characterized as natural ecologic niches of monkeypox because of swamps and rain forests (2,4). Satellite imagery during 2000–2016 shows a substantial increase in built-up areas and farmland in southern Nigeria, created at the expense of these forested areas (28). This expansion of developed areas increases the likelihood of reservoir animals, such as rodents, rabbits, and primates, being displaced from their natural habitat and living among humans, thus increasing interspecies contact (29). Past serologic surveys found higher seroprevalence of OPXV-specific IgG among residents of forested habitat, suggesting frequent exposure to MPXV and other OPXV (5,30,31). This evidence is further supported by the disproportionate prevalence among men in this outbreak (male:female ratio = 3:1), because predominantly men perform most high-risk occupations dealing with wild animals, such as hunting and trading bush meat (4,32). In addition, expansion of urban transport networks may have contributed to widespread transmission in this outbreak, because states with >10 confirmed cases tended to be converging points for major roads (Figure 2, panel B).

Of note, an increasing number of cases were detected in drier savannahs in the northern zones, which are not typical ecologic niches of MPXV (2). This finding is possibly because more animal-human interfaces are occurring outside of MPXV natural habitats because of savannah being clearing for farming and settlement. In fact, savannah-to-agricultural land transition constituted the largest segment of land conversion in Nigeria during 1975–2013 (33). Moreover, interstate railway lines and highways may have enabled patients from monkeypox clusters to travel north from southern locations and subsequently infect local residents.

Several models have conceptualized zoonotic transmission as a multistage process with several bottlenecks that can influence the probability of spillover (8,34). In these models, host-specific and pathogen-specific factors determine how many pathogens are released into the environment and how long they survive. Individual human behaviors determine the probability and dose of exposure; individual human physiology and immunity determine the probability and severity of infection upon exposure (8). In other words, although urbanization and land conversion increase the frequency of animal exposure and the average exposure dose, human immunity potentially opposes this effect by lowering the probability of infection. At the same time, although smallpox vaccination may provide partial protection, sufficiently large infectious inoculum, through prolonged or frequent animal contact, can overcome such protection and manifest symptomatically (24,35).

Although no cases were reported in Nigeria during 1978–2017, because of the high prevalence of smallpox vaccination among the 1970s cohort, mild and asymptomatic infections might have occurred but gone unreported. In addition, the West Africa clade is associated with lower virulence (36), which could have enabled the disease to spread through mild or asymptomatic cases not captured by passive surveillance. In fact, before the 2017 outbreak, monkeypox was not in the Integrated Disease Surveillance and Response system list of reportable diseases (37). Serologic surveys of West Africa populations revealed active levels of IgG suggestive of routine exposure to OPXV, albeit without patients recalling symptoms or having scars (6,30). The resurgence of monkeypox in Nigeria in 2017, although seemingly unprecedented, may be the result of alignment of several control gaps in the spillover process, driven by a combination of factors: modern urbanization, urban densification, waning of immunity among vaccinated residents, and accumulation of unvaccinated cohorts.

This study is subject to some limitations because our model was built on several assumptions. We assumed that the base population in 1970 started with a uniform immunity level of 85%; in reality, persons vaccinated before 1970 would have had a lower immunity level at the start of the model and persons vaccinated from 1970–1980 would have started with a higher immunity level at a later year. We assumed that 77.2% vaccination coverage was uniform across all states, but our uniform vaccination coverage and protection levels represent a simplified averaging of heterogeneous rates of coverage across states. Finally, for our model, we assumed that no vaccination campaigns occurred in Nigeria after 1970. In fact, several vaccination campaigns were conducted during 1969–1980 in Nigeria (38), but there was insufficient data on these campaigns’ frequency and coverage to accurately quantify their effects on the population immunity level. The model also did not account for changing kinetics of antibodies and T-cells in persons receiving a booster dose (39).

Next, in the absence of state-specific population growth rates, we were unable to simulate rural-urban migration in the model, which resulted in an urban-rural growth differential and could lead to differential increase in the susceptible population between states (40,41). However, accurately estimating this effect would require expanding the parameters of a future model to account for population migration between states, data that are not publicly available. Last, the estimated rate of serologic immunity decline we used had a wide confidence interval in the source study (13); that would have increased the margin of error for our estimates of individual and population immunity levels. The model would benefit from future studies that more accurately estimate rates of immunity decline.

The wide geographic spread of the 2017 outbreak in Nigeria was likely driven by the lower level of residual OPXV immunity, population growth, an increase in the proportion of susceptible persons, and potential spillover events at the animal-human interfaces caused by human settlements encroaching into forested areas. The initial spillovers may have been followed by rapid human-to-human transmission enabled by high population density and a growing immunologically naive population fully susceptible to MPXV. High prevalence among working adults 21–40 years of age, born after universal vaccination programs were discontinued, suggests that declining population immunity plays a substantial role in the reemergence of monkeypox.

Fewer monkeypox cases were diagnosed in 2020, which other researchers have attributed to the self-limiting nature of MPXV human transmission (e.g., because of nonairborne mode of transmission, low probability of infection per contact) (4,42). However, we cannot rule out the possibility of future mutations that might enable sustained human-to-human transmission or adoption of more cosmopolitan animal reservoir hosts. Such occurrences would present substantial public health risks. These ongoing risks highlight the importance of serosurveillance to understand the extent of OPXV endemicity within the population. The role of vaccination in preventing monkeypox is being considered, and clinical trials for healthcare workers are underway (43,44). In the absence of seroprevalence data in Nigeria, this study provides an alternative method to estimate the residual level of vaccine immunity and adds another perspective to the discourse on monkeypox reemergence in West Africa.