The research, published as an open access study in the International Journal for Equity in Health, is an ecological analysis covering every district in the territory over fifteen consecutive years. Rather than relying on simple crude death counts, the team built annual period life tables for each district, the demographic gold standard for estimating life expectancy at birth. The methodological centerpiece is a Bayesian adaptation of a model known as TOPALS, short for Tool for Projecting Age-Specific Rates using Linear Splines. TOPALS works by modeling age-specific mortality rates as smooth spline functions relative to a standard mortality schedule, which stabilizes estimates in small populations where the number of deaths at individual ages can be sparse and noisy. The Bayesian layer adds probability distributions around every estimate, allowing the researchers to quantify uncertainty rather than pretending each district figure is known with perfect precision. From the posterior median estimates of life expectancy at birth, the team then computed how widely the district values spread around their average.

The headline numbers are striking. The median district-level life expectancy at birth climbed from 82.3 years in 2010 to 86.4 years in 2024, a gain of more than four years in a decade and a half, achieved in a population that was already among the longest-lived on Earth. That trajectory, however, was not a smooth upward slope. The study documents a territory-wide decline in life expectancy during the 2022 Omicron wave, when Hong Kong experienced one of the deadliest per-capita COVID-19 outbreaks of any wealthy society, driven largely by deaths among older residents with low vaccination coverage at the time. The demographic shock is clearly visible in the reconstructed life tables, demonstrating how sharply an epidemic wave in an aging city can carve years off a summary mortality indicator that usually creeps upward by fractions of a year.

Perhaps the most consequential finding concerns inequality between districts. The researchers measured dispersion using the standard deviation of district-level life expectancy estimates, a straightforward statistic with a powerful interpretation: the larger the standard deviation, the bigger the gap between the longest-lived and shortest-lived neighborhoods. Before the pandemic, that gap widened steadily. Among females, the standard deviation of life expectancy at birth doubled from 1.4 years to 2.9 years by 2020. Among males it nearly tripled, rising from 1.0 year to 2.9 years over the same span. In practical terms, by 2020 a woman’s expected lifespan could differ by several years depending on which of Hong Kong’s eighteen districts she called home, a gap comparable to the difference between some entire countries.

The pandemic then reshaped that landscape in unexpected ways. Dispersion declined in 2021 and, after the devastating Omicron wave of 2022, remained lower than its peak. That pattern is demographically intriguing. Pandemics often compress or distort geographic inequality in mortality because the hardest-hit communities can be scattered rather than concentrated, and because the elderly, who dominate COVID-19 deaths, are distributed unevenly across districts in ways that do not track general socioeconomic gradients. The authors’ data suggest that Hong Kong’s district-level inequality in longevity peaked around 2020 and then narrowed, a result that resists any simple narrative in which the pandemic uniformly amplified every pre-existing health disparity.

One district stands out for persistent disadvantage: Yau Tsim Mong. This dense cluster of neighborhoods on the Kowloon peninsula, home to some of the oldest housing stock and most socially diverse populations in the city, remained persistently at the bottom of the longevity rankings across the study period. The persistence matters. While overall dispersion narrowed after 2020, the fact that one district stayed disadvantaged for years signals a structural vulnerability rather than random fluctuation. In a compact territory served by a single health system, a district that lags consistently in life expectancy points to place-based factors, housing conditions, income composition, access to services, or population churn, that island-wide averages cannot capture.

Just as interesting is what the spatial analysis did not find. Using Moran’s I, a standard statistic for detecting spatial autocorrelation, the team searched for geographic clustering of high or low life expectancy and found no clear pattern. In many cities, longevity maps form broad contiguous belts, with wealthy suburbs ringed by poorer peripheries or an affluent core surrounded by struggling outer boroughs. Hong Kong defies that template. Its long-lived and short-lived districts are interwoven rather than segregated into coherent regions, a fragmentation likely shaped by the territory’s extreme vertical density, where public housing estates, luxury towers, and aging walk-up buildings can sit within a few hundred meters of one another. Inequality, in other words, is localized, not zonal.

When the researchers explored associations between district life expectancy and a set of structural and social determinants, most conventional candidates failed to show a statistically detectable relationship. The single exception was residential mobility: the five-year in-migration rate of a district showed a statistically detectable association with life expectancy at birth, with an estimated effect of 1.141 years, and a 95 percent credible interval running from 0.211 to 2.071 years. Districts that received more new residents over five years tended to have higher life expectancy, even after accounting for uncertainty in the Bayesian estimates. The finding is exploratory and correlational, and the authors are careful about interpretation, but it is demographically plausible. In-migration into a district often selects for younger, healthier, and more economically mobile individuals, a classic healthy-migrant effect, while districts with little inflow may be aging in place or losing their younger residents. Mobility, in this reading, is not a cause of longevity so much as a fingerprint of who ends up living where.

The study carries implications well beyond Hong Kong. For one, it demonstrates a technical template that other small-area demographers can adopt: Bayesian TOPALS modeling produces stable, uncertainty-aware life tables for small populations, making annual district-level surveillance of mortality feasible even where death counts are modest. For another, it challenges the assumption that a national or citywide life expectancy figure tells residents what they can expect. In Hong Kong, the spread between districts has at times been nearly three standard deviation years wide, meaning the city’s world-record longevity is genuinely experienced differently across neighborhoods. The authors argue that annual district-level monitoring and place-based assessment are needed to identify persistent local vulnerability before it becomes entrenched, rather than discovering it years later in retrospect.

There is also a subtle scientific payoff in the pandemic-era data. Because the study covers 2010 to 2024, it captures a full pre-pandemic trend, the sharp Omicron reversal, and the partial recovery afterward, all at fine geographic resolution. That temporal depth shows how fragile even the most favorable mortality trends can be, and how quickly a single epidemic wave can erase years of progress while simultaneously reshaping the geography of survival. As populations worldwide age and dense cities remain the dominant form of human settlement, the Hong Kong case suggests that longevity records will increasingly be contested not just between countries but within them, block by block. A city that adds more than four years of expected life in fifteen years can still harbor neighborhoods where residents fall years short of their neighbors, and only sustained, statistically sophisticated surveillance makes that hidden inequality visible in time to act.

Subject of Research: Spatial inequality in district-level life expectancy across Hong Kong from 2010 to 2024

Article Title: Spatial inequalities of life expectancy in Hong Kong, 2010–2024

Article References: Chen, X., Chen Sun, C., Cao, Y., Chan, S. S., & Bishai, D. (2026). Spatial inequalities of life expectancy in Hong Kong, 2010–2024. International Journal for Equity in Health. https://doi.org/10.1186/s12939-026-03036-1

Image Credits: AI Generated

DOI: 10.1186/s12939-026-03036-1

Keywords: Hong Kong, life expectancy, health inequality, Bayesian TOPALS, spatial demography, Moran's I, COVID-19, Omicron wave, Yau Tsim Mong, period life tables, residential mobility, equity in health