Survival and life expectancy inequality by gender in Thai provinces: Trends from 2015 to 2023

Sergei Scherbov; Wiraporn Pothisiri; Orawan Prasitsiriphon; Dalkhat M. Ediev; Warren Sanderson

doi:10.1371/journal.pone.0348587

Abstract

Background

Gender disparities in survival and life expectancy are indicative of broader health and socio-economic inequalities. This study examines temporal trends and subnational variations in gendered survival outcomes across Thailand’s 77 provinces between 2015 and 2023, emphasizing the impacts of the COVID-19 pandemic and environmental factors. For the first time, comprehensive provincial life tables have been produced for all Thai provinces for this period.

Methods

Using civil registration data, life tables were constructed, from which life expectancy at birth (), at age 65 (), and survival probabilities between ages 20 and 65 were derived. Generalized additive models (GAM) with Poisson likelihood were employed to estimate mortality rates and analyze temporal trends and regional disparities. All analyses were conducted separately by sex. National-level estimates were computed as population-weighted aggregates of provincial estimates, from which life table indicators were derived consistently. Uncertainty intervals were estimated for key indicators.

Results

From 2015 to 2023, the national gender gap in life expectancy at birth widened from 7.0 years to a peak of 7.7 years in 2021, primarily driven by increased male mortality during the COVID-19 pandemic. Provincial variations were substantial, with gender gaps ranging from 5.7 to 9.5 years. PM2.5 exposure measured in 2024 showed a moderate to strong negative correlation with life expectancy, highlighting significant environmental impacts.

Conclusions

Persistent and geographically uneven gender disparities in life expectancy underscore the necessity for localized, gender-sensitive interventions targeting male mortality and environmental health risks.

Citation: Scherbov S, Pothisiri W, Prasitsiriphon O, Ediev DM, Sanderson W (2026) Survival and life expectancy inequality by gender in Thai provinces: Trends from 2015 to 2023. PLoS One 21(5): e0348587. https://doi.org/10.1371/journal.pone.0348587

Editor: Khin Thet Wai, Freelance Consultant, Myanmar, MYANMAR

Received: May 20, 2025; Accepted: April 17, 2026; Published: May 13, 2026

Copyright: © 2026 Scherbov et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Population data used in this study are publicly available from the Bureau of Registration Administration statistical database at: https://stat.bora.dopa.go.th/stat/statnew/statMenu/newStat/home.php. Mortality data were obtained from the Bureau of Registration Administration, Department of Provincial Administration, Ministry of Interior, Thailand. These data contain sensitive individual-level administrative records and are subject to legal and institutional restrictions imposed by the Ministry of Interior. Access to the data is governed by an internal data access review committee within the Bureau of Registration Administration. Data access inquiries may be directed to the Registration Technology Administration and Development Division, Bureau of Registration Administration, Department of Provincial Administration, Ministry of Interior, Thailand (59 Moo 11, Bueng Thong Lang, Lam Luk Ka District, Pathum Thani 12150, Thailand; Tel: +66 2 791 7443).

Funding: This research was funded in whole by the Thailand Science Research and Innovation Fund, Chulalongkorn University (Grant number: SOC_FF_68_034_5100_001). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Life expectancy is a cornerstone indicator of population health, reflecting the cumulative impacts of socio-economic development, healthcare access, environmental quality, and individual behaviors. Globally, women tend to outlive men [1]—a phenomenon attributed to biological advantages (such as hormonal protection) and differences in lifestyle and risk behaviors [2,3].

Recent estimates confirm that the female–male life expectancy gap continues to be substantial, despite its fluctuation over the century [4]. In 2021, the global average life expectancy was 73.8 years for women and 68.4 years for men—a difference of 5.4 years [2]. Evidence also indicates that the gap is strongly influenced by global and regional shocks. For instance, the COVID-19 pandemic disproportionately increased male mortality in many settings, contributing to a temporary widening of the gap [5–7]. In addition, regional crises—such as armed conflict [8], political instability (including female’s political participation [9], economic collapse [10], environmental disasters [11,12], and disease outbreaks [13,14]—have had pronounced localized effects on survival patterns. These events often exacerbate health system weaknesses, restrict access to care, and deepen existing gender-based vulnerabilities, contributing to fluctuations in the life expectancy gap across countries and regions.

While global and national trends provide valuable context, growing evidence indicates that national-level averages can obscure significant subnational disparities in health and mortality outcomes. This limitation is particularly relevant for understanding gender gaps in life expectancy. Thailand, the focus of this study, illustrates this issue clearly. National-level estimates from 2019−2022 show that the female–male life expectancy gap in Thailand ranges from 7.6 to 7.8 years [15]—substantially wider than in many high-income countries [16]. However, in 2021, provincial-level data reveal even greater variation: in some provinces, the gap reaches 8.9–9.5 years, while in others it narrows to just 5.7–6.5 years [15]. These disparities could have been further exacerbated by recent mortality shocks, particularly during the COVID-19 pandemic, which disproportionately increased male mortality [6,7,14].

This study is motivated by two key gaps in existing literature. First, with a few notable exceptions [17,18], relatively little research has examined gender disparities in life expectancy in middle-income or developing countries, where structural inequalities may amplify health risks. Among the limited body of work, only one study has explicitly analyzed the gendered effects of the COVID-19 pandemic on mortality trajectories [5]. Second, even fewer have generated subnational life tables to examine within-country variation in gendered survival—a limitation driven partly by data availability.

To address these gaps, our study constructs life tables for all 77 provinces in Thailand over a nine-year period (2015–2023), offering a detailed account of survival patterns and gender gaps across regions. We focus on three key indicators: life expectancy at birth (), life expectancy at age 65 (), and the probability of survival from age 20–65—a proxy for adult mortality risk. Specifically, we address the following questions: (1) What are the levels of and for females and males across provinces, and how do they compare with national averages? and (2) How have gender differences in these indicators evolved over time, particularly during the COVID-19 period?

Although not the primary focus of this study, we also explore potential drivers of regional variation in the gender gap by examining the correlation of provincial gap estimates with male and female life expectancy and with ambient air pollution levels. This province-level analysis aims to fill an important evidence gap and support more targeted health and development strategies in Thailand and comparable settings.

Materials and methods

Data sources

This retrospective study covers all 77 Thai provinces over the period from 2015 to 2023. Annual population and mortality data were obtained from the Bureau of Registration Administration under Thailand’s Ministry of Interior (MOI). Both datasets are derived from Thailand’s civil registration system and are available by single year of age, sex, province, and calendar year.

Population data

Annual mid-year population counts were used to represent the population at risk. Population data are available by age, sex, year, and province of registration and were used both as exposure denominators in the mortality models and as weights in the aggregation of provincial estimates to the national level.

The population data include Thai and non-Thai individuals recorded in Thailand’s household registration system. The “non-Thai” population refers to individuals who do not hold Thai nationality but are registered in the system, including long-term residents from ethnic minority and migrant groups, foreign spouses of Thai citizens registered in their spouse’s household, and children born in Thailand to non-Thai parents who do not acquire Thai nationality by birth. As non-Thais constitute approximately 1.4% of the total population, their inclusion has negligible effect on the mortality and life table estimates.

Mortality data

Mortality data, covering Thai and non-Thai deaths, includes individual-level registered deaths occurring between January 1, 2015, and December 31, 2023, with information on age, sex, province of registration, and date of death. No adjustments were made for potential delays in death reporting, as previous evidence indicates that the reporting delays accounted for only 0.03% of deaths over a 12-year period [19].

The completeness of Thailand’s official death registration system has improved significantly over time. According to the National Survey of Population Change in 2015−16, 99.3% of deaths were officially registered [20]. Given this high level of completeness, registered deaths are considered a reliable basis for estimating age- and sex-specific mortality patterns at both national and provincial levels.

PM2.5 exposure data

PM2.5 exposure data were obtained from the public download portal of the Office of Space Technology Development and Geo-Informatics (GISTDA) PM2.5 monitoring system. At the time of data retrieval, the only publicly available province-level dataset covered calendar year 2024. Twelve monthly files (January–December 2024) were combined to construct a daily province-level time series, from which annual exposure was computed as the mean of daily concentrations by province. PM2.5 concentrations are reported in µg/m³ and are derived from satellite observations integrated with ground monitoring data. Because PM2.5 exposure was available only for 2024, it is treated as a cross-sectional indicator of relative provincial exposure rather than a time-varying covariate.

Statistical model

A large body of literature models subnational variation in mortality by treating the observed number of deaths as count data and assuming a Poisson distribution, while accounting for systematic effects of age, calendar time, and geography on mortality levels and age profiles [21,22].

We fit generalized additive model (GAMs) to death counts using a Poisson likelihood and a log link. Let denote the observed number of deaths at age in year and province (estimated separately by sex) and let denote the corresponding exposure (population at risk/person-years). We assume

where is the expected number of deaths. To adjust for differences in exposure across cells, we include as an offset and model

This implies , so can be interpreted as the modeled mortality rate for the relevant age interval.

The linear predictor is specified to allow flexible age patterns that vary over time and across geography:

where is a year-specific smooth function of age (a separate smooth age profile for each year), is a region-specific smooth function of age (a separate smooth age profile for each region to which province belongs), and is a parametric fixed effect for each year–province combination. Conceptually, captures temporal evolution in the age pattern of mortality, while captures systematic regional differences in age-specific mortality profiles. The year–province interaction absorbs additional province-specific annual deviations in overall mortality level beyond what is explained by the smooth age-profile components. This perspective can be viewed as a flexible extension of Lee–Carter-type approaches to modelling age-specific mortality, adapted to subnational settings and richer covariate structures [23].

Smooth functions were represented using cubic regression splines. For each year- and region-specific age smooth, the basis dimension was set to . Smoothing parameters were estimated by Restricted Maximum Likelihood (REML) to balance fit and smoothness. Model diagnostics were used to evaluate basis adequacy and fit, including residual patterns over age and year and checks for overdispersion. Sensitivity analyses with alternative basis dimensions () yielded very similar goodness-of-fit metrics and substantively similar fitted age profiles; therefore, the final specification was guided by visual diagnostics and demographic plausibility.

Models were fitted separately by sex. In both models, we excluded infants (age 0) and restricted the analysis to ages 1–85. From the fitted GAMs, we obtained age-specific mortality rates at the provincial level by sex and year via

National age-specific mortality rates were derived as population-weighted aggregates of provincial estimates using age–sex–year-specific exposures as weights.

Uncertainty estimation

Uncertainty in age-specific mortality rates was derived from the estimated covariance matrix of the GAM. For each province–year–sex–age combination, an approximate 95% uncertainty interval for the fitted mortality rate was computed as:

where denotes the standard error of the fitted rate. Lower and upper mortality schedules constructed from these bounds were then propagated through the full life table calculations to obtain corresponding uncertainty intervals for , , and . The same oldest-age adjustment procedure was applied consistently to the lower and upper mortality schedules to ensure coherent uncertainty propagation at advanced ages. For brevity, we report uncertainty estimates only for 2023 in Supplementary Materials. Uncertainty levels in other years were comparable to those observed in 2023.

Life table computations

The age- and sex-specific mortality rates estimated from the GAM were used to construct period life tables for each province and year, following standard procedures (Preston et al., 2001) [24]. We produced complete life tables with single-year age intervals, starting at age 0 and closing with an open-ended interval at ages 101 + .

Because all age groups are single years, mortality rates were converted to probabilities of death () using

where is the average fraction of the interval lived by those who die within the interval. For all ages except infancy and the open-ended interval, we set (deaths occur on average halfway through the year). For age 0, was computed using the Keyfitz empirical approximation () [25]. For the open-ended age interval, was set to the inverse of the death rate in that interval.

All remaining life table functions, including survivorship (), person-years lived (), life expectancy at birth () and at age 65 (), were derived using standard demographic relationships. Survival between ages 20 and 65 was defined as the conditional survival probability—the proportion surviving from exact age 20 to exact age 65:

All life table computations were implemented consistently across provinces, years, and sexes to ensure comparability of the resulting indicators. National life tables and derived indicators were constructed from aggregated national mortality schedules.

Treatment of specific age groups

Special consideration was given infant mortality due to its distinct dynamics relative to mortality at older ages. Infant mortality (age 0) was not included in the smooth age functions, as its shape and determinants differ from those at ages 1 and above. Operationally, age 0 was treated as a separate age group, and the observed age-0 mortality rate – computed using registered infant deaths and the mid-year population at age 0 as the exposure denominator – was used directly rather than smoothed. This approach is deemed appropriate because Thailand’s vital registration system is generally considered to be of high quality, and under current Thai mortality levels, variation in has a negligible impact on overall life expectancy.

To assess the robustness of life expectancy estimates to alternative treatments of infant mortality, we conducted a sensitivity analysis examining different handling of age-0 mortality and infant assumptions. To illustrate the limited influence of infant mortality under current Thai mortality conditions, we note that the minimum age-0 exposure in our province-year series is 554 (men) and 555 (women). Even under a relatively high hypothetical mortality rate for Thailand at age 0 (=0.01), the standard deviation of the rate implied by binomial variation is small; three standard deviations correspond to a change on the order of 6 × 10 ⁻ ⁵ in the mortality rate, which has a negligible effect on the life table and on . For life table computations, we follow the Chiang approach, and for age 0 we estimate using the standard Keyfitz adjustment . Using alternative infant approximations (e.g., Preston-type formulations) produces negligible differences in under Thai mortality levels. See Scherbov and Ediev (2011) for additional resource of life table accuracy for small populations [26].

Mortality rates at advanced ages also require special treatment. In particular, mortality rates at ages 85 and above often show a paradoxical decline, a phenomenon associated with data limitations such as age misreporting and declining data quality at older ages [27,28]. To address this issue, we applied a logistic adjustment to the oldest-age mortality schedule, generating the probability of death () on the logit scale. Specifically, for each province–year–sex combination, we fitted a logistic relationship to over ages 65–84, an age range for which mortality patterns are generally well observed and linear in logits. The fitted model was then used to predict for ages 85–100, producing a smooth and monotonic extension of the mortality schedule at the oldest ages. The transition from observed to adjusted mortality occurs at age 85 as a hard threshold, but continuity is ensured because the adjustment is anchored to the immediately preceding observed ages. For the terminal open-ended age interval, we set .

To evaluate the plausibility and impact of the oldest-age extrapolations, we conducted a robustness check comparing life expectancy at age 65 () derived with and without the oldest-age extension (see Supplementary S1 Fig). In most cases, life expectancy without adjustment is strongly overestimated. As an additional check, we cross-checked our approach using the constrained extrapolation method [29], in combination with life expectancy estimates derived from previously proposed approaches [28,30] for the oldest old population (results not shown). While these alternative methods produced broadly consistent results, in line with our logistic regression estimates, they also showed inherent instability, with frequent outliers, due to their sensitivity to fluctuations in population growth rates. We therefore retained the logistic adjustment for its robustness and interpretability.

Fig 1 illustrates the extension of mortality rates at older ages and the overall fit of the model to observed mortality rates for ages 1–85, using selected provinces with different population sizes to highlight associated uncertainty. For illustration, we present four cases: Bangkok (males, 2020), Saraburi (males, 2020), Tak (females, 2020), and Ranong (females, 2020). In some provinces—such as Ranong for females—mortality rates in several age groups are recorded as zero because no deaths were registered. In these situations, the regional effect becomes particularly important for producing stable and accurate mortality-rate estimates.

Download:

Fig 1. Observed and fitted mortality rates for selected Thai provinces in 2020.

https://doi.org/10.1371/journal.pone.0348587.g001

Ethical considerations

As this study relies on publicly accessible de-identified secondary data collected through official channels, ethical approval was not required. All data handling complied with relevant national data protection guidelines.

Results

Our analysis of Thai provincial data from 2015 to 2023 reveals significant temporal trends and regional disparities in survival outcomes. In the following subsections, we detail the trends in life expectancy at birth () and at age 65 (), the evolution of gender gaps over time, discuss the probability to survive from age 20 to age 65, the impact of the COVID‐19 pandemic, and possible drivers behind the gender gap variation. The complete set of results incorporated into an interactive map of Thai provinces could be found in online repository at https://sergio007.shinyapps.io/intmap/.

National trends in life expectancy at birth (e₀)

In Table 1, between 2015 and 2023, Thailand’s national estimates reveal a steady improvement in life expectancy at birth until 2020. In 2015, life expectancy at birth was 79.6 years for females and 72.6 years for males. By 2020, these figures peaked at around 80.7 years for females and 73.4 years for males, coinciding with the country’s early and effective response to the COVID-19 pandemic such as physical distancing, mask wearing, hand hygiene, widespread testing, and public health communication [31].

Download:

Table 1. National and province-level distribution of life expectancy at birth (

) by sex, Thailand, 2015–2023.

https://doi.org/10.1371/journal.pone.0348587.t001

However, this progress was disrupted during 2021–2022, when Thailand experienced significant waves of COVID-19. The confirmed COVID-related death toll reached 21,693 in 2021 and 11,971 in 2022 [32], contributing to the decline in life expectancy during those years. Female life expectancy dropped to 79.2–79.5 years, while male life expectancy dropped to 71.5–71.8 years-reversing the gains achieved in the preceding years. For 2023, province- and sex-specific estimates are reported with 95% uncertainty intervals derived from mortality uncertainty (Supplementary S1 Table).

Provincial trends in life expectancy at birth

Fig 2 presents province-level distributions of life expectancy at birth () by sex from 2015 to 2023 using box-and-whisker plots. Across provinces, the median follows a temporal pattern consistent with national trend. The boxplots also highlight persistent spatial inequality in , indicated by the width of the interquartile ranges (IQRs) and the length of the whiskers. In most years, the male distributions are more dispersed than the female distributions, suggesting greater interprovincial heterogeneity in mortality conditions among men.

Download:

Fig 2. Provincial variation in life expectancy at birth (

) in Thailand, 2015–2023, by sex. Boxplots summarize the distribution of provincial estimates each year, the length of box shows the interquartile range (IQR), the horizontal line in the IQR represents the median

across provinces, the whiskers indicate the minimum value at 1.5*IQR of the 25^th percentile, and the maximum value at 1.5*IQR of the 75^th percentile, and the dots denote outlying provinces with unusually low or high

.

https://doi.org/10.1371/journal.pone.0348587.g002

Based on Table 1 and its details provided in online repository, the difference between the highest and lowest provincial ranged from 5.6–7.8 years among women and 6.5–8.1 years among men. Spatial clustering is evident. Across the study period, Bangkok and surrounding provinces were in general at or above the national average . A high share of provinces in Southern Thailand also remained at or above the national average (typically 11–12 of 14 provinces).

In 2020—the year with the highest national for both gender—the highest for women was observed in Phatthalung (85.2 years) and the lowest provincial occurred in Nakhon Phanom (78.0 years) for women. For men, the highest provincial was observed in Bangkok (77.8 years), and the lowest provincial occurred in Kalasin (70.2 years). Bangkok with surrounding provinces and 11 out of 14 provinces in the South exhibited female life expectancy levels at or above the national average. Likewise, Bangkok with surrounding provinces and 12 out of 14 provinces in the South (excluding Narathiwat and Yala) were at or above the male life expectancy at the national average.

In contrast, a clear clustering of lower longevity is observed in the Northeast and North. In 2020, 19 out of 20 provinces in the Northeast (for both sexes), and 14 of 17 provinces in North for men, and 15 of 17 provinces in the North for women had life expectancy levels below national average (See maps in online repository).

We also observed that several provinces with above-average life expectancy such as Bangkok and surrounding provinces, Chon Buri, and Ranong, exhibited signs of recovery in 2022–2023, suggesting a possible rebound in survival outcomes following the pandemic’s peak (See maps in online repository).

These spatial patterns provide a foundation for further examination of gender gaps and regional disparities in life expectancy, particularly in the context of pandemic-related shocks.

Gender gap in life expectancy at birth

Analysis of national data from 2015 to 2023 reveals a persistent gender gap in life expectancy at birth, with females consistently outliving males (Table 1). In 2015, the national-level gender gap in life expectancy at birth was 7.0 years. This gap widened slightly to 7.3 years by 2020. The gap increased more markedly in 2021—when COVID-19 had its most severe impact on Thailand—reaching 7.7 years. By 2023, the gender gap had narrowed slightly to 7.5 years, though it remained higher than the pre-pandemic level. This trend mirrors patterns observed in the United States, where COVID-19 was identified as a major contributor to the widening gender gap in life expectancy [7].

All province-year life expectancy estimates and derived gender-gap measures are available in the online repository. Using these province-year estimates, we find substantial heterogeneity at the provincial level, although the overall trajectory aligns with the national pattern. In 2015 the widest gender gap was recorded in Singburi (8.9 years), while the narrowest was in Loei (5.3 years). By 2020, Phatthalung emerged as the province with the widest gender gap (9.3 years), while Satun exhibited the narrowest (4.3 years). In 2022, Phatthalung again reported the highest gender gap at 9.5 years, suggesting that the pandemic may have disproportionately impacted male survival in the province. In contrast, Narathiwat had the lowest gender gap at 5.9 years.

Fig 3 visualizes these disparities in 2023, with the left panel showing male life expectancy at birth and the right panel illustrating the gender gap. Provinces in the North and Northeast generally exhibit lower male life expectancy and larger gender gaps, while Southern provinces such as Yala, Pattani, and Narathiwat report both higher male life expectancy and narrower gender gaps. These spatial patterns reinforce the need to explore how socio-economic, cultural, behavioral, and environmental factors intersect to shape gendered survival outcomes—particularly during and after pandemic-related health system disruptions.

Download:

Fig 3. Provincial patterns of male life expectancy at birth and gender gap in life expectancy at birth at birth, Thailand, 2023.

Maps were produced using administrative boundary data provided by geoBoundaries and made available under the Creative Commons Attribution 4.0 (CC BY 4.0) license [33].

https://doi.org/10.1371/journal.pone.0348587.g003

An analysis of the correlation patterns provides further insight into the dynamics underlying these gender differences (Fig 4). From 2015 to 2023, the correlation between male life expectancy at birth and the gender gap was consistently strong and negative, reaching −0.7 in 2023. This indicates that lower male life expectancy is strongly associated with a wider gender gap, suggesting that male-specific mortality factors are a primary driver of the disparity. In contrast, the correlation between female life expectancy at birth and the gender gap was statistically insignificant and fluctuated around zero, implying that variations in female longevity play a minimal role in shaping the gap.

Download:

Fig 4. Year-specific association between provincial life expectancy at birth and the gender gap in life expectancy in Thailand, 2015–2023.

For each year, we calculated Pearson’s correlation coefficient () across provinces between male (blue) or female (orange) provincial life expectancy at birth () and the provincial gender gap in life expectancy at birth (female – male, ). Shaded bands show 95% confidence intervals for based on Fisher’s z transformation.

https://doi.org/10.1371/journal.pone.0348587.g004

Trends in life expectancy at age 65 and gender gap

As shown in Table 2 with full province-year results in the online repository, the geographic distribution of life expectancy at age 65 is broadly consistent with the spatial clustering observed for life expectancy at birth (Table 1). National life expectancy at age 65 () improved gradually before the COVID-19 pandemic, with noticeable differences between females and males. In 2015, the national was 19.6 years for females and 17.2 years for males—a gender gap of around 2.4 years. The life expectancy at age 65 increased steadily and peaked in 2020, when women enjoyed a life expectancy of 20.3 years and men 17.8 years, resulting in an increased gap of 2.6 years at age 65. However, the relaxation of state responses to COVID-19 during 2021–2022 led to a shift in the trend, reversing earlier improvements. In 2022, life expectancy at age 65 declined to 19.1 years for females and 16.3 years for males, resulting in a gender gap of 2.9 years. This widening occurred not because women were unaffected, but because male life expectancy declined more sharply, reflecting differential vulnerability to COVID-19 related mortality [5,6].

Download:

Table 2. National and province-level distribution of life expectancy at birth (

) by sex, Thailand, 2015–2023.

https://doi.org/10.1371/journal.pone.0348587.t002

Importantly, these trends were not uniform across regions. In 2020, at the provincial level, Phatthalung recorded the highest female life expectancy at age 65 (23.9 years), while Nakhon Phanom reported the lowest (18.2 years). For males, Bangkok (20.2 years) and southern provinces such as Satun (19.3 years) and Phuket (19.1 years) exhibited relatively high life expectancy at age 65. In contrast, Kalasin in the Northeast and Lamphun in the North reported the lowest male values (both 16.0 years).

Provincial heterogeneity is also evident in the gender gap at age 65. In 2020, the gap was most pronounced in Phatthalung at 4.9 years, while Tak and Mukdahan had the narrowest gaps (0.8 years and 1.0 year, respectively). These regional disparities in both absolute survival and gender gaps suggest underlying structural differences in healthcare access [34], prevalence of chronic conditions [35], and behavioral health risks [36], all of which likely shaped the pandemic’s impact on elderly survival.

Probability to survive from age 20 to age 65

Although life expectancy at birth and the probability of surviving from age 20 to age 65 are closely related, the latter provides insights into adult mortality and workforce health (data is available in the online repository or Supplementary S3 Table for 2023).

At the national level, the probability of surviving from age 20 to age 65 in Thailand exhibited a generally stable pattern over time, with gradual improvements before the COVID-19 pandemic and a notable decline in 2021. In 2020, when survival probabilities peaked, the national average stood at 88.3% for females and 74.6% for males. However, the impact of the pandemic in 2021–2022 led to a decline, with survival probabilities dropping to 86.9% for females and 72.0% for males, before partially recovering in 2023. While these shifts are modest, they align with broader evidence that male survival was more adversely affected by the pandemic, reinforcing observed gender disparities in life expectancy [5,6].

At the provincial level, our data reveal notable variability in the probability of surviving from age 20 to age 65 years across Thailand. While the overall time trend aligns with national averages, certain provinces consistently outperform or underperform compared to the national survival rate.

Among the provinces with the high survival probabilities, Bangkok consistently recorded values well above the national average. In 2020—when survival rates were at their peak nationally—Bangkok reported probabilities of approximately 91.2% for females and 80.9% for males, higher than the national figures. In 2021, Bangkok experienced a decline, with survival dropping to 88.4% for females and 74.9% for males, though it remained above the national trend.

Conversely, Sa Kaew—a relatively poor eastern province bordering Cambodia—and Phrae—a small mountainous province in northern Thailand—consistently exhibited some of the lowest survival probabilities from age 20–65. In 2020, Sa Kaew reported 85.6% survival for females, while Phrae recorded 69.5% for males. By 2022, these figures declined to 83.5% for females in Sa Kaew and 68.1% for males in Phrae. These provinces remained in the lower quintile of survival across all years in the study period, reflecting persistently lower midlife survival outcomes.

Association between PM2.5 and life expectancy

Analysis of provincial PM2.5 levels and life expectancy across Thailand revealed notable regional disparities. Urban centers such as Bangkok recorded average annual PM2.5 concentrations exceeding 20 μg/m³, while many southern provinces maintained lower levels, often below 15 μg/m³. Northern provinces exhibited elevated PM2.5 levels, largely attributable to seasonal agricultural burning and industrial activity.

To examine this relationship more systematically, we computed Pearson correlation coefficients between recorded PM2.5 values (from 2024—the only available provincial dataset) and average female life expectancy at birth from 2015 to 2023. The results, presented in Fig 5, reveal a moderate to strong negative correlation, with coefficients ranging from –0.5 to –0.6. This indicates that provinces with higher levels of PM2.5 tended to have lower female life expectancy. During the height of the COVID-19 pandemic (2021–2022), the correlation between the decline in life expectancy and PM2.5 levels was the weakest over the study period.

Download:

Fig 5. Year-specific correlation between regional PM2.5 exposure and life expectancy at birth (

) in Thailand, 2015–2023. Lines show Pearson’s correlation coefficient (

) across regions between regional PM2.5 and regional life expectancy at birth (

) for males (blue) and females (orange), computed using complete cases only. The horizontal dashed line indicates

.

https://doi.org/10.1371/journal.pone.0348587.g005

Discussion

This study provides new empirical evidence on gender disparities in life expectancy across Thailand’s 77 provinces from 2015 to 2023, with particular attention to provincial-level deviations from national patterns. By constructing subnational life tables using national vital registration data, we systematically examine temporal shifts in gender gaps during the COVID-19 pandemic and explore the role of socio-economic, healthcare, and environmental factors, especially PM2.5 exposure, in shaping survival inequalities. These findings address the underexplored issue of within-country gender gaps in middle-income settings.

Between 2015–2019, Thailand experienced modest but steady improvements in life expectancy. Life expectancy at birth increased from 79.6 to 80.2 years for females and from 72.6 to 72.8 years for males. Similarly, life expectancy at age 65 years rose from 19.6 to 20.3 for females, and from 17.2 years to 17.8 years for males. These gains led to a slight widening of the gender gap from 7.0 to 7.4 years in life expectancy at birth, and from 2.4 to 2.5 years at age 65. The findings are consistent with broader trends observed across Southeast Asia [37], where men consistently live shorter than women [38,39].

However, this trajectory was disrupted by the COVID-19 pandemic. Our findings show that the widening gender gap was largely driven by excess male mortality, with the COVID-19 pandemic acting as an amplifying factor between 2020–2023. In 2020, Thailand initially gained survival advantages through effective public health responses such as physical distancing, mask wearing, hand hygiene, and public health communication [31] which contributed to a modest increase of 0.5–0.6 years in life expectancy at birth and temporarily stabilized the gender gap. However, by 2022, the relaxation of these measures coincided with a sharper decline in male life expectancy (−1.8 years) compared to females (−1.5 years), resulting in the gender gap widening to 7.7 years [2,40]. Similar patterns were observed in life expectancy at age 65, with males experiencing a larger drop. By 2023, male life expectancy rebounded slightly faster than that of females. This pattern is confirmed by our correlation analysis (see Fig 3). These results are consistent with global studies suggesting that men were more vulnerable to COVID-19-related mortality [5,6,41,42].

Our findings of regional life expectancy align with previous studies that documented geographical variation in mortality across Thailand using national vital registration data [35,43]. Provinces in the North and Northeast consistently reported lower life expectancy, a pattern likely driven by structural socio-economic disadvantages. These regions have a higher proportion of older adults and elevated poverty rates [44], both of which are associated with increased risks and reduced access to health-promoting resources [36,43,45].

Unlike provinces in the North and Northeast, where economic disadvantage is associated with lower life expectancy, the Southern region—despite having the highest poverty rate since 2017 [44]—has consistently reported higher life expectancy. This apparent paradox may be attributed to distinct socio-cultural and behavioral factors. Prior research indicates that the southern region—particularly provinces with a higher proportion of Muslim residents—tends to exhibit healthier lifestyles. These include lower levels of alcohol consumption [46], greater intake of fruit and vegetable intake [36], and lower rates of road traffic accidents, particularly in provinces like Satun, Pattani, and Yala [36]. Collectively, these behaviors may contribute to more favorable health outcomes.

Moreover, provinces with predominantly Muslim populations may also experience narrower gender gaps in life expectancy. Among Thai men, approximately 30.4% of total disease burden is attributable to smoking and alcohol consumption, whereas these factors contribute minimally to disease burden among women [47]. In predominantly Muslim provinces—where religious practices discourage such behaviors—male mortality risks may be comparatively lower, contributing to smaller gender gaps in life expectancy observed in provinces such as Narathiwat, Pattani, and Yala. This supports the hypothesis that gender differences in behavioral risk exposures play a significant role in shaping survival disparities.

This spatial heterogeneity in gender gaps may reflect the interaction between biological and behavioral factors. Estrogen provides enhanced immune system, while the presence of two X chromosomes may offer greater genetic resilience [39,42]. In terms of health behavior, Thai women were significantly less likely to smoke across all age groups and regions [36,48]. Between 1990 and 2019, the smoking prevalence among Thai women declined by 46.5%, compared to a 25.5% reduction among men [49]. Furthermore, the proportion of adults engaging in physical activity—especially among those aged 60 and above—increased from 58.6% in 2012 to 73.4% in 2019 [36]. These biological and behavioral advantages underpin the consistent female advantage in life expectancy across all ages. This partly explains the observed resilience of female life expectancy during the pandemic, even as male life expectancy declined more sharply. Nevertheless, the gender gap becomes slightly narrower at older ages, possibly due to selective survival and reduced male–female differences in risk exposure later in life [50].

Moreover, socio-economic status remains a critical determinant of health inequalities. Our findings that provinces with lower socio-economic development, such as many in the Northeast and North of Thailand, tended to exhibit lower life expectancy and wider gender gaps, align with the evidence that low-income and rural areas typically face compounded disadvantages: limited access to quality healthcare, fewer educational opportunities, environmental hazards, and unstable employment [51–53]. Moreover, COVID-19 exacerbated existing socio-economic vulnerabilities. Similar to patterns observed internationally [54,55], regions with higher poverty, weaker health infrastructure, and greater proportions of disadvantaged groups faced disproportionately higher mortality impacts, particularly among men who tend to engage in riskier behaviors, have greater biological vulnerabilities, and are overrepresented in high-risk occupations [5–7,17,41,56].

Variations in healthcare access and quality further contributed to differential survival outcomes across provinces [34,57]. During the COVID-19 pandemic, widespread strain on healthcare systems led to disruptions in essential services for non-communicable diseases (NCDs) such as diagnosis, treatment, rehabilitation, and palliative care [58,59]. Infection control measures, including lockdowns and work-from-home policies [31], may have further limited healthcare access [58], disproportionately affecting male survival risks. Our findings suggest that provinces with stronger healthcare infrastructure, such as Bangkok and surrounding metropolitan areas, exhibited faster recovery in life expectancy trends compared to rural and underserved areas.

Environmental factors, particularly exposure to fine particulate matter (PM2.5), may further influence survival probabilities. Provinces in the North, such as Chiang Mai and Lampang, reported higher levels of PM2.5 concentrations, correlating with elevated mortality from chronic obstructive pulmonary disease (COPD) [60], and lower life expectancy observed in our findings. Long-term exposure to PM2.5 is strongly associated with increased risks of cardiovascular and respiratory mortality [61–64]. Moreover, emerging evidence suggests that although ischemic heart disease and stroke are more prevalent among men [36,65], women may exhibit greater biological vulnerability to the cardiovascular effects of particulate pollution [64], potentially contributing to regional gender differences in survival outcomes.

Strengths and limitations

A key strength of our study is the use of detailed provincial data spanning a nine-year period, which allows for an in-depth analysis of temporal trends and regional heterogeneity. However, several limitations must be acknowledged. First, our analysis relies on administrative mortality records, which may be subject to reporting biases that vary by province. Second, although our GAM framework effectively captures nonlinear and spatial effects, residual confounding from unmeasured variables—such as detailed socio-economic indicators—may persist. Third, while we reference existing literature on PM2.5, the absence of direct integration of detailed provincial pollution data limits our ability to quantify its specific contribution to regional survival differences. Furthermore, this study did not aim to determine whether environmental and behavioral effects might be confounded. This is an important direction for future research. Finally, recent studies have raised methodological concerns regarding the use of traditional period life tables for estimating changes in life expectancy during periods of crisis such as pandemics [66]. In such contexts, hybrid life expectancy—combining period and cohort perspectives—may provide a more accurate measure of population survival under rapidly changing mortality conditions.

Conclusion

This study shows that gender disparities in life expectancy across Thailand’s provinces remain persistent and geographically uneven. The COVID-19 pandemic magnified these differences, particularly where structural and environmental vulnerabilities were greatest. Addressing gendered survival gaps requires attention to local health system capacity, behavioral risk factors, and environmental exposures. Tailored health policies—especially those targeting male health risks and pollution-related burdens—can help reduce inequalities and improve population health equity.

Supporting information

S1 Fig. Sensitivity of GAM-derived to oldest-old (85+) closure.

Boxplots summarize, by year (2015–2023), the difference in regional obtained from the GAM-fitted mortality schedule after applying an 85 + logit-closure adjustment compared with the unadjusted GAM-based estimate (Δ = adjusted − unadjusted). Male and female results are shown in separate panels; the zero line denotes no effect. Outliers with |Δ| > 20 years were excluded. The overwhelmingly negative differences highlight the biases in the original data caused by age exaggeration, and emphasize the importance of adjustment to obtain more reliable life expectancy estimates.

https://doi.org/10.1371/journal.pone.0348587.s001

(PNG)

S1 Table. Life expectancy at birth () by province and sex, Thailand, 2023.

Values shown as mean (lower, upper); bounds are 95% interval limits derived from mortality uncertainty.

https://doi.org/10.1371/journal.pone.0348587.s002

(DOCX)

S2 Table. Life expectancy at age 65 () by province and sex, Thailand, 2023.

Values shown as mean (lower, upper); bounds are 95% interval limits derived from mortality uncertainty.

https://doi.org/10.1371/journal.pone.0348587.s003

(DOCX)

S3 Table. Probability of surviving from age 20–65 () by province and sex, Thailand, 2023.

Values shown as mean (lower, upper); bounds are 95% interval limits derived from mortality uncertainty.

https://doi.org/10.1371/journal.pone.0348587.s004

(DOCX)

Acknowledgments

The authors wish to thank the Bureau of Registration Administration, the Ministry of Interior (Thailand) for providing data used in this manuscript.

References

1. Barford A, Dorling D, Davey Smith G, Shaw M. Life expectancy: women now on top everywhere. BMJ. 2006;332(7545):808. pmid:16601021
- View Article
- PubMed/NCBI
- Google Scholar
2. Dattani S, Rodés-Guirao L. Why do women live longer than men?. Our World in Data. 2023. Accessed 2025 April 27. https://archive.ourworldindata.org/20260417-153549/why-do-women-live-longer-than-men.html
3. Ginter E, Simko V. Women live longer than men. Bratisl Lek Listy. 2013;114(2):45–9. pmid:23331196
- View Article
- PubMed/NCBI
- Google Scholar
4. Feraldi A, Zarulli V, Mazzuco S, Giudici C. Functional data analysis approach in population studies: an application to the gender gap in life expectancy. Qual Quant. 2024;58:5125–50.
- View Article
- Google Scholar
5. Beegle K, Demombynes G, de Walque D, Gubbins P, Veillard J. COVID-19 increased existing gender mortality gaps in high-income more than middle-income countries. Int J Infect Dis. 2024;148:107167. pmid:39233048
- View Article
- PubMed/NCBI
- Google Scholar
6. Castro MC, Gurzenda S, Turra CM, Kim S, Andrasfay T, Goldman N. Reduction in life expectancy in Brazil after COVID-19. Nat Med. 2021;27(9):1629–35. pmid:34188224
- View Article
- PubMed/NCBI
- Google Scholar
7. Yan BW, Arias E, Geller AC, Miller DR, Kochanek KD, Koh HK. Widening gender gap in life expectancy in the US, 2010-2021. JAMA Intern Med. 2024;184(1):108–10.
- View Article
- Google Scholar
8. Plümper T, Neumayer E. The unequal burden of war: the effect of armed conflict on the gender gap in life expectancy. Int Orga. 2006;60(3):723–54.
- View Article
- Google Scholar
9. Macmillan R, Shofia N, Sigle W. Gender and the politics of death: female representation, political and developmental context, and population health in a cross-national panel. Demography. 2018;55(5):1905–34. pmid:30128940
- View Article
- PubMed/NCBI
- Google Scholar
10. Gutmann J, Matthias N, Neumeier F. Sanctioned to death? The impact of economic sanctions on life expectancy and its gender gap. J Dev Stud. 2021;57(1):139–62.
- View Article
- Google Scholar
11. Neumayer E, Plümper T. The gendered nature of natural disasters: the impact of catastrophic events on the gender gap in life expectancy, 1981–2002. Ann Assoc Am Geogr. 2007;97(3):551–66.
- View Article
- Google Scholar
12. Murillo MQ, Tan S. Discovering the differential and gendered consequences of natural disasters on the gender gap in life expectancy in Southeast Asia. Nat Hazards Earth Syst Sci Discuss. 2017;2017:1–18.
- View Article
- Google Scholar
13. Masquelier B, Reniers G. AIDS and the gender gap in life expectancy in Africa. Popul Societ. 2018;(554):1–4.
- View Article
- Google Scholar
14. Noymer A, Garenne M. The 1918 influenza epidemic’s effects on sex differentials in mortality in the United States. Popul Dev Rev. 2000;26(3):565–81. pmid:19530360
- View Article
- PubMed/NCBI
- Google Scholar
15. International Health Policy Program. Life expectancy and health adjusted life expectancy of Thai population in 2019 - 2021 [in Thai]. Nonthaburi: International Health Policy Program; 2023. https://le-hale.bodthai.net/
16. United Nations Department of Economic and Social Affairs Population Division. World population prospects 2024. New York: United Nations; 2024. https://population.un.org/wpp/
17. De Walque D, Filmer D. Trends and socioeconomic gradients in adult mortality around the developing world. Popul Dev Rev. 2013;39(1):1–29.
- View Article
- Google Scholar
18. Pinho-Gomes A-C, Peters SAE, Woodward M. Gender equality related to gender differences in life expectancy across the globe gender equality and life expectancy. PLOS Glob Public Health. 2023;3(3):e0001214. pmid:36963039
- View Article
- PubMed/NCBI
- Google Scholar
19. Pothisiri W, Prasitsiriphon O, Apakupakul J, Ploddi K. Gender differences in estimated excess mortality during the COVID-19 pandemic in Thailand. BMC Public Health. 2023;23(1):1900. pmid:37784059
- View Article
- PubMed/NCBI
- Google Scholar
20. National Statistical Office of Thailand. The 2015-2016 survey of population change [in Thai]. Bangkok: National Statistical Office; 2017.
21. Alexander M, Zagheni E, Barbieri M. A flexible bayesian model for estimating subnational mortality. Demography. 2017;54(6):2025–41. pmid:29019084
- View Article
- PubMed/NCBI
- Google Scholar
22. Czado C, Delwarde A, Denuit M. Bayesian Poisson log-bilinear mortality projections. Insur Math Econ. 2005;36(3):260–84.
- View Article
- Google Scholar
23. Lee RD, Carter LR. Modeling and forecasting US mortality. J Am Stat Assoc. 1992;87(419):659–71.
- View Article
- Google Scholar
24. Preston S, Heuveline P, Guillot M. Demography: measuring and modeling population processes. Malden, MA: Blackwell Publishers; 2001.
25. Keyfitz N. Finding probabilities from observed rates or how to make a life table. Am Stat. 1970;24(1):28–33.
- View Article
- Google Scholar
26. Scherbov S, Ediev D. Significance of life table estimates for small populations: simulation-based study of estimation errors. Demogr Res. 2011;24(22):527–50.
- View Article
- Google Scholar
27. Ediev DM. Age exaggeration ruses: Infrequent age overstatement distorts the mortality curve at old age. In: Skiadas CH, Skiadas C, editors. Quantitative methods in demography: methods and related applications in the Covid-19 era. Springer; 2022. 189–205.
28. Horiuchi S, Coale AJ. A simple equation for estimating the expectation of life at old ages. Popul Stud (Camb). 1982;36(2):317–26. pmid:22077276
- View Article
- PubMed/NCBI
- Google Scholar
29. Ediev DM. Constrained mortality extrapolation to old age: an empirical assessment. Eur J Popul. 2018;34(3):441–57. pmid:30147209
- View Article
- PubMed/NCBI
- Google Scholar
30. Mitra S. Estimating the expectation of life at older ages. Popul Stud. 1984;38(2):313–9.
- View Article
- Google Scholar
31. Ministry of Public Health. Strategy: managing the new wave of the Covid-19 epidemic. Nonthaburi: Ministry of Public Health; 2021.
32. SAT-MOPH. Situation of COVID-19 patients in Thailand. 2023. https://ddc.moph.go.th/covid19-dashboard/
33. Runfola D, Anderson A, Baier H, Crittenden M, Dowker E, Fuhrig S, et al. geoBoundaries: a global database of political administrative boundaries. PLoS One. 2020;15(4):e0231866. pmid:32330167
- View Article
- PubMed/NCBI
- Google Scholar
34. Tonboot S, Wattanadumrong B. Situation of inequality in health utilization among Thai elderly in 2020. In: Business creativity and innovation in the age of transformation: proceedings of the 8th business economics and communications international conference (BECIC 2021). 2021.
- View Article
- Google Scholar
35. Faramnuayphol P, Chongsuvivatwong V, Pannarunothai S. Geographical variation of mortality in Thailand. J Med Assoc Thai. 2008;91(9):1455–60. pmid:18843878
- View Article
- PubMed/NCBI
- Google Scholar
36. Thai Health Project. Thai health 2023: Thailand’s commitment in COP and responses to climate change. Nakhon Pathom: Institute for Population and Social Research, Mahidol University; 2023.
37. United Nations Economic and Social Commission for Asia and the Pacific. Asia-Pacific report on population ageing 2022: trends, policies and good practices regarding older persons and population ageing. Bangkok: United Nations; 2022.
38. Ho C. Asian men’s health: gender disparity and life expectancy. J Surg Acad. 2014;4:5–8.
- View Article
- Google Scholar
39. Regitz‐Zagrosek V. Sex and gender differences in health. EMBO Reports. 2012;13(7):596–603.
- View Article
- Google Scholar
40. Thornton J. WHO report shows that women outlive men worldwide. BMJ. 2019;365:l1631. pmid:30952650
- View Article
- PubMed/NCBI
- Google Scholar
41. Peckham H, de Gruijter NM, Raine C, Radziszewska A, Ciurtin C, Wedderburn LR, et al. Male sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU admission. Nat Commun. 2020;11(1):6317. pmid:33298944
- View Article
- PubMed/NCBI
- Google Scholar
42. Rehman S, Ravinayagam V, Nahvi I, Aldossary H, Al-Shammari M, Amiri MSA, et al. Immunity, sex hormones, and environmental factors as determinants of COVID-19 disparity in women. Front Immunol. 2021;12:680845. pmid:34484179
- View Article
- PubMed/NCBI
- Google Scholar
43. Aungkulanon S, Tangcharoensathien V, Shibuya K, Bundhamcharoen K, Chongsuvivatwong V. Post universal health coverage trend and geographical inequalities of mortality in Thailand. Int J Equity Health. 2016;15(1):190. pmid:27876056
- View Article
- PubMed/NCBI
- Google Scholar
44. National Economic and Social Development Council. Situation analysis report on poverty and inequality in Thailand. Bangkok: National Economic and Social Development Council; 2024.
45. Yiengprugsawan V, Lim LL-Y, Carmichael GA, Seubsman S-A, Sleigh AC. Tracking and decomposing health and disease inequality in Thailand. Ann Epidemiol. 2009;19(11):800–7. pmid:19560371
- View Article
- PubMed/NCBI
- Google Scholar
46. Wakabayashi M, McKetin R, Banwell C, Yiengprugsawan V, Kelly M, Seubsman S, et al. Alcohol consumption patterns in Thailand and their relationship with non-communicable disease. BMC Public Health. 2015;15:1297. pmid:26704520
- View Article
- PubMed/NCBI
- Google Scholar
47. International Health Policy Program. Burden of disease attributable to risk factors in Thailand 2019 [in Thai]. Nonthaburi: International Health Policy Program; 2023.
48. Jitnarin N, Kosulwat V, Rojroongwasinkul N, Boonpraderm A, Haddock CK, Poston WSC. Socioeconomic status and smoking among thai adults: results of the National Thai Food Consumption Survey. Asia Pac J Public Health. 2011;23(5):672–81. pmid:20460275
- View Article
- PubMed/NCBI
- Google Scholar
49. GBD 2019 Tobacco Collaborators. Spatial, temporal, and demographic patterns in prevalence of smoking tobacco use and attributable disease burden in 204 countries and territories, 1990–2019: a systematic analysis from the Global Burden of Disease Study 2019. Lancet. 2021;397(10292):2337–60.
50. Oksuzyan A, Juel K, Vaupel JW, Christensen K. Men: good health and high mortality. Sex differences in health and aging. Aging Clin Exp Res. 2008;20(2):91–102. pmid:18431075
- View Article
- PubMed/NCBI
- Google Scholar
51. National Academies of Sciences, Engineering, and Medicine. The future of nursing 2020-2030: charting a path to achieve health equity. Flaubert JL, Le Menestrel S, Williams DR, Wakefield MK, editors. Washington (DC): National Academies Press (US); 2021.
52. Khavjou O, Tayebali Z, Cho P, Myers K, Zhang P. Rural-urban disparities in state-level diabetes prevalence among US adults, 2021. Prev Chronic Dis. 2025;22:E05. pmid:39819894
- View Article
- PubMed/NCBI
- Google Scholar
53. Passarelli-Araujo H, de Souza GM. Urban-rural health disparities in Brazil: do sociodemographic attributes play a role?. Prev Med. 2023;175:107679. pmid:37625652
- View Article
- PubMed/NCBI
- Google Scholar
54. Johri M, Agarwal S, Khullar A, Chandra D, Pratap VS, Seth A. The first 100 days: how has COVID-19 affected poor and vulnerable groups in India?. Health Promot Inter. 2021;36(6):1716–26. pmid:34002217
- View Article
- PubMed/NCBI
- Google Scholar
55. Rönkkö R, Rutherford S, Sen K. The impact of the COVID-19 pandemic on the poor: insights from the Hrishipara diaries. World Dev. 2022;149:105689. pmid:34697515
- View Article
- PubMed/NCBI
- Google Scholar
56. Scully EP, Haverfield J, Ursin RL, Tannenbaum C, Klein SL. Considering how biological sex impacts immune responses and COVID-19 outcomes. Nat Rev Immunol. 2020;20(7):442–7. pmid:32528136
- View Article
- PubMed/NCBI
- Google Scholar
57. Kosulwat V. The nutrition and health transition in Thailand. Public Health Nutr. 2002;5(1A):183–9. pmid:12027283
- View Article
- PubMed/NCBI
- Google Scholar
58. Gadsden T, Downey LE, Vilas VDR, Peiris D, Jan S. The impact of COVID-19 on essential health service provision for noncommunicable diseases in the South-East Asia region: a systematic review. Lancet Reg Health Southeast Asia. 2022;1:100010. pmid:35769108
- View Article
- PubMed/NCBI
- Google Scholar
59. Kharroubi SA, Diab-El-Harake M. Sex-differences in COVID-19 diagnosis, risk factors and disease comorbidities: a large US-based cohort study. Front Public Health. 2022;10:1029190. pmid:36466473
- View Article
- PubMed/NCBI
- Google Scholar
60. Odton P, Choonpradub C, Bundhamcharoen K. Geographical variations in all-cause mortality in Thailand. Southeast Asian J Trop Med Public Health. 2010;41:1209–19.
- View Article
- Google Scholar
61. Greenpeace Southeast Asia. Greenpeace Southeast Asia annual report 2022. Bangkok: Greenpeace Southeast Asia; 2022.
62. Guo J, Chai G, Song X, Hui X, Li Z, Feng X, et al. Long-term exposure to particulate matter on cardiovascular and respiratory diseases in low- and middle-income countries: a systematic review and meta-analysis. Front Public Health. 2023;11:1134341. pmid:37056647
- View Article
- PubMed/NCBI
- Google Scholar
63. Lelieveld J, Klingmüller K, Pozzer A, Pöschl U, Fnais M, Daiber A, et al. Cardiovascular disease burden from ambient air pollution in Europe reassessed using novel hazard ratio functions. Eur Heart J. 2019;40(20):1590–6. pmid:30860255
- View Article
- PubMed/NCBI
- Google Scholar
64. Zhang J, Wang X, Yan M, Shan A, Wang C, Yang X, et al. Sex differences in cardiovascular risk associated with long-term PM2.5 exposure: a systematic review and meta-analysis of cohort studies. Front Public Health. 2022;10:802167. pmid:35186842
- View Article
- PubMed/NCBI
- Google Scholar
65. Institute for Health Metrics and Evaluation IHME. Global burden of disease 2021: findings from the GBD 2021 study. Seattle, WA: IHME; 2024.
66. Sanderson WC, Scherbov S. The effect of the COVID-19 pandemic on life expectancy in the USA: an application of hybrid life expectancy. Biol Methods Protoc. 2023;8(1):bpad025. pmid:37929068
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Barford A, Dorling D, Davey Smith G, Shaw M. Life expectancy: women now on top everywhere. BMJ. 2006;332(7545):808. pmid:16601021
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Dattani S, Rodés-Guirao L. Why do women live longer than men?. Our World in Data. 2023. Accessed 2025 April 27. https://archive.ourworldindata.org/20260417-153549/why-do-women-live-longer-than-men.html

[ref3] 3. Ginter E, Simko V. Women live longer than men. Bratisl Lek Listy. 2013;114(2):45–9. pmid:23331196
View Article
PubMed/NCBI
Google Scholar

[7] View Article

[8] PubMed/NCBI

[9] Google Scholar

[ref4] 4. Feraldi A, Zarulli V, Mazzuco S, Giudici C. Functional data analysis approach in population studies: an application to the gender gap in life expectancy. Qual Quant. 2024;58:5125–50.
View Article
Google Scholar

[11] View Article

[12] Google Scholar

[ref5] 5. Beegle K, Demombynes G, de Walque D, Gubbins P, Veillard J. COVID-19 increased existing gender mortality gaps in high-income more than middle-income countries. Int J Infect Dis. 2024;148:107167. pmid:39233048
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref6] 6. Castro MC, Gurzenda S, Turra CM, Kim S, Andrasfay T, Goldman N. Reduction in life expectancy in Brazil after COVID-19. Nat Med. 2021;27(9):1629–35. pmid:34188224
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref7] 7. Yan BW, Arias E, Geller AC, Miller DR, Kochanek KD, Koh HK. Widening gender gap in life expectancy in the US, 2010-2021. JAMA Intern Med. 2024;184(1):108–10.
View Article
Google Scholar

[22] View Article

[23] Google Scholar

[ref8] 8. Plümper T, Neumayer E. The unequal burden of war: the effect of armed conflict on the gender gap in life expectancy. Int Orga. 2006;60(3):723–54.
View Article
Google Scholar

[25] View Article

[26] Google Scholar

[ref9] 9. Macmillan R, Shofia N, Sigle W. Gender and the politics of death: female representation, political and developmental context, and population health in a cross-national panel. Demography. 2018;55(5):1905–34. pmid:30128940
View Article
PubMed/NCBI
Google Scholar

[28] View Article

[29] PubMed/NCBI

[30] Google Scholar

[ref10] 10. Gutmann J, Matthias N, Neumeier F. Sanctioned to death? The impact of economic sanctions on life expectancy and its gender gap. J Dev Stud. 2021;57(1):139–62.
View Article
Google Scholar

[32] View Article

[33] Google Scholar

[ref11] 11. Neumayer E, Plümper T. The gendered nature of natural disasters: the impact of catastrophic events on the gender gap in life expectancy, 1981–2002. Ann Assoc Am Geogr. 2007;97(3):551–66.
View Article
Google Scholar

[35] View Article

[36] Google Scholar

[ref12] 12. Murillo MQ, Tan S. Discovering the differential and gendered consequences of natural disasters on the gender gap in life expectancy in Southeast Asia. Nat Hazards Earth Syst Sci Discuss. 2017;2017:1–18.
View Article
Google Scholar

[38] View Article

[39] Google Scholar

[ref13] 13. Masquelier B, Reniers G. AIDS and the gender gap in life expectancy in Africa. Popul Societ. 2018;(554):1–4.
View Article
Google Scholar

[41] View Article

[42] Google Scholar

[ref14] 14. Noymer A, Garenne M. The 1918 influenza epidemic’s effects on sex differentials in mortality in the United States. Popul Dev Rev. 2000;26(3):565–81. pmid:19530360
View Article
PubMed/NCBI
Google Scholar

[44] View Article

[45] PubMed/NCBI

[46] Google Scholar

[ref15] 15. International Health Policy Program. Life expectancy and health adjusted life expectancy of Thai population in 2019 - 2021 [in Thai]. Nonthaburi: International Health Policy Program; 2023. https://le-hale.bodthai.net/

[ref16] 16. United Nations Department of Economic and Social Affairs Population Division. World population prospects 2024. New York: United Nations; 2024. https://population.un.org/wpp/

[ref17] 17. De Walque D, Filmer D. Trends and socioeconomic gradients in adult mortality around the developing world. Popul Dev Rev. 2013;39(1):1–29.
View Article
Google Scholar

[50] View Article

[51] Google Scholar

[ref18] 18. Pinho-Gomes A-C, Peters SAE, Woodward M. Gender equality related to gender differences in life expectancy across the globe gender equality and life expectancy. PLOS Glob Public Health. 2023;3(3):e0001214. pmid:36963039
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref19] 19. Pothisiri W, Prasitsiriphon O, Apakupakul J, Ploddi K. Gender differences in estimated excess mortality during the COVID-19 pandemic in Thailand. BMC Public Health. 2023;23(1):1900. pmid:37784059
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref20] 20. National Statistical Office of Thailand. The 2015-2016 survey of population change [in Thai]. Bangkok: National Statistical Office; 2017.

[ref21] 21. Alexander M, Zagheni E, Barbieri M. A flexible bayesian model for estimating subnational mortality. Demography. 2017;54(6):2025–41. pmid:29019084
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref22] 22. Czado C, Delwarde A, Denuit M. Bayesian Poisson log-bilinear mortality projections. Insur Math Econ. 2005;36(3):260–84.
View Article
Google Scholar

[66] View Article

[67] Google Scholar

[ref23] 23. Lee RD, Carter LR. Modeling and forecasting US mortality. J Am Stat Assoc. 1992;87(419):659–71.
View Article
Google Scholar

[69] View Article

[70] Google Scholar

[ref24] 24. Preston S, Heuveline P, Guillot M. Demography: measuring and modeling population processes. Malden, MA: Blackwell Publishers; 2001.

[ref25] 25. Keyfitz N. Finding probabilities from observed rates or how to make a life table. Am Stat. 1970;24(1):28–33.
View Article
Google Scholar

[73] View Article

[74] Google Scholar

[ref26] 26. Scherbov S, Ediev D. Significance of life table estimates for small populations: simulation-based study of estimation errors. Demogr Res. 2011;24(22):527–50.
View Article
Google Scholar

[76] View Article

[77] Google Scholar

[ref27] 27. Ediev DM. Age exaggeration ruses: Infrequent age overstatement distorts the mortality curve at old age. In: Skiadas CH, Skiadas C, editors. Quantitative methods in demography: methods and related applications in the Covid-19 era. Springer; 2022. 189–205.

[ref28] 28. Horiuchi S, Coale AJ. A simple equation for estimating the expectation of life at old ages. Popul Stud (Camb). 1982;36(2):317–26. pmid:22077276
View Article
PubMed/NCBI
Google Scholar

[80] View Article

[81] PubMed/NCBI

[82] Google Scholar

[ref29] 29. Ediev DM. Constrained mortality extrapolation to old age: an empirical assessment. Eur J Popul. 2018;34(3):441–57. pmid:30147209
View Article
PubMed/NCBI
Google Scholar

[84] View Article

[85] PubMed/NCBI

[86] Google Scholar

[ref30] 30. Mitra S. Estimating the expectation of life at older ages. Popul Stud. 1984;38(2):313–9.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref31] 31. Ministry of Public Health. Strategy: managing the new wave of the Covid-19 epidemic. Nonthaburi: Ministry of Public Health; 2021.

[ref32] 32. SAT-MOPH. Situation of COVID-19 patients in Thailand. 2023. https://ddc.moph.go.th/covid19-dashboard/

[ref33] 33. Runfola D, Anderson A, Baier H, Crittenden M, Dowker E, Fuhrig S, et al. geoBoundaries: a global database of political administrative boundaries. PLoS One. 2020;15(4):e0231866. pmid:32330167
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref34] 34. Tonboot S, Wattanadumrong B. Situation of inequality in health utilization among Thai elderly in 2020. In: Business creativity and innovation in the age of transformation: proceedings of the 8th business economics and communications international conference (BECIC 2021). 2021.
View Article
Google Scholar

[97] View Article

[98] Google Scholar

[ref35] 35. Faramnuayphol P, Chongsuvivatwong V, Pannarunothai S. Geographical variation of mortality in Thailand. J Med Assoc Thai. 2008;91(9):1455–60. pmid:18843878
View Article
PubMed/NCBI
Google Scholar

[100] View Article

[101] PubMed/NCBI

[102] Google Scholar

[ref36] 36. Thai Health Project. Thai health 2023: Thailand’s commitment in COP and responses to climate change. Nakhon Pathom: Institute for Population and Social Research, Mahidol University; 2023.

[ref37] 37. United Nations Economic and Social Commission for Asia and the Pacific. Asia-Pacific report on population ageing 2022: trends, policies and good practices regarding older persons and population ageing. Bangkok: United Nations; 2022.

[ref38] 38. Ho C. Asian men’s health: gender disparity and life expectancy. J Surg Acad. 2014;4:5–8.
View Article
Google Scholar

[106] View Article

[107] Google Scholar

[ref39] 39. Regitz‐Zagrosek V. Sex and gender differences in health. EMBO Reports. 2012;13(7):596–603.
View Article
Google Scholar

[109] View Article

[110] Google Scholar

[ref40] 40. Thornton J. WHO report shows that women outlive men worldwide. BMJ. 2019;365:l1631. pmid:30952650
View Article
PubMed/NCBI
Google Scholar

[112] View Article

[113] PubMed/NCBI

[114] Google Scholar

[ref41] 41. Peckham H, de Gruijter NM, Raine C, Radziszewska A, Ciurtin C, Wedderburn LR, et al. Male sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU admission. Nat Commun. 2020;11(1):6317. pmid:33298944
View Article
PubMed/NCBI
Google Scholar

[116] View Article

[117] PubMed/NCBI

[118] Google Scholar

[ref42] 42. Rehman S, Ravinayagam V, Nahvi I, Aldossary H, Al-Shammari M, Amiri MSA, et al. Immunity, sex hormones, and environmental factors as determinants of COVID-19 disparity in women. Front Immunol. 2021;12:680845. pmid:34484179
View Article
PubMed/NCBI
Google Scholar

[120] View Article

[121] PubMed/NCBI

[122] Google Scholar

[ref43] 43. Aungkulanon S, Tangcharoensathien V, Shibuya K, Bundhamcharoen K, Chongsuvivatwong V. Post universal health coverage trend and geographical inequalities of mortality in Thailand. Int J Equity Health. 2016;15(1):190. pmid:27876056
View Article
PubMed/NCBI
Google Scholar

[124] View Article

[125] PubMed/NCBI

[126] Google Scholar

[ref44] 44. National Economic and Social Development Council. Situation analysis report on poverty and inequality in Thailand. Bangkok: National Economic and Social Development Council; 2024.

[ref45] 45. Yiengprugsawan V, Lim LL-Y, Carmichael GA, Seubsman S-A, Sleigh AC. Tracking and decomposing health and disease inequality in Thailand. Ann Epidemiol. 2009;19(11):800–7. pmid:19560371
View Article
PubMed/NCBI
Google Scholar

[129] View Article

[130] PubMed/NCBI

[131] Google Scholar

[ref46] 46. Wakabayashi M, McKetin R, Banwell C, Yiengprugsawan V, Kelly M, Seubsman S, et al. Alcohol consumption patterns in Thailand and their relationship with non-communicable disease. BMC Public Health. 2015;15:1297. pmid:26704520
View Article
PubMed/NCBI
Google Scholar

[133] View Article

[134] PubMed/NCBI

[135] Google Scholar

[ref47] 47. International Health Policy Program. Burden of disease attributable to risk factors in Thailand 2019 [in Thai]. Nonthaburi: International Health Policy Program; 2023.

[ref48] 48. Jitnarin N, Kosulwat V, Rojroongwasinkul N, Boonpraderm A, Haddock CK, Poston WSC. Socioeconomic status and smoking among thai adults: results of the National Thai Food Consumption Survey. Asia Pac J Public Health. 2011;23(5):672–81. pmid:20460275
View Article
PubMed/NCBI
Google Scholar

[138] View Article

[139] PubMed/NCBI

[140] Google Scholar

[ref49] 49. GBD 2019 Tobacco Collaborators. Spatial, temporal, and demographic patterns in prevalence of smoking tobacco use and attributable disease burden in 204 countries and territories, 1990–2019: a systematic analysis from the Global Burden of Disease Study 2019. Lancet. 2021;397(10292):2337–60.

[ref50] 50. Oksuzyan A, Juel K, Vaupel JW, Christensen K. Men: good health and high mortality. Sex differences in health and aging. Aging Clin Exp Res. 2008;20(2):91–102. pmid:18431075
View Article
PubMed/NCBI
Google Scholar

[143] View Article

[144] PubMed/NCBI

[145] Google Scholar

[ref51] 51. National Academies of Sciences, Engineering, and Medicine. The future of nursing 2020-2030: charting a path to achieve health equity. Flaubert JL, Le Menestrel S, Williams DR, Wakefield MK, editors. Washington (DC): National Academies Press (US); 2021.

[ref52] 52. Khavjou O, Tayebali Z, Cho P, Myers K, Zhang P. Rural-urban disparities in state-level diabetes prevalence among US adults, 2021. Prev Chronic Dis. 2025;22:E05. pmid:39819894
View Article
PubMed/NCBI
Google Scholar

[148] View Article

[149] PubMed/NCBI

[150] Google Scholar

[ref53] 53. Passarelli-Araujo H, de Souza GM. Urban-rural health disparities in Brazil: do sociodemographic attributes play a role?. Prev Med. 2023;175:107679. pmid:37625652
View Article
PubMed/NCBI
Google Scholar

[152] View Article

[153] PubMed/NCBI

[154] Google Scholar

[ref54] 54. Johri M, Agarwal S, Khullar A, Chandra D, Pratap VS, Seth A. The first 100 days: how has COVID-19 affected poor and vulnerable groups in India?. Health Promot Inter. 2021;36(6):1716–26. pmid:34002217
View Article
PubMed/NCBI
Google Scholar

[156] View Article

[157] PubMed/NCBI

[158] Google Scholar

[ref55] 55. Rönkkö R, Rutherford S, Sen K. The impact of the COVID-19 pandemic on the poor: insights from the Hrishipara diaries. World Dev. 2022;149:105689. pmid:34697515
View Article
PubMed/NCBI
Google Scholar

[160] View Article

[161] PubMed/NCBI

[162] Google Scholar

[ref56] 56. Scully EP, Haverfield J, Ursin RL, Tannenbaum C, Klein SL. Considering how biological sex impacts immune responses and COVID-19 outcomes. Nat Rev Immunol. 2020;20(7):442–7. pmid:32528136
View Article
PubMed/NCBI
Google Scholar

[164] View Article

[165] PubMed/NCBI

[166] Google Scholar

[ref57] 57. Kosulwat V. The nutrition and health transition in Thailand. Public Health Nutr. 2002;5(1A):183–9. pmid:12027283
View Article
PubMed/NCBI
Google Scholar

[168] View Article

[169] PubMed/NCBI

[170] Google Scholar

[ref58] 58. Gadsden T, Downey LE, Vilas VDR, Peiris D, Jan S. The impact of COVID-19 on essential health service provision for noncommunicable diseases in the South-East Asia region: a systematic review. Lancet Reg Health Southeast Asia. 2022;1:100010. pmid:35769108
View Article
PubMed/NCBI
Google Scholar

[172] View Article

[173] PubMed/NCBI

[174] Google Scholar

[ref59] 59. Kharroubi SA, Diab-El-Harake M. Sex-differences in COVID-19 diagnosis, risk factors and disease comorbidities: a large US-based cohort study. Front Public Health. 2022;10:1029190. pmid:36466473
View Article
PubMed/NCBI
Google Scholar

[176] View Article

[177] PubMed/NCBI

[178] Google Scholar

[ref60] 60. Odton P, Choonpradub C, Bundhamcharoen K. Geographical variations in all-cause mortality in Thailand. Southeast Asian J Trop Med Public Health. 2010;41:1209–19.
View Article
Google Scholar

[180] View Article

[181] Google Scholar

[ref61] 61. Greenpeace Southeast Asia. Greenpeace Southeast Asia annual report 2022. Bangkok: Greenpeace Southeast Asia; 2022.

[ref62] 62. Guo J, Chai G, Song X, Hui X, Li Z, Feng X, et al. Long-term exposure to particulate matter on cardiovascular and respiratory diseases in low- and middle-income countries: a systematic review and meta-analysis. Front Public Health. 2023;11:1134341. pmid:37056647
View Article
PubMed/NCBI
Google Scholar

[184] View Article

[185] PubMed/NCBI

[186] Google Scholar

[ref63] 63. Lelieveld J, Klingmüller K, Pozzer A, Pöschl U, Fnais M, Daiber A, et al. Cardiovascular disease burden from ambient air pollution in Europe reassessed using novel hazard ratio functions. Eur Heart J. 2019;40(20):1590–6. pmid:30860255
View Article
PubMed/NCBI
Google Scholar

[188] View Article

[189] PubMed/NCBI

[190] Google Scholar

[ref64] 64. Zhang J, Wang X, Yan M, Shan A, Wang C, Yang X, et al. Sex differences in cardiovascular risk associated with long-term PM2.5 exposure: a systematic review and meta-analysis of cohort studies. Front Public Health. 2022;10:802167. pmid:35186842
View Article
PubMed/NCBI
Google Scholar

[192] View Article

[193] PubMed/NCBI

[194] Google Scholar

[ref65] 65. Institute for Health Metrics and Evaluation IHME. Global burden of disease 2021: findings from the GBD 2021 study. Seattle, WA: IHME; 2024.

[ref66] 66. Sanderson WC, Scherbov S. The effect of the COVID-19 pandemic on life expectancy in the USA: an application of hybrid life expectancy. Biol Methods Protoc. 2023;8(1):bpad025. pmid:37929068
View Article
PubMed/NCBI
Google Scholar

[197] View Article

[198] PubMed/NCBI

[199] Google Scholar

Figures

Abstract

Background

Methods

Results

Conclusions

Introduction

Materials and methods

Data sources

Population data

Mortality data

PM2.5 exposure data

Statistical model

Uncertainty estimation

Life table computations

Treatment of specific age groups

Ethical considerations

Results

National trends in life expectancy at birth (e₀)

Provincial trends in life expectancy at birth

Gender gap in life expectancy at birth

Trends in life expectancy at age 65 and gender gap

Probability to survive from age 20 to age 65

Association between PM2.5 and life expectancy

Discussion

Strengths and limitations

Conclusion

Supporting information

S1 Fig. Sensitivity of GAM-derived to oldest-old (85+) closure.

S1 Table. Life expectancy at birth () by province and sex, Thailand, 2023.

S2 Table. Life expectancy at age 65 () by province and sex, Thailand, 2023.

S3 Table. Probability of surviving from age 20–65 () by province and sex, Thailand, 2023.

Acknowledgments

References