This study offers a comprehensive analysis of life expectancy patterns across Western European border regions, using novel granular mortality data to identify potentially successful regions. We examined life expectancy trends in 277 border regions in 12 Western European countries from 1995 to 2019, measuring improvements in these regions by comparing them with neighbouring border regions, non-border regions of the same country, and all selected countries combined. This analysis yielded two major results. First, despite overall increases in life expectancy in all border regions, the pace of improvement varied notably across countries and regions, revealing clusters of successful and less successful border regions regarding mortality convergence. Second, we found that disparities between border regions of neighbouring countries were more pronounced than those between border and non-border regions within the same country, suggesting that, despite European integration efforts, national and regional contexts significantly shape mortality patterns across Europe.

Convergence and divergence patterns of life expectancy trajectories from 1995 to 2019 between individual border regions (NUTS-3) and the third reference group, the average of all 12 selected countries

The observed increases in life expectancy in all border regions are consistent with previous studies on mortality improvements at national [12, 36], provincial [13, 19], and district [23] levels. Despite these overall gains, the absolute differences in life expectancy between individual border regions remain large, suggesting that there is no uniform European pattern. Although greater gains in life years in certain regions could be explained by European integration, the persistent large differences between them are most likely determined by country- and region-specific factors. Moreover, the differences in life expectancy were more pronounced when comparing border regions with those of neighbouring countries, as opposed to the non-border regions of the same country. This further underscores the importance of the national and regional context in shaping mortality differences, despite ongoing harmonisation efforts at the EU level. EU cohesion policies have been an important tool in addressing regional disparities. Between 2014 and 2020, approximately €347 billion, accounting for nearly one-third of the total EU budget, was allocated to this policy [37]. A substantial portion of these funds is designated for the European Regional Development Fund, which aims to mitigate developmental disparities across European regions, prioritising those that are disadvantaged [38]. Despite these efforts, there seems to be limited success in achieving the intended outcomes, as reflected in persistent mortality disparities. While life expectancy has increased overall, the timing and magnitude of these improvements differ across regions. Some border regions show strong improvements in life expectancy, while others lag compared to neighbouring border regions, non-border regions or the Western European average.

At the national level, Vallin and Meslé have identified vanguards– countries leading in the international ranking– and laggards– countries following the vanguards at slower or similar improvement rates [16]. Sauerberg and colleagues confirmed the complex landscape of mortality development by highlighting the dichotomy of leading and lagging regions at the district level [23]. While these previous studies have focused primarily on the East-West mortality gap, our research reveals that persistent mortality differences also exist within and across Western European border regions. As a result, we identified distinct clusters of leading and lagging border regions concerning mortality convergence.

For example, the Alpine regions, especially along the Swiss and Italian borders, show increasing advantages in life expectancy, thus exceeding the Western European average, for men. Additionally, northern Italian border regions are not only catching up but are even outperforming the predominantly southern, non-border regions. However, this trend may not be solely attributed to their border status but rather to broader regional patterns in Italy. Previous research has found gains in various provinces and districts of northern Italy [13, 23]. Nevertheless, male mortality in the north has not always been low and leading. This reciprocal north-south shift in mortality has been attributed to decreases in lung cancer and ischemic heart diseases in the north [39]. Moreover, the economically prosperous northern regions have been more effective in reducing their initially high mortality levels among adult men. Examining the Swiss side of the Alpine border region reveals the highest life expectancy figures, which is consistent with Switzerland’s status as one of the wealthiest nations in Europe, as measured by GDP per capita [40]. This economic prosperity is correlated with life expectancy outcomes [2, 4]. In addition, the specific characteristics of the Alpine environment, particularly its mountainous terrain, have been associated with lower mortality rates, as residing at high altitudes may confer health benefits [41].

Furthermore, we detected a convergence in life expectancy between the Nordic border regions, with the Finnish and Danish border regions catching up with their neighbouring Swedish border regions. Sweden’s global ranking has declined as mortality at higher ages decreases more slowly at older ages [42]. Improvement for Finland has been much faster than in Sweden [43], resulting in mortality convergence at national and regional levels. This is in line with our findings, which suggest convergence in the Finnish-Swedish border region.

The Finnish border region has made progress in catching up with its neighbouring Swedish regions and the rest of Finland, yet it still lags behind. This region, located in Finland’s far north, suffers from mortality disadvantages typically associated with peripheral areas. Previous research highlights higher mortality rates in northern Finland compared to western Finland [43]. The underlying reasons for these patterns remain somewhat elusive, though a higher prevalence of risky health behaviours in the north-east and possible genetic differences have been proposed as contributing factors [44].

Within the economic centre of Western Europe, we observed another cluster of lagging regions along the western German border. These areas reported the lowest life expectancy in 2019 and showed minimal gains in life expectancy for both sexes over time. This trend reflects previous studies that identified Germany as “one of the worst performers among high-income countries” regarding life expectancy at the national level [45]. Factors contributing to this include higher mortality rates from cardiovascular diseases and potential shortcomings in primary care and disease prevention. We also observed overall negative trends for western German border regions compared with border regions in neighbouring countries and the Western European average. For the latter, these disadvantaging trends extend to the Dutch, Belgian, and parts of the French border regions among females. This finding is particularly surprising given that, as mentioned previously, these border regions are located within a highly industrialised area of Europe, also known as the European dorsal [46]. This densely populated corridor, stretching from north-western England through the Benelux countries and western Germany to northern Italy, is known for its economic prosperity and advanced infrastructure. These factors could arguably have a positive impact on life expectancy. However, environmental and socio-economic factors may counteract these benefits. For example, industrial areas often face higher pollution levels, which leads to lower life expectancy [47]. Additionally, these regions have experienced significant deindustrialisation since the 1970s, leading to high unemployment [48] and selective out-migration, contributing to higher mortality rates as healthier, working-age individuals may move elsewhere in search of better opportunities [49]. Yet, these disadvantaging trends in life expectancy apply only to women.

Our sensitivity check reveals that while some regional classifications change when using a shorter time frame, the overall patterns largely remain consistent. For example, the border regions between Italy and France, as well as Italy and Switzerland, shift from showing convergence to displaying only minor changes. In contrast, for women in the German-Dutch and German-Belgian regions, the classification changes from minor changes to divergence, indicating short-term fluctuations. Additionally, the Danish border regions have shown increasing advantages since 2000, whereas they were previously categorised as experiencing decreasing disadvantages over a longer period. Despite these specific shifts, our sensitivity check confirms that the results remain stable in most cases when the starting point is adjusted to 2000. Most importantly, when comparing the border regions to the Western European average, the broader spatial patterns and cluster structures remain unchanged, with only a few isolated instances of category shifts.

While our findings are robust across various starting years, a two-time-point approach for determining convergence should be used with caution. This approach groups regions based on differences in life expectancy at the start and end points, potentially overlooking dissimilar intermediate trends. Although this approach helps summarise long-term patterns and offers easier interpretability, it may obscure short-term temporal dynamics. Moreover, our study utilised highly granular geographical data at the district level, providing a unique opportunity to explore an understudied group of regions: cross-border regions. This approach allowed for the identification of localised clusters of high and low mortality, providing valuable insights for targeted health interventions. However, using small-area death counts posed some challenges, particularly in sparsely populated regions. Sparse data can lead to high variability and potential instability in model estimates. To address these fluctuations, we employed penalised spline models [29], which effectively smooth irregularities in spatial or temporal data while preserving the underlying data structures [50] Despite their advantages, smoothing techniques can lead to over- or under-smoothing of critical trends. In areas with very few deaths, mortality patterns may be more influenced by random variation, potentially concealing important local differences [51]. However, as our observation period spans 24 years, individual outliers are unlikely to alter the overall pattern. Furthermore, our research investigated long-term mortality trends within a consistent demographic and epidemiological framework, intentionally excluding the COVID-19 period from our analysis. The pandemic has introduced exceptional short-term fluctuations in mortality rates, which may be especially relevant for border regions. These dynamics are inherently different from the trends observed before the pandemic. Consequently, incorporating these disparate periods could obscure the structural patterns we intend to scrutinise. Finally, our analysis is confined to the border regions of Western Europe, which have benefited from an extended period of open borders and are characterised by comparatively low mortality rates. The inclusion of Eastern European border regions could potentially uncover additional patterns worthy of investigation in future research; however, such inclusion falls outside the scope of our study.

Despite these limitations, our study adds to the literature in several ways. It provides a novel perspective on mortality convergence through the lens of cross-border life expectancy trends. Furthermore, we identify distinct clusters of leading and lagging regions across national borders, such as the high-performing Alpine regions of Italy and Switzerland, and the underperforming areas in the old industrial belt.

Our study adopts an intuitive approach to promote accessibility, yet its findings also carry significant policy implications. Our findings align with ongoing EU efforts to harmonise living standards, including health outcomes, and underline the challenges that remain in achieving this goal. While EU initiatives such as the directive on patient’s rights in cross-border healthcare aim to enhance healthcare coordination and accessibility across borders [11], their full potential has yet to be realised [52]. Furthermore, our findings suggest that differences between countries play a greater role in life expectancy disparities than a region’s location within a country, whether inland or along the border. These insights should be considered when developing new European Union directives and regional initiatives to enhance health integration.

To address these challenges, the EU could consider revising the allocation criteria for regional and cohesion funds, which currently rely heavily on GDP measures [37] and is motivated by economic convergences. Incorporating health indicators, such as mortality measures, might better target regions facing significant health disparities. Furthermore, the EU could explore developing tailored regional policies that extend national borders to address specific public health challenges in border areas. Cross-border regions, given their unique positioning, serve as experimental laboratories for policy learning and collaboration, enabling lagging regions to learn from the more successful. In this regard, initiatives like Interreg, which foster cross-border cooperation in healthcare, infrastructure, and economic development, are vital for promoting regional integration. Our study serves as an initial step in evaluating the impact of such initiatives by identifying ongoing disparities and pinpointing areas where closer collaboration could lead to meaningful health improvements.

Furthermore, the persistent disparities in Western European border regions highlight the important influence of structural and contextual factors, such as economic conditions, healthcare accessibility, and demographic composition, on these trends. It is essential to address these underlying differences to understand why some regions continue to fall behind, even as broader improvements are made. Thus, to improve the effectiveness of EU funding tools, future research should focus on structural and contextual factors and seek to disentangle the role of EU, national and regional policies. Cross-border regions offer a great setting to do this. To our knowledge, this is the first study that offers new insights into how regional trajectories of border regions may converge or diverge in the context of continued European integration. Finally, it provides a departure point for future studies aimed at unravelling the complex factors shaping health outcomes in European border regions.