Under-5 mortality trends and leading causes of death in Greece, 2016-2020: A nationwide study Authors: Maria Tzoraki, Eva-Maria Igoumenidi, Georgios Naltsas, Georgia Kourlaba, Tania Siahanidou, Dimitra Metallinou Published: 2026 ABSTRACT ObjectivesTo examine recent trends and cause-specific patterns of under-5 mortality in Greece and to identify demographic and regional variations using ICD-10–based national data.Study designNationwide population-based retrospective study.MethodsWe conducted a nationwide analysis of all under-5 deaths recorded by the Hellenic Statistical Authority (ELSTAT) from 2016 to 2020. Neonatal mortality rate (NMR), infant mortality rate (IMR), and under-5 mortality rate (U5MR) were calculated per 1000 live births. Causes of death were analyzed using individual ICD-10 codes, and age-, sex-, and region-specific patterns were examined. Mortality trends were contextualized using data from the World Health Organization (WHO), United Nations Inter-Agency Group for Child Mortality Estimation (UN IGME), and Eurostat.ResultsA total of 1877 under-5 deaths were recorded, 58.3% of which occurred during the neonatal period. Mortality declined steadily, with U5MR decreasing by 22.9%, from 4.97 to 3.82 per 1000 live births. Deaths were predominantly attributed to perinatal (P00-P96) and congenital anomalies (Q00-Q99) ICD-10 categories (>70%). Mortality was higher among males (56%) and concentrated in regions hosting major tertiary hospitals. In 2020, Greece's U5MR ranked in the mid-range among EU countries.ConclusionsDespite declining under-5 mortality, preventable neonatal causes remain prominent in Greece. Strengthening perinatal and neonatal care and targeted prevention strategies are essential to further improve outcomes and align with leading European countries. Introduction The under-5 mortality rate (U5MR), the probability of dying before five years of age per 1000 live births, is a fundamental indicator of population health, health-system performance and socioeconomic development [1]. Global child survival has improved substantially over the past three decades: between 1990 and 2019, the U5MR declined by nearly 60%, from 93 to 38 deaths per 1000 live births [2]. Despite this progress, preventable deaths remain a major global health challenge, with almost five million children under-5 dying in 2020, and the neonatal period accounting for nearly half of these deaths [3]. In recognition of this persistent burden, the Sustainable Development Goals (SDGs) call for all countries to reduce neonatal mortality to ≤12 and under-5 mortality to ≤25 deaths per 1000 live births by 2030 [4]. High-income European countries have achieved some of the lowest U5MR globally. According to estimates from the United Nations Inter-agency Group for Child Mortality Estimation (UN IGME), several Western and Southern European countries reported U5MRs below 4 deaths per 1000 live births in 2020, including Italy (2.9) and Spain (3.2), while higher rates persisted in countries such as France (4.5) and in several Eastern European countries, including Romania (6.7) and Bulgaria (6.5) [2]. Greece's U5MR of approximately 3.8 deaths per 1000 live births places the country in the mid-range of European Union Member States [2]. Consistent with this pattern, Eurostat data indicate that Greece's infant mortality rate (IMR) was approximately 3.4 deaths per 1000 live births in 2020, comparable to France (3.5), slightly higher than Italy and Spain (both at 2.5 - 2.7), and substantially lower than rates observed in Eastern European countries, such as Romania and Bulgaria (at 5.6 - 5.7 deaths per 1000 live births) [5]. Evidence from national time-trend analyses and burden-of-disease assessments further indicates that infant and child mortality trends in Greece have been sensitive to broader socioeconomic conditions, with the economic crisis of 2008 - 2016 associated with a temporary disruption of previously declining mortality patterns [6,7]. In high-income settings, the causes of death in early childhood differ markedly from those in low- and middle-income countries. Deaths are predominantly due to neonatal complications, especially prematurity, low birth weight, perinatal asphyxia and congenital anomalies, while infectious diseases and external causes are rare [8,9]. Greece reflects this profile but also has several distinguishing characteristics that may influence mortality patterns. The country has one of the highest caesarean section rates in Europe, exceeding 50%, a factor that has been associated with increased risks of iatrogenic prematurity and neonatal morbidity [10,11]. In addition, geographic disparities exist, with specialized neonatal intensive care units (NICUs) being concentrated in urban centers, raising concerns regarding timely access to advanced perinatal and neonatal care for families residing in rural or island regions. Despite the relevance of these structural and epidemiological factors, evidence on under-5 mortality in Greece remains limited. Existing national analyses have focused primarily on infant mortality trends, often without detailed ICD-10 cause-specific categorization or regional breakdowns. To date, no study has comprehensively examined recent national under-5 mortality data in Greece using detailed age-, sex-, cause-, and region-specific analyses. The present study addresses this gap by analyzing all deaths among children under-5 years of age in Greece between 2016 and 2020 using mortality records from the Hellenic Statistical Authority (ELSTAT). We describe trends in neonatal, infant, and under-5 mortality; identify leading causes of death; explore demographic and geographic patterns; and situate Greece's mortality profile within the broader European and global context. This study provides an updated nationwide epidemiological assessment of early childhood mortality in Greece using detailed ICD-10 cause-specific data and aims to inform public health priorities for reducing preventable deaths. Methods Data on all deaths among children under-5 years of age in Greece from 2016 to 2020 (01/01/2016 - 31/12/2020) were obtained directly from ELSTAT following an official request [12]. The dataset included the year of death, age at death, sex, nationality, place of death, and the underlying cause of death coded according to the International Classification of Diseases, 10th Revision (ICD-10) [13]. Corresponding annual live births were retrieved from ELSTAT Vital Statistics reports [[14], [15], [16], [17], [18]] for the same period and were used as denominators for mortality rate calculations. For the calculation of mortality indicators, infant deaths were defined as all deaths occurring before one completed year of age and therefore included both neonatal deaths (<28 completed days) and post-neonatal deaths (28 days to <1 year). For cause-specific analyses, infant deaths were examined separately for the neonatal and post-neonatal periods in order to distinguish deaths occurring before and after 28 completed days of life. To facilitate international comparisons, we used publicly available estimates from the UN IGME, UNICEF's Levels and Trends in Child Mortality reports, and Eurostat demographic databases [2,5,19,20]. Geographical analyses were performed according to the Nomenclature of Territorial Units for Statistics (NUTS-2) classification of Greece, as defined by ELSTAT in the 2020 Vital Statistics report [18]. Greece's 13 NUTS-2 regions include: Eastern Macedonia and Thrace, Central Macedonia, Western Macedonia, Thessaly, Epirus, Ionian Islands, Western Greece, Central Greece, Peloponnese, Attica, North Aegean, South Aegean, and Crete. Definitions ●Live birth: Any complete expulsion or extraction of a fetus from the mother, regardless of gestational age, that shows at least one sign of life - such as heartbeat, respiration, or movement - consistent with United Nations standards [1].●Neonatal death: Death of a live-born neonate occurring before 28 completed days of life [1,14].●Infant death: Death occurring before 1 completed year of life [1,14].●Under-5 death: Death occurring before 5 completed years of life [1,14]. ●Live birth: Any complete expulsion or extraction of a fetus from the mother, regardless of gestational age, that shows at least one sign of life - such as heartbeat, respiration, or movement - consistent with United Nations standards [1]. Live birth: Any complete expulsion or extraction of a fetus from the mother, regardless of gestational age, that shows at least one sign of life - such as heartbeat, respiration, or movement - consistent with United Nations standards [1]. ●Neonatal death: Death of a live-born neonate occurring before 28 completed days of life [1,14]. Neonatal death: Death of a live-born neonate occurring before 28 completed days of life [1,14]. ●Infant death: Death occurring before 1 completed year of life [1,14]. Infant death: Death occurring before 1 completed year of life [1,14]. ●Under-5 death: Death occurring before 5 completed years of life [1,14]. Under-5 death: Death occurring before 5 completed years of life [1,14]. Classification of causes of death Underlying causes of death were coded according to the International Classification of Diseases, 10th Revision (ICD-10), which was applied consistently throughout the study period. Minor updates introduced by ELSTAT after 2018 followed revisions recommended by the World Health Organization [13]. All analyses were conducted using the underlying cause of death at the individual ICD-10 code level, as recorded by ELSTAT, without aggregation into broader cause categories, in order to preserve the original official classification system, including unspecified categories where applicable. Mortality indicators We calculated three standard mortality indicators:●Neonatal mortality rate (NMR): Number of deaths occurring before 28 completed days of life per 1000 live births [20].●IMR: Number of deaths occurring before 1 completed year of life per 1000 live births [21].●U5MR: Number of deaths occurring before 5 completed years of life per 1000 live births [22]. ●Neonatal mortality rate (NMR): Number of deaths occurring before 28 completed days of life per 1000 live births [20]. Neonatal mortality rate (NMR): Number of deaths occurring before 28 completed days of life per 1000 live births [20]. ●IMR: Number of deaths occurring before 1 completed year of life per 1000 live births [21]. IMR: Number of deaths occurring before 1 completed year of life per 1000 live births [21]. ●U5MR: Number of deaths occurring before 5 completed years of life per 1000 live births [22]. U5MR: Number of deaths occurring before 5 completed years of life per 1000 live births [22]. Annual values for all three indicators were derived using ELSTAT mortality data as numerators and corresponding annual live birth counts as denominators. Statistical analysis Annual NMR, IMR, U5MR were calculated per 1000 live births, with 95% confidence intervals estimated assuming a Poisson distribution for the observed number of deaths. Temporal trends during 2016 - 2020 were evaluated using Poisson regression models with calendar year entered as a continuous variable and the number of live births used as the offset variable. Mortality rate ratios (MRRs), annual percentage change (APC), 95% confidence intervals (CIs), and p-values were calculated. Statistical analyses were performed using standard statistical methods for count data. A p-value <0.05 was considered statistically significant. Ethics and data protection This study was conducted using anonymized national mortality and birth data obtained from the ELSTAT [12] following an official data request. All required confidentiality and data-protection agreements were reviewed and signed prior to data release. The dataset contained de-identified individual-level administrative records with no direct or indirect personal identifiers and no possibility of re-identification of individuals. In accordance with the General Data Protection Regulation (EU Regulation 2016/679), anonymized data are not considered personal data and therefore fall outside the scope of data protection legislation (Recital 26). As the study involved secondary analysis of anonymized administrative data, formal approval by a research ethics committee was not required. Consequently, the study was exempt from ethics committee review. Results Overall mortality trends and sex differences During 2016 - 2020, a total of 1877 deaths among children under-5 years of age were recorded in Greece. Mortality rates declined overall during the study period, although year-to-year variation was observed (Fig. 1).Fig. 1Trends in NMR, IMR, and U5MR in Greece, 2016 - 2020. Rates are expressed per 1000 live births. Trends in NMR, IMR, and U5MR in Greece, 2016 - 2020. Rates are expressed per 1000 live births. The NMR decreased from 2.85 to 2.30, the IMR from 4.17 to 3.24 and the U5MR from 4.97 to 3.82 per 1000 live births between 2016 and 2020. Poisson regression analyses demonstrated statistically significant downward trends for IMR and U5MR, whereas the reduction in NMR did not reach conventional statistical significance (Table 1). Neonatal deaths consistently accounted for the largest proportion of under-5 deaths.Table 1Temporal trend analysis of mortality indicators in Greece, 2016 - 2020.Table 1Indicator2016 rate2020 rateAPC (%)MRR95% CIp-valueNMR2.852.30−3.50.9650.925-1.0060.092IMR4.173.24−4.30.9570.924-0.9910.013U5MR4.973.82−4.20.9580.928-0.9890.008Abbreviations: APC: annual percentage change; CI: confidence interval; IMR: infant mortality rate; MRR: mortality rate ratio; NMR: neonatal mortality rate; U5MR: under-5 mortality rate. Table 1Temporal trend analysis of mortality indicators in Greece, 2016 - 2020.Table 1Indicator2016 rate2020 rateAPC (%)MRR95% CIp-valueNMR2.852.30−3.50.9650.925-1.0060.092IMR4.173.24−4.30.9570.924-0.9910.013U5MR4.973.82−4.20.9580.928-0.9890.008Abbreviations: APC: annual percentage change; CI: confidence interval; IMR: infant mortality rate; MRR: mortality rate ratio; NMR: neonatal mortality rate; U5MR: under-5 mortality rate. Temporal trend analysis of mortality indicators in Greece, 2016 - 2020. Abbreviations: APC: annual percentage change; CI: confidence interval; IMR: infant mortality rate; MRR: mortality rate ratio; NMR: neonatal mortality rate; U5MR: under-5 mortality rate. Sex-specific analyses showed persistently higher mortality rates among males across all indicators. Male NMR declined from 3.44 to 2.66 per 1000 live births between 2016 and 2020, compared with a decline from 2.22 to 1.91 among females. Similarly, male IMR decreased from 5.04 to 3.86, compared with 3.27 to 2.57 among females, while male U5MR declined from 5.97 to 4.44, compared with 3.96 to 3.17 among females. Across all years, mortality rates remained consistently higher among males, although both sexes followed similar downward trends. Males also accounted for a greater proportion of under-5 deaths than females across all study years and age groups (56% vs. 44%) (Fig. 2).Fig. 2Sex distribution of under-5 deaths in Greece, 2016 - 2020. Sex distribution of under-5 deaths in Greece, 2016 - 2020. Age-specific mortality Of the 1877 under-5 deaths recorded during 2016 - 2020, 1094 deaths (58.3%) occurred during the neonatal period (<28 days), underscoring the concentration of mortality in the earliest phase of life. In contrast, deaths among children aged from 1 year to <5 years were comparatively uncommon, reflecting the substantial decline in mortality risk after infancy (Table 2).Table 2Under-5 deaths by age group, Greece 2016 - 2020. Percentages are calculated using total under-5 deaths for each year as the denominator.Table 2YearNeonatal deaths (<28d)Post-neonatal deaths (28d to < 1 y)Deaths from 1y to < 5yTotal under-5 deaths2016265 (57.4%)122 (26.4%)75 (16.2%)4622017210 (58.5%)96 (26.7%)53 (14.8%)3592018206 (58.7%)94 (26.8%)51 (14.5%)3512019218 (57.2%)95 (24.9%)68 (17.8%)3812020195 (60.2%)80 (24.7%)49 (15.1%)324Abbreviations: d: days; m: months; y: year. Table 2Under-5 deaths by age group, Greece 2016 - 2020. Percentages are calculated using total under-5 deaths for each year as the denominator.Table 2YearNeonatal deaths (<28d)Post-neonatal deaths (28d to < 1 y)Deaths from 1y to < 5yTotal under-5 deaths2016265 (57.4%)122 (26.4%)75 (16.2%)4622017210 (58.5%)96 (26.7%)53 (14.8%)3592018206 (58.7%)94 (26.8%)51 (14.5%)3512019218 (57.2%)95 (24.9%)68 (17.8%)3812020195 (60.2%)80 (24.7%)49 (15.1%)324Abbreviations: d: days; m: months; y: year. Under-5 deaths by age group, Greece 2016 - 2020. Percentages are calculated using total under-5 deaths for each year as the denominator. Abbreviations: d: days; m: months; y: year. Causes of death Causes of death are presented as leading individual ICD-10 codes to preserve diagnostic specificity and ensure transparency and reproducibility. Given the heterogeneity of conditions encompassed within broad ICD-10 chapters, particularly congenital anomalies and perinatal disorders, analyses were conducted at the individual code level without aggregation into broader cause categories. Among neonates (<28 days), mortality was predominantly attributable to perinatal and prematurity-related conditions (Table 3). The leading causes were extreme immaturity (P07.2) and extremely low birth weight (P07.0), which together accounted for more than 30% of neonatal deaths. These were followed by respiratory distress syndrome of the newborn (P22.0) and congenital heart malformations (Q24.9). Chorioamnionitis (P02.7) was the fifth most common cause.Table 3Leading individual ICD-10 causes of death among neonates (<28 days), Greece 2016 - 2020.Table 3ICD-10 codeCause of deathDeaths (n)% of total (n = 1094)P07.2Extreme immaturity of newborn18016.5P07.0Extremely low birth weight newborn16715.3P22.0Respiratory distress syndrome of newborn797.2Q24.9Congenital malformation of heart, unspecified383.5P02.7Newborn affected by chorioamnionitis383.5Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = neonatal deaths, n = 1094. Table 3Leading individual ICD-10 causes of death among neonates (<28 days), Greece 2016 - 2020.Table 3ICD-10 codeCause of deathDeaths (n)% of total (n = 1094)P07.2Extreme immaturity of newborn18016.5P07.0Extremely low birth weight newborn16715.3P22.0Respiratory distress syndrome of newborn797.2Q24.9Congenital malformation of heart, unspecified383.5P02.7Newborn affected by chorioamnionitis383.5Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = neonatal deaths, n = 1094. Leading individual ICD-10 causes of death among neonates (<28 days), Greece 2016 - 2020. Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = neonatal deaths, n = 1094. Collectively, these conditions represented nearly half of all neonatal deaths recorded between 2016 and 2020, underscoring the dominant role of prematurity, perinatal complications, and congenital anomalies in early-life mortality. Among infants aged 28 days to <1 year of age, mortality was primarily driven by congenital and respiratory causes, with prematurity continuing to exert a substantial influence (Table 4). The leading contributors were congenital heart malformations (Q24.9), interstitial pulmonary disease (J84.9), and extreme immaturity (P07.2), which together accounted for approximately 20% of all infant deaths. Sudden infant death syndrome (R95) and pneumonia (J18.9) were also notable, though less common contributors.Table 4Leading individual ICD-10 causes of death among post-neonatal infants (28 days to <1 year), Greece 2016 - 2020.Table 4ICD-10 codeCause of deathDeaths (n)% of total (n = 487)Q24.9Congenital heart malformation, unspecified367.4J84.9Interstitial pulmonary disease, unspecified336.8P07.2Extreme immaturity of newborn275.5R95Sudden infant death syndrome204.1J18.9Pneumonia, unspecified organism153.1Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = infant deaths(28 days to <1 year), n = 487. Table 4Leading individual ICD-10 causes of death among post-neonatal infants (28 days to <1 year), Greece 2016 - 2020.Table 4ICD-10 codeCause of deathDeaths (n)% of total (n = 487)Q24.9Congenital heart malformation, unspecified367.4J84.9Interstitial pulmonary disease, unspecified336.8P07.2Extreme immaturity of newborn275.5R95Sudden infant death syndrome204.1J18.9Pneumonia, unspecified organism153.1Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = infant deaths(28 days to <1 year), n = 487. Leading individual ICD-10 causes of death among post-neonatal infants (28 days to <1 year), Greece 2016 - 2020. Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = infant deaths(28 days to <1 year), n = 487. These findings indicate that congenital anomalies and respiratory conditions become increasingly prominent beyond the neonatal period, while the impact of prematurity remains evident during infancy. Deaths in children aged from 1 year to < 5 years followed a distinct etiological profile (Table 5). The leading causes included congenital heart malformations (Q24.8) and malignant neoplasms of the brain (C71.9), each contributing approximately 4–8% of deaths in this age group. External causes were also prominent, particularly drowning (T75.1), multiple injuries (T06.8), and intracranial trauma (S06.9). This pattern reflects the shift in mortality determinants beyond infancy, with deaths increasingly associated with chronic conditions and preventable injuries.Table 5Leading individual ICD-10 causes of death among children 1 to <5 years, Greece 2016 - 2020.Table 5ICD-10 codeCause of deathDeaths (n)% of total (n = 296)Q24.8Congenital heart malformation, other specified237.8C71.9Malignant neoplasm of brain, unspecified124.1T75.1Effects of drowning/non-fatal submersion124.1T06.8Multiple injuries, unspecified113.7S06.9Intracranial injury, unspecified103.4Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = deaths aged 1 to <5 years, n = 296. Table 5Leading individual ICD-10 causes of death among children 1 to <5 years, Greece 2016 - 2020.Table 5ICD-10 codeCause of deathDeaths (n)% of total (n = 296)Q24.8Congenital heart malformation, other specified237.8C71.9Malignant neoplasm of brain, unspecified124.1T75.1Effects of drowning/non-fatal submersion124.1T06.8Multiple injuries, unspecified113.7S06.9Intracranial injury, unspecified103.4Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = deaths aged 1 to <5 years, n = 296. Leading individual ICD-10 causes of death among children 1 to <5 years, Greece 2016 - 2020. Abbreviations: n: number. Percentages are calculated using the total number of deaths in the relevant age group as the denominator. Denominator = deaths aged 1 to <5 years, n = 296. Geographical distribution The geographical distribution of under-5 deaths between 2016 and 2020 broadly reflected the population structure and regional distribution of live births in Greece, which are heavily concentrated in Attica and Central Macedonia. Most deaths occurred in Attica (n = 992) and Central Macedonia (n = 361), where the country's principal tertiary maternity and pediatric hospitals are located. Moderate numbers were observed in Western Greece, Crete, and Eastern Macedonia and Thrace, whereas all other regions recorded fewer than 100 deaths over the five-year period. Across all regions, neonatal deaths predominated, accounting for nearly 60% of all under 5 deaths. Rural and island regions reported lower absolute numbers of deaths, likely reflecting smaller birth cohorts and dependence on referral pathways to tertiary hospitals. Because regional analyses were based on place of death rather than place of residence, the observed concentration of deaths in metropolitan regions should be interpreted cautiously, as it may primarily reflect referral patterns and the centralization of specialized neonatal and pediatric care services rather than true regional differences in mortality risk (Table 6).Table 6Regional distribution of under-5 deaths by age group.Table 6Region (NUTS-2)Neonatal deaths (<28 d)Post-neonatal deaths (28d to < 1 y)Deaths 1y to < 5yEastern Macedonia &Thrace65304Attica580265147North Aegean522Western Greece552822Western Macedonia353Epirus25104Thessaly471516Ionian Islands232Central Macedonia2158363Crete752511South Aegean1544Peloponnese687Central Greece11011Abbreviations: NUTS-2: Nomenclature of Territorial Units for Statistics, d: days, m: months, y: year. Table 6Regional distribution of under-5 deaths by age group.Table 6Region (NUTS-2)Neonatal deaths (<28 d)Post-neonatal deaths (28d to < 1 y)Deaths 1y to < 5yEastern Macedonia &Thrace65304Attica580265147North Aegean522Western Greece552822Western Macedonia353Epirus25104Thessaly471516Ionian Islands232Central Macedonia2158363Crete752511South Aegean1544Peloponnese687Central Greece11011Abbreviations: NUTS-2: Nomenclature of Territorial Units for Statistics, d: days, m: months, y: year. Regional distribution of under-5 deaths by age group. Abbreviations: NUTS-2: Nomenclature of Territorial Units for Statistics, d: days, m: months, y: year. Nationality Across all age groups, most deaths occurred among children of Greek nationality. Greek nationals accounted for 82.7% of neonatal deaths, 80.6% of deaths among infants aged from 28 days and younger than one year, and 80.4% of deaths among children aged 1 to <5 years, with children of other nationalities accounting for the remaining proportions. Discussion This nationwide analysis provides an updated assessment of child mortality in Greece and demonstrates an overall decline in infant and under-5 mortality between 2016 and 2020, alongside year-to-year variation. By 2020, the U5MR had fallen to 3.8 per 1000 live births, placing Greece in the mid-range among EU countries. Despite this progress, a substantial proportion of deaths remain attributable to preventable or potentially modifiable factors, underscoring the need for continued public health action. Patterns and leading causes The concentration of mortality in the neonatal period highlights the centrality of perinatal care to child survival. Perinatal conditions (P00 - P96) and congenital anomalies (Q00 - Q99) accounted for more than 70% of under-5 deaths, mirroring patterns observed in other high-income settings [1,2]. Prematurity and low birth weight were especially prominent contributors, reflecting both the inherent vulnerability of newborns and the relatively high prevalence of preterm births in Greece. Congenital anomalies, particularly congenital heart disease, chromosomal abnormalities, and diaphragmatic hernia, also remained major causes of mortality across infancy and early childhood. These findings align with global evidence showing that as infectious causes decline, neonatal complications and congenital disorders dominate mortality profiles in high- and middle-income countries [3,23]. As highlighted by Lawn et al. [24], up to 70% of neonatal deaths may be preventable with high-quality antenatal, intrapartum, and immediate postnatal care, including timely corticosteroid administration, skilled birth attendance and essential newborn care. The low proportion of infection-related deaths in Greece reflects strong vaccination coverage and robust postnatal healthcare but also indicates that further reductions in mortality will depend primarily on optimizing perinatal and neonatal services. Comparison with European and international data In 2020, Greece's U5MR of 3.8 per 1000 live births remained higher than that of the best-performing EU countries, including Italy and Spain, but lower than rates reported in several Eastern European countries. At the global level, Greece has already achieved the SDG 3.2 targets of ≤25 under-5 deaths and ≤12 neonatal deaths per 1000 live births [7,25]. Comparable studies from China reveal similar cause-of-death structures, reinforcing the universality of neonatal and congenital causes as predominant drivers of early childhood mortality once infectious diseases recede [9,26]. These parallels emphasize the importance of sustained investments in perinatal and neonatal care even in countries with relatively low child mortality [3,23]. Sex differences The male predominance in under-5 mortality observed in our study (56%) is consistent with international literature and aligns with well-documented biological and physiological vulnerabilities among male newborns [23,[26], [27], [28]]. These include greater susceptibility to respiratory distress, higher rates of preterm delivery and slower physiological adaptation after birth [2,3,23]. International estimates indicate that male infants have a 10–15% higher risk of dying before age five than females, even in settings with universal healthcare access [2]. In Greece, where gender-based disparities in healthcare are minimal, these differences likely reflect intrinsic biological vulnerability. Nevertheless, recognition of sex-specific risks remains clinically relevant, particularly for optimizing respiratory support strategies and the management of premature and high-risk neonates. Health system factors The prominence of neonatal causes of death underscores the critical influence of Greece's health system organization. In Greece, NICUs are heavily concentrated in Athens and Thessaloniki, the two largest cities, contributing to favorable overall neonatal outcomes but creating regional disparities in timely access to advanced perinatal care [12,19]. Greece's complex geography, including numerous islands and remote mainland areas, makes neonatal transport a persistent operational challenge. Evidence from countries with similar geographic constraints, such as Japan, Norway, and Scotland, demonstrates that coordinated neonatal retrieval systems with standardized protocols and dedicated air and ground teams can substantially improve outcomes and reduce inequities [[29], [30], [31], [32]]. Developing similar coordinated national or regional neonatal transport networks, integrated through the National Center for Emergency Assistance and supported by structured referral pathways, could strengthen equity and align Greece more closely with European best practice [19,33]. Another significant system-level issue is Greece's high caesarean section rate, exceeding 50%, compared with less than 20% in Nordic and Western European countries [19]. Previous studies have linked non-medically indicated caesarean delivery with iatrogenic prematurity and neonatal respiratory morbidity [3,26]. However, the present dataset did not include obstetric or patient-level perinatal variables and therefore no direct association between caesarean section rates and neonatal mortality can be established. Consequently, Greece's high caesarean section rate should be interpreted as a possible contextual factor warranting further investigation using linked perinatal and obstetric data. Evidence-based obstetric care practices may contribute to reductions in prematurity-related adverse neonatal outcomes [7,19]. Addressing unnecessary caesarean deliveries remains an important health-system priority through auditing indications, adherence to WHO intrapartum care guidelines and strengthening midwife-led care models that promote physiologic labor practices [26]. Historical and socioeconomic context Greece's long-term mortality trajectory reflects substantial progress but also sensitivity to broader economic and social conditions. Infant mortality declined markedly from the 1950s to reach historically low levels by 2008, before temporary reversals and stagnation were observed during the 2008 - 2016 financial crisis, highlighting the vulnerability of perinatal and child health services to austerity measures and reductions in healthcare financing [7]. While mortality declined again after 2016 [12], the prior reversal emphasizes the importance of resilient perinatal and neonatal care systems that can withstand socioeconomic instability. Beyond clinical and health-system factors, child survival is also influenced by broader social determinants, including socioeconomic conditions, access to care, maternal education, migration status, and regional inequalities. Recent international research has also highlighted the impact of structural socioeconomic disadvantage on child health and developmental outcomes later in life [34]. Although the present dataset did not include individual-level socioeconomic indicators, future studies linking mortality data with perinatal, socioeconomic, and healthcare-access variables would be valuable for identifying potentially avoidable inequalities in early childhood outcomes. Limitations This study has several limitations that should be considered when interpreting the findings. First, causes of death were derived from national death certificates and coded using ICD-10 and therefore, depend on the accuracy and consistency of medical certification and coding practices. Misclassification is possible, particularly for conditions with overlapping clinical features such as complications of prematurity, unspecified congenital anomalies or sudden infant death syndrome. In addition, several leading causes of death were recorded using unspecified ICD-10 categories (e.g., unspecified congenital heart malformations or unspecified pneumonia). These represent valid ICD-10 diagnostic entities routinely used in mortality surveillance and do not necessarily indicate coding errors. Nevertheless, the use of unspecified categories may reduce etiological granularity and limit more detailed interpretation of cause-specific mortality patterns. Second, the use of administrative mortality data precluded adjustment for important clinical and perinatal factors. Consequently, although temporal trend analyses were performed, the available data do not permit causal inference or adjustment for important clinical and perinatal confounders. The relatively short five-year observation period also limits the statistical power of trend analyses and reduces the ability to distinguish sustained long-term changes from short-term year-to-year fluctuations. Third, regional analyses were based on place of death rather than place of residence. As a result, the observed concentration of deaths in metropolitan regions likely reflects referral patterns and the concentration of tertiary perinatal services in major urban centers rather than true regional differences in mortality risk. In addition, small numbers in certain regions and age groups, particularly among children aged 1-4 years, may result in unstable estimates for rare causes of death. Finally, although analysis at the individual ICD-10 code level preserves diagnostic specificity, it may limit direct comparability with international studies that report aggregated cause-of-death categories. Despite these limitations, the nationwide coverage and standardized coding of the data provide a robust overview of under-5 mortality patterns in Greece. Future perspectives Demographic projections indicate that Greece will continue to face profound challenges related to low fertility rates and population ageing [14]. Although these shifts may alter broader health system priorities, maintaining high-quality perinatal and neonatal care must remain central to national health planning. Strengthening prenatal screening programs, improving the early detection and management of congenital anomalies, and expanding neonatal care capacity, particularly in underserved regions, will be essential for sustaining progress. Additional gains may also be achieved through public health initiatives focused on safe infant sleep practices, injury prevention, and maintaining high vaccination coverage. Collectively, these strategies could help Greece continue its trajectory of improving early childhood survival and align outcomes with the best-performing European countries. Conclusions Between 2016 and 2020, under-5 mortality in Greece declined overall, with statistically significant reductions in infant and under-5 mortality rates, although year-to-year variation persisted. Neonatal mortality also decreased, but trends should be interpreted cautiously given the short observation period. Prematurity, congenital anomalies, and, less frequently, injuries and SIDS remained important contributors to early childhood mortality. Greece has achieved the SDG 3.2 targets for neonatal and under-5 mortality and performs favorably within the European context. Nevertheless, further reductions in preventable deaths will require sustained investment in equitable perinatal and neonatal care, improved prenatal screening and early diagnosis, strengthened national prevention strategies, and further evaluation of obstetric care practices, including the high national caesarean section rate. Overall, the findings highlight the importance of accessible, resilient, and high-quality maternal, perinatal, and child health services for sustaining progress in early childhood survival and reducing remaining inequalities. Author contributions Maria Tzoraki: Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization, Project administration. Eva-Maria Igoumenidi: Methodology, Investigation, Data curation, Writing – original draft. Georgios Naltsas: Formal analysis, Writing – original draft, Visualization. Georgia Kourlaba: Methodology, Writing - review & editing, Supervision. Tania Siahanidou: Methodology, Writing - review & editing, Supervision. Dimitra Metallinou: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Supervision, Project administration. Declaration of generative AI and AI-assisted technologies in the manuscript preparation process During the preparation of this work the authors used ChatGPT in order to polish the language, correct grammar and improve readability and transitions between sections. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article. Ethical statement This study was conducted using fully anonymized national mortality and birth data obtained from the Hellenic Statistical Authority (ELSTAT) following an official data request and the completion of all required confidentiality and data-protection agreements. The dataset contained de-identified administrative records with no direct or indirect personal identifiers and no possibility of re-identification. In accordance with the General Data Protection Regulation (EU Regulation 2016/679, Recital 26), anonymized data do not constitute personal data. As the study involved the secondary analysis of fully anonymized administrative data, formal approval by a research ethics committee was not required. Consequently, the study was exempt from ethics committee review. Funding Open access funding provided by HEAL-Link Greece. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.