4.1. Factors Associated with Dinaric Tallness
The findings of this research naturally raise many questions. The most fundamental one is whether this local phenomenon can be explained by some environmental factors, or if we should look for exceptional genetic predispositions of the local population. Height is, by nature, a highly heritable trait [19
], but it is also a very sensitive indicator of living conditions and the role of environment can be huge. Indeed, following the Industrial Revolution, the height of European nations has increased by 10–17 cm since the end of the 19th century [3
]. Detailed analyses of this process have been performed e.g., for Germany [21
], Norway [22
], and the United States [23
]. This increase in stature is driven by several key factors that are closely associated with the rising GDP (gross domestic product) per capita: better nutrition (mainly high-quality proteins from milk, pork, and eggs), declining rates of total fertility (which determine the distribution of resources within families), the absence of infectious diseases that exhaust the growth capacity of the child’s body (which is reflected in lower child mortality), urbanization (which facilitates better access to resources and healthcare), and social equality [3
Previous studies [12
] have shown that when viewed from the perspective of these major environmental factors, the height of the Western Balkan countries is a striking anomaly, in both the European and global context. In fact, a regression model of six socio-economic and three nutritional variables in 119 countries [18
] showed that the predicted male height for Bosnia and Herzegovina (173.5 cm) deviated the most by far (by 7.7 cm) from the true, observed height (181.2 cm), and it was even lower than the predicted height for Albania (177.0 cm). The positive residual in the case of Serbia (5.1 cm) was the third highest in this global sample (176.1 cm vs. 181.2 cm) (after Haiti, which is heavily dependent on international aid). These numbers remain practically the same even in a smaller sample of 96 countries, after the addition of the Gini index (a measure of social inequality) and the replacement of HDI (Human Development Index) with IHDI (inequality-adjusted HDI): 7.4 cm for Bosnia and Herzegovina and 5.1 cm for Serbia.
These eccentric results are easy to understand, when we consider that the GDP in the Netherlands for 2013 was 49,242 USD per capita, whereas that of Montenegro was more than three-times lower (14,870 USD per capita), and the GDP in Bosnia and Herzegovina reached only 12,011 USD per capita in 2015 [24
] (Figure 3
). Even more important are the statistics of dietary protein quality, assessed using the FAOSTAT database [25
]. Indeed, the ‘protein index’ (a ratio between the daily supply of proteins from dairy and pork/wheat), which is the strongest nutritional predictor of height in 44 European countries (r
= 0.62, p
< 0.001; see [18
]), is mostly very low in the Western Balkans (Figure 4
). In any case, it is far below the level of the Netherlands, which has been, until recently, characterized by the highest dietary protein quality in the world. Only Montenegro and Slovenia were among the top 15 ranked European countries between 2010 and 2019, and Croatia got above the European mean only very recently. What is even more stunning is the fact that the generation surveyed in the Dinaric Alps was often growing up in very difficult conditions during a post-war period, when these protein indices were substantially lower.
The knowledge of these dietary factors can illuminate the unexpectedly short stature of local Muslims in Bosnia and Herzegovina, who do not consume pork for religious reasons [12
]. This paradox is particularly apparent in the capital of Sarajevo, where we observed a 2.2 cm difference (181.8 cm vs. 184.0 cm) between the major Muslim part of the city and the eastern Serbian part (Istočno Sarajevo). This regional anomaly also occurred in the Muslim enclave of Goražde, and in the Muslim part of Herzegovina between Konjic and Mostar, where we found 2–3 cm shorter averages than in the neighboring Croatian and Serbian regions.
However, the only variable that meaningfully explains Dinaric height as a whole is genetic—the frequencies of Y haplogroup (male lineage) I-M170. Our updated data (Grasgruber et al.—in review) show that this haplogroup correlates strongly positively with male height both within the seven countries of the Dinaric Alps (r =
0.80, p =
0.030) and within 55 countries of Europe and the Near East (r =
0.73, p <
0.001). In contrast, Y haplogroups of Near Eastern origin (E-M96, G-M201, J-M304) predict short stature in the Dinaric Alps (r
= −0.70, p
= 0.079), and this is especially true for J-M304 (r
= −0.88, p
= 0.008), which is also the strongest correlate of shortness in Europe and the Near East (r
= −0.86, p
< 0.001). In accordance with these findings, the inclusion of I-M170 and the three Near Eastern Y haplogroups dramatically improves the best regression model of male height in Europe: from adjusted R2
= 0.426 to 0.721 [18
]. Other European Y haplogroups have a more restricted geographical distribution but they also show geographical relationships with height (see Figure 5
A,B and [20
Y haplogroup I-M170 generally constitutes a very interesting case because its origin is very old and can be traced as far back as to the Upper Paleolithic Gravettian culture [26
]. I-M170 was also closely associated with the post-glacial expansion of Epigravettian (Late Gravettian) groups from the refugium around the Adriatic Sea and became the predominant Y haplogroup in Mesolithic Europe [27
]. At present, it reaches a global frequency peak in Herzegovina (70.9%) [28
], where it is overwhelmingly represented by the Balkan sub-branch I2a1a-P37.2 and particularly by downstream mutations of I2a1a2-M423 [29
]. This dominance of I2a1a2-M423 would result from a relatively recent (possibly even post-Neolithic) founder effect [28
]. Still, the history of I-M170 in the Dinaric Alps remains enigmatic and due to the scarcity of well-preserved skeletal material, it will be difficult to capture its evolution over time. In fact, the website of Ancient Human DNA [32
] does not register a single prehistoric sample from Bosnia and Herzegovina and Montenegro [32
]. To our knowledge, the oldest occurrence of I-M170 (I2a1a2-M423) in the Dinaric area was documented in the Bezdanjača Cave (Lika-Senj County, Adriatic Croatia) and was indirectly dated to ~1200 cal. BC [33
]. A more recent analysis of Bronze Age individuals from this cave reportedly found another case of I2a1a2-M423 (I. Lazaridis, personal communication). However, Utevska [34
] doubted the local origin of I2a1a-P37.2 in theWestern Balkans and hypothesized that it had expanded ~3000 years ago from the area east of the Carpathians.
In any case, the genetic explanation would fit the distribution of height in the Dinaric Alps because the area with the tallest statures does not follow the border of the limestone bedrock (as hypothesized by Coon), but rather, it lies deeper behind the main mountain range that served as a barrier to the genetic flow from the European mainland. Environmental explanations fail especially in Montenegro, where people in the richer coastal regions around the capital of Podgorica are much shorter than those in the remote mountain areas of Kolašin and Šavnik lying mostly on flysch sediments [35
]. Only Albania is an exception because here, we observe a completely different geographical trend between the short-statured mountainous hinterland (especially Elbasan and Kukës) and the taller Adriatic coast. Age-stratified regional data from the Albania DHS 2017–2018 (Figure 6
) indicate that the main cause of this anomaly must be sought in Albania’s long-term economic stagnation and isolation. In other words, height in many Albanian regions has long been conserved at the medieval level (~170 cm in men) due to deep economic underdevelopment, and it is natural that the improvement of living conditions during the last three decades was accompanied by a very fast positive trend in the most affluent coastal counties. In contrast, Elbasan and Kukës remained the poorest counties in Albania [36
4.2. Current Genomic Evidence
Given that Y haplogroups are uniparental markers that best serve the purpose of genetic genealogy and do not accurately reflect autosomal ancestry, the genetic explanation of the Dinaric phenomenon requires an analysis of autosomal DNA that would identify specific genetic loci linked to the selection for height. Comparisons of this sort are already available thanks to genome-wide association studies (GWAS) and have shown that height is highly polygenic—depending on the combination of a large number of genes, each of which explains only a very small part of the total variability [37
]. At present, hundreds of potentially height-associated SNPs (single nucleotide polymorphisms) are known, and most of them were reported by the GIANT consortium (Genetic Investigation of Anthropometric Traits). Their combined effect (polygenic height scores) is also used as the predictor of height at the population level. However, the identification of such SNPs is difficult due to the confounding role of environment [39
]. Although the accuracy of the polygenic height scores has been improving at the individual level [40
], their success at the population level is still rather mixed because height-associated SNPs are population-specific and most GWAS have been performed on Europeans [41
]. Another drawback of these studies is that they pay virtually no attention to the Western Balkans. As a result, the number of publicly available individual genomes from the Western Balkans is relatively small and considering that they are not sorted according to regions in the POPRES database [43
], a direct testing of the genetic hypothesis was not possible.
Contemporary genetic studies aimed at the evolution of height in Europe (e.g., [39
]) touched the Western Balkan region at best only superficially and are not very helpful either. Their results generally predict tall statures in ancient populations emerging out of the Epigravettian refugium (genetic cluster Villabruna/WHG) and in the nomadic Eneolithic cultures from the East European steppe (genetic cluster Yamnaya), and attribute low genetic predispositions to Near Eastern agriculturalists. This result agrees with the Y chromosomal picture but curiously, polygenic height scores in the modern Western Balkan populations are only moderate or below average. For example, the preprint by Berg et al. [41
] estimated medium height in present-day Croats and predicted the highest values in Icelanders, Englishmen, and Scots. At the same time, English and Scottish males reach only ~178 cm, despite a long historical lead in industrial development, whereas Croats are almost 3 cm taller. This discrepancy raises legitimate questions about the accuracy and population-specificity of these predictions. A later study by Sohail et al. [39
] used presumably unconfounded SNPs based on British individuals from the UK Biobank, but these markers also produced only low-to-medium polygenic scores in a small POPRES sample from the former Yugoslavia (n
= 44). This finding is even more problematic because height means in the Western Balkans are among the highest that have ever been documented in any human group, and five countries of the former Yugoslavia are among the top 13 tallest in the world. Therefore, it is very difficult to imagine that the local populations would be endowed with below-average genetic predispositions in the European context. An alternative explanation must work with the possibility that polygenic height scores derived from Western Europeans are not usable for the Western Balkans and the genetic variation in height in Europe is underestimated.
4.3. Future Perspectives
The suboptimal socio-economic and nutritional statistics summarized in the present study suggest that the region of the Western Balkans has not yet reached its maximum potential in terms of physical stature. In fact, the projection of correlation lines between male height and the frequencies of I-M170 indicates that well-nourished males in Herzegovina and southern Dalmatia could potentially reach an astonishing average height of ~190 cm [12
]. This value, however seemingly improbable, is already not too far from the urban means that we documented in Makarska (187.6 cm), Imotski (186.2 cm), and Čapljina (185.9 cm). The current development of dietary protein quality appears the most optimistic in the case of Croatia and Montenegro (Figure 4
). Judging from the measurements of recruits in the capital of Podgorica born during the 1960s [45
], the height of Montenegrin men has been increasing at a rate of 1.7 cm/decade. Croatian boys measured during nationwide surveys in 1980–1984 and 2006–2008 grew by 2.9 cm [46
], until their growth stopped (or even reversed) during the economic recession in the late 1990s [47
]. The current positive trend in the dietary protein quality predicts that their height should increase again.
The most interesting situation can be observed in Albania, where the ‘protein index’ has been rising relatively fast and steadily since the early 1990s. Provided that all the lagging Albanian regions reach sufficiently high nutritional and socio-economic standards, we expect that the north-to-south gradient in height, which was once reported by Coon [6
], will emerge again, and the height difference between Albania and Montenegro should also decrease. The future potential of men from northern Albania can probably be best illustrated by the example of contemporary men from Western Kosovo. However, as already mentioned, the drawback of the Muslim diet in Albania and Bosnia and Herzegovina lies in the negligible consumption of pork, whose nutritional value is not easy to replace. Given that Albanians already consume the highest amount of dairy proteins in the world [25
], the values of the ‘protein index’ in Albania are unlikely to rise much higher.
In Slovenia, we observed only negligible increases in height means in annual school surveys (G. Starc, personal communication), which is in accordance with the recent stagnation of the dietary protein quality. Bosnia and Herzegovina and Serbia have recovered from the negative consequences of the Balkan wars (1991–1999), but the consumption of high-quality proteins in these two countries is still rather limited. Nevertheless, 18-year-old boys from the high schools in Tuzla (Bosnia and Herzegovina) grew by 2 cm between 1980 and 2003, despite war hardships [48
], and judging from our data, their height has further increased from 178.8 cm in 2003 to 180.5 cm in 2015. This is a remarkable pace of the secular trend, given the fact that the diet of Bosnian Muslims does not include pork, which is much more consumed in Herzegovina. The quality of nutrition in North Macedonia is similarly low and it is noteworthy that the height of young men was 4 cm lower than in neighboring Serbia. The high share of Albanians in the total population (~25%) could be one of the reasons, but the relatively short mean height documented in Southern and Eastern Serbia, as well as in southern Kosovo, testifies that this tendency towards shorter statures is typical of the whole area and stems from genetic factors.