Our results show that in 1974–2003 MM incidence increased in both genders among the Sardinian population. Cancer Registry data confirmed such finding limited to the northern area of the region and the last decade covered by the database we used. There was no further increase in the subsequent years. The steeper slope of the regression line among the elderly, and the observed association between proportion of elderly among the resident population and MM incidence, suggest that the increasing aging of the Sardinian population and the increasing access of the elderly to specialized medical care over the study period might have contributed to a more frequent diagnosis of the disease along the years. However, the same upward time trend was also observed among the population aged < 65, indicating that other factors might have played a role, possibly interacting with genetic susceptibility to generate the steeper upward time trend we observed in high-risk areas.
Seven communes stand out with the highest probability of their rate exceeding that of the overall population of the region, four located in the northwestern area (Benetutti, Bitti, Oschiri, and Perfugas) and three scattered in the north west and central areas. These are mainly agricultural areas, six with archeological remains indicating their origin dating back to prehistory, and one, Arborea, built in 1928 over a reclaiming land, previously covered by marshes in a malaria endemic area. Most of the population hosted in this new city came in the early years of its foundation from Veneto and Friuli, two regions in north east Italy, and created a flourishing livestock and agricultural crop economy. This town is also known for the large size of its livestock, with two thirds of Sardinian cattle raised in its land. However, we did not find a relationship between presence of large cattle farms and MM incidence over the whole region.
The finding of an excess risk associated with having a first-degree relative affected by MM, particularly among men, and African Americans, supports a role of genetic factors [20]. On the other hand, about 17% of MM heritability seems explained by the known gene variants [21]. Besides, based on results from molecular biology studies, aberrant class switch recombination occurring early in the natural history of MM suggests that environmental factors, such as high doses of ionizing radiation, and occupational exposure in the farming and petrochemical industries, might also contribute to increase risk [22]. The DNA damage resulting from environmental exposures would interact with the class switch recombination process to increase the risk of chromosomal translocations, oncogene deregulation, and malignant transformation [22]. In an analysis of MM risk related to occupation, a moderate increase in risk was reported in association with contact with livestock [23]. Also, gardeners and nursery workers combined, but not other farming jobs, metal processors, female cleaners, and occupations with high level exposure to organic solvents showed a moderate increase in risk [24]. Among lifestyle factors, a moderate alcohol intake might would reportedly convey protection [25].
Although Sardinia is well known for the genetic peculiarities of its population [11], we are not aware of genetic investigations aiming to identify the varying prevalence of gene polymorphisms implicated in MM. The small town of Arborea, with his peculiar modernist architecture, is home for about 4,000 inhabitants, a large fraction of whom preserved their original language, diet, and habits. This population has a different ethnic origin than the rest of the Sardinian population, but it is unclear whether this might be related to the excess incidence of MM therein observed. Nonetheless, the incidence for the resident population (both genders), standardized based on the world population, was 4.6 x 10− 5, 5.05 among men, and 4.2 among women. The corresponding rates in the Veneto region cancer registry were 4.5 for men and 3.7 for women in the IARC CI5 10th Edition [26], and 4.1 for men and 2.7 for women in its 11th Edition [2]. The corresponding figures in the Friuli Cancer Registry were 3.8 for men and 2.9 for women [26], and 3.3 for men and 2.5 for women [2], consistent with what observed in the small city of Arborea in 1974–2003.
The lack of exact correspondence with Cancer Registry data might be due to the fact that the database we used include the cases resulting from final clinical and pathological work up, i.e. those who accessed a dedicated haematology unit that could perform the necessary tests, such as platelets count, light chains in serum and urine, bone marrow aspiration execution and evaluation, bone marrow biopsy, and cytochemistry. These might not include all the incident cases, also because of co-occurrence of MM with other diseases among the elderly, which might result in underdiagnosis. However, this would affect mostly small villages far apart from the specialized haematology units, located in the major urban centres; however, the elevated risk was mainly observed in small towns, which would contrast this hypothesis.
For the same reason, post-diagnosis relocation of the families seems unlikely to have occurred. The exact address at the time of diagnosis was missing for 58/1606 patients (3.6%); it seems also unlikely that this might have affected the overall pattern.
An advantage of our study is that the diagnoses were all reviewed by the same expert haematologist (GB), thus preventing bias due to the varying diagnostic ability by time and geographic area and minimizing and spreading equally the probability of misdiagnosis over the whole region and along the study period. The substantial similarity between the incidence data calculated from our database, and those from Cancer Registry data, although limited to the last decade covered by the clinical database we used and to the northern part of the island, supports the completeness of its records.