[Report on health status of residents in areas with industrial, mining or military sites in Sardinia, Italy].

UNLABELLED: The work described in the present report has been requested by the Secretary of Hygiene, Health and Social Welfare of the Sardinia Region (Italy). It has been carried out by the Regional Epidemiological Observatory within the domain of ESA (Epidemiology Development and Environment) and with the support of the European Union. Eighteen areas (for a total of 73 municipalities) were identified a priori as "potentially polluted", accounting for a population of 917,977 in 2001 census (56% of the total population of Sardinia). The areas have been named after the most important town, as listed below (in brackets rounded 2001 population), major activities in industrial areas are briefly described.

INDUSTRIAL AREAS: Portoscuso (59,000). Processing of aluminium and other metals. Foundry. Power plants. Dismissed mines (mainly coal mining, lead, zinc). Plants for storing and treating special wastes. Italian Law 349/1986 classified this area as "at high risk of environmental crisis" and classified some plants as being "at high technological risk" (Norma Seveso Decree 334/1999). The area is part of the Sulcis National Restoration site. San Gavino (24,000). Industrial and commercial activities. Lead and zinc foundry. Dairy factories. Food industry. Sarroch (52,000). Petrochemical and refinery industry. Power plants. Mining. Incinerator. Plants for storing and treating special wastes. Gas and mineral oil deposits. Ottana (15,000). Chemical industry. Production of plastics and synthetic fibres. Denim production. Porto Torres (168,000). Chemical industry: production of basic chemicals (benzene, toluene, ethylene, propylene and others), polyethylene, elastomers and vinyl chloride. Textile industry. First and second category landfills. Some plants have been classified "at high technological risk" (Norma Seveso Decree 334/1999). The area is a National Restoration site. The town of Sassari is included. Tortolì (23,000). Construction of steel structures for offshore facilities of the oil and gas industry. Paper industry. Tempio Pausania (21,000). Cork production. Stone quarries. Macomer (17,000). Textile industry (velvet). First and second category landfills. Incinerator.

MINING AREAS: Arbus (30,000). Extraction of zinc, lead and silver. Iglesias (39,000). Extraction of zinc, lead and silver.

MILITARY SITES: Teulada (16,000). La Maddalena (11,000). Naval army shipyards. Salto di Quirra (31,000). Mining area.

URBAN AREAS: Cagliari (299,000). Petrochemical plants, port, airport. Nuoro (37,000). Olbia (47,000). Port and airport. Oristano (31,000). Sassari (121,000).

RESULTS: THE COMPARISON SARDINIA-ITALY: In 1997-2001, the age-standardized mortality rate (x1,000 person-years) among males was higher than in Italy (84.4 vs 80.8) while the reverse occurred in females (50.9 vs 52.0). Ill defined causes of death were 1.4% in males and 2.5% in females (vs corresponding estimates of 1.1% and 1.4% in Italy). Compared to Italian national data, regional age-standardized estimates were higher in Sardinia for infectious diseases (23% in males and 12% in females), respiratory diseases (22% and 14%: pneumoconiosis was more than 6 times more frequent in Sardinia than in Italy), diseases of the digestive system (26% and 9%: for liver cirrhosis, the excess was 33% in males and 9% in females; corresponding figures for liver cancer were 13% and 16%), breast cancer in females (5%). On the other hand, regional mortality rates were lower than the national rates for cardiovascular diseases (-1.3% and -7.4% in males and females respectively), all cancers considered as a whole (-9% and -7%) and lung cancer (-5% and -32%). Regional and national death rates for non Hodgkin lymphoma in both sexes and for leukaemia in females were almost identical, whereas the latter rate in males was slightly higher in Sardinia than in Italy (9.4 vs 8.4 x100,000 person-years). Particularly in men, the differences in mortality rates from all causes and from cardiovascular, respiratory diseases and lung cancer among the four traditional Provinces (Cagliari, Nuoro, Oristano and Sassari) were greater than the difference between Sardinia and Italy. Remarkably enough, also death rates from lymphohaemopoietic tumours were more heterogeneous within Sardinia.

RESULTS IN THE INVESTIGATED AREAS: Rates of hospital discharges in Sardinia showed a high variability, which is partly attributable to differences in the availability of both hospital beds and alternative forms of care. This heterogeneity must be taken into account in the interpretation of rates of hospital discharge. These were relatively high in some areas (Cagliari, Iglesias, Portoscuso, Tortolì) and low in others (Olbia, Porto Torres, Sassari). All the reported observed/expected ratios were based on material deprivation adjusted figures. All the estimated statistics were reported with 90% Confidence Interval.

INDUSTRIAL AREAS: In 1997-2001, deaths from respiratory diseases were significantly in excess in males in Portoscuso (obs/exp 205/124.77) and in San Gavino (69/46.77). Deaths from pneumoconiosis were recorded sporadically, with the exception of Portoscuso, where the excess was impressive (obs/exp 112/30.46). SMRs for lung cancer in males ranged between 0.62 in Ottana and 1.22 in San Gavino, with statistically significant departure from expected values in Portoscuso and Sarroch (both with SMR significantly in excess in males: 1.24). In Porto Torres mortality from all causes was in significant excess in both sexes (SMRs 1.04 in males and 1.09 in females), for respiratory diseases (1.08 and 1.28), for diseases of the digestive system (1.13 and 1.21), for all cancers (1.04 and 1.09). Liver cancer deaths were also in excess in both sexes (SMRs 1.18 and 1.21). The latter finding is confirmed by incidence rates from the local cancer registry. Among industrial areas, Porto Torres was also the one with a stronger evidence of an excess of deaths from lymphohaemopoietic cancer in males (obs/exp 99/83.60) and females (73/68.20).

MINING AREAS: These areas are characterized by statistically significant excesses of mortality in males, largely caused by non neoplastic respiratory conditions (obs/exp 119/86.41 in Iglesias and 156/62.55 in Arbus). In recent years, deaths from pneumoconiosis averaged 20 per year in Arbus and 10 per year in Iglesias. Lung cancer in males was also significantly in excess in both areas (obs/exp 72/56.38 in Arbus and 108/72.14 in Iglesias). There is a time trend (1981-2001) towards a decrease of mortality from respiratory conditions, which nevertheless remains largely in excess over the regional average also in the most recent period.

MILITARY AREAS: Statistically significant excesses of deaths and hospital discharges for non Hodgkin lymphoma were detected in La Maddalena (mortality, 1981-2001, in males 17 observed cases vs 6.13 expected, in females 8/5.64). In Salto di Quirra in 1997-2001 deaths from myeloma (in males 5/2.3) and leukaemias were increased in both sexes (total obs/exp 20/13.3, statistically non significant).

URBAN AREAS: Urban areas in Sardinia are relatively well developed with high values of socioeconomic indicators. The health profile in Cagliari and Sassari is typical of towns of the Western world. In Cagliari there is a higher mortality for colorectal, breast, cervical and lung cancer.

CONCLUSIONS: Environmental (non occupational) pollution might explain some of the observed excesses of disease in the investigated industrial areas of Sardinia, particularly in women, less likely to be exposed to hazards in the work environment, whereas in the mining areas studied the disease pattern suggests a major role of occupational exposures. On the other hand, the causal links between disease occurrence and exposures in the screened military areas remain uncertain. The disease patterns in the cities of Sardinia are likely to be associated with lifestyle and urban pollution. Historically, southern Italian Regions have been characterized by an advantage over the rest of the country in terms of health, but during the last decade such advantage tended to vanish. Sardinia confirms this secular trend. However in the most recent years studied, overall age-standardized mortality rate in Sardinian females still remains lower than Italian average, but this is not the case for males any more. Differences in the health profile between residents in different areas of Sardinia have been found to be far greater than the difference between Sardinia as a whole and Italy. A major contribution to intraregional differences is given by the 18 investigated areas where excesses were registered for: respiratory diseases (including cancer) in the industrial areas of Portoscuso, Sarroch and Porto Torres, and in the mining areas; diseases of the digestive tract, liver cancer and lymphohaemopoietic cancer in the area of Porto Torres; cancer of the lymphohaemopoietic system in some military areas; cancers of the colon and rectum, lung, breast and uterus in some of the major cities of the Region.