• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
Appl Environ Microbiol. Oct 2005; 71(10): 5805–5813.
PMCID: PMC1265926

Legionella Contamination in Hot Water of Italian Hotels


A cross-sectional multicenter survey of Italian hotels was conducted to investigate Legionella spp. contamination of hot water. Chemical parameters (hardness, free chlorine concentration, and trace element concentrations), water systems, and building characteristics were evaluated to study risk factors for colonization. The hot water systems of Italian hotels were strongly colonized by Legionella; 75% of the buildings examined and 60% of the water samples were contaminated, mainly at levels of ≥103 CFU liter−1, and Legionella pneumophila was the most frequently isolated species (87%). L. pneumophila serogroup 1 was isolated from 45.8% of the contaminated sites and from 32.5% of the hotels examined. When a multivariate logistic model was used, only hotel age was associated with contamination, but the risk factors differed depending on the contaminating species and serogroup. Soft water with higher chlorine levels and higher temperatures were associated with L. pneumophila serogroup 1 colonization, whereas the opposite was observed for serogroups 2 to 14. In conclusion, Italian hotels, particularly those located in old buildings, represent a major source of risk for Legionnaires' disease due to the high frequency of Legionella contamination, high germ concentration, and major L. pneumophila serogroup 1 colonization. The possible role of chlorine in favoring the survival of Legionella species is discussed.

Bacteria of the genus Legionella normally inhabit freshwater or wet soil, but the major reservoirs are man-made aquatic environments, particularly warm water systems (13, 20, 22, 41, 51). In freshwater, legionellae survive as intracellular parasites of free-living protozoans, which are their natural hosts (1, 9, 25, 49). In building water systems, legionellae are also associated with biofilms that provide shelter and nutrients and support their survival and multiplication even outside a host cell (32, 38, 39). These peculiar abilities to grow are responsible for the frequent contamination of artificial water systems, as well as difficulties in eradicating legionellae from contaminated water systems and the lack of biocide efficacy (8, 11, 21, 24, 31, 34). The accumulation of bacteria on pipeline surfaces and biofilm formation are influenced by many factors, such as an inadequate flow rate or stagnation of the water, surface materials and roughness, the concentration and quality of nutrients and disinfectants, temperature, and the hydraulics of the system (5, 46, 61). A better understanding of the risk factors for Legionella colonization in artificial water systems could help in the development of control strategies for prevention of legionellosis. For instance, Legionella contamination of domestic hot water was found to be associated with a centralized system, a greater distance from the heat source, and an older water plant, whereas copper in the water had a protective effect (10).

Different distributions of Legionella species and serogroups have been observed depending on the water type. Significant amounts of Legionella spp. were detected in groundwater and potable water, whereas Legionella pneumophila was found to be adapted to warm water systems, in which this organism multiplies most efficiently (18, 36, 37). The relative distributions of different Legionella species may have a substantial impact on public health as the majority of Legionnaires' disease is caused by L. pneumophila, and within this species the most prevalent serogroup is serogroup 1 (26).

Tourism is a major industry in many European countries and is sensitive to health threats. In 1987, a European surveillance scheme for travel-associated Legionnaires' disease was introduced to detect cases (19, 28) and control and prevent Legionella infections associated with tourism in both indigenous and foreign persons (15, 44). In 2002, the highest number of reported cases in tourists was associated with travel in Italy (132 of 676 cases), where 35 clusters of travel-associated Legionnaires' disease were identified from July 2002 to October 2003, mainly involving hotels and residences located in 14 Italian regions (47).

Due to the high prevalence of cases associated with travel in Italy, we decided to estimate Legionella spp. contamination in a representative number of Italian hotels. A multicenter investigation was carried out to identify Legionella spp. in hot water samples and to study risk factors associated with microbial contamination in Italian hotels. The aims of this study were to quantify the frequency and severity of contamination, to evaluate the characteristics of buildings and water distribution systems associated with Legionella colonization and their relative roles in favoring or protecting against contamination, and to suggest preventive measures to limit colonization. Other aquatic bacteria may compete with Legionella for growth in the aquatic environment (23, 55), so we also evaluated Pseudomonas spp. colonization and total microbial counts at 36°C and 22°C. Finally, risk factors for the presence of different Legionella species and serogroups with special reference to L. pneumophila serogroup 1 were examined due to the potential impact on public health.


Sample collection.

From September 2003 through July 2004, 119 water samples were collected from 40 hotels located in five towns (Milan, Modena, Bologna, Naples, and Bari) that are representative of different Italian regions (northern, central, and southern Italy). The hotels were selected on the basis of the water distribution systems in the cities, the characteristics of the buildings, and hotel cooperation. Three to five water samples were taken from each hotel, depending on the number of rooms. Hot water samples were collected from bathroom outlets (showerheads or bath taps) in three sterile glass bottles (1 liter each) without flaming after a brief flow time to eliminate cold water inside the tap or the flexible pipe of the shower. In order to neutralize residual free chlorine, sodium thiosulfate was added to sterile bottles for bacteriological analysis, whereas acid-preserved glass bottles were used for chemical determinations. The methods used for sample processing and the storage conditions have been described elsewhere (10).

Microbiological analysis.

Legionellae were isolated by using the procedure described in the Italian guidelines (4). Culture and identification of L. pneumophila were carried out by using the ISO 11731 method (3).

Briefly, 2 liters of a water sample was concentrated by membrane filtration (0.2-μm-pore-size polyamide filter; Millipore, Billerica, MA). The filter membrane was resuspended in 10 ml of the original sample water (15). Two 0.1-ml aliquots of the original and concentrated specimens (heat treated and untreated, diluted 1:10 and undiluted) were spread on duplicate plates containing MWY selective medium (Oxoid Ltd., Basingstoke, Hampshire, United Kingdom). The plates were incubated at 36°C in a humidified environment with 2.5% CO2 at least for 10 days and examined beginning on day 5 with a dissecting microscope. All colonies on plates containing ≤10 colonies and 10 to 20 random colonies from other plates were subcultured on BCYE (with cysteine) and CYE (cysteine-free) media (Oxoid) for ≥2 days. Colonies grown on BCYE were subsequently identified by an agglutination test (Legionella latex test; Oxoid). This test allows separate identification of L. pneumophila serogroup 1 and serogroups 2 to 14 and detection of seven species of non-L. pneumophila legionellae (polyvalent) which have been implicated in human disease, L. longbeachae, L. bozemanii 1 and 2, L. dumoffii, L. gormanii, L. jordanis, L. micdadei, and L. anisa. For selected L. pneumophila isolates, serotyping was also performed by agglutination with commercial specific monoclonal antibodies (Pro-lab Diagnostics, Canada). The results were expressed as CFU liter−1, and the detection limit of the procedure was 25 CFU liter−1 (mean of two plates).

Accuracy and recovery performance were periodically established by using water samples supplemented with different Legionella species and concentrations distributed in a quality program organized by the National Health Institute, as well as commercial certified specimens (Oxoid).

The total microbial counts at 36°C and 22°C were determined twice by the pour plate method on plate count agar (Oxoid). The plates were incubated at 36°C for 48 h or at 22°C for 72 h.

In order to isolate Pseudomonas spp., 100-ml and 10-ml water samples were filtered through 0.45-μm-pore-size membranes (Millipore, Billerica, MA). If the number of bacteria was high, suitable dilutions were prepared. The membranes were placed on Pseudomonas CFC agar (Oxoid) and incubated at 30°C for 48 h. Each type of oxidase-positive colony was counted.

Chromosomal PFGE analysis.

L. pneumophila environmental isolates were subjected to pulsed-field gel electrophoresis (PFGE) using a contour-clamped homogeneous electric field system (CHEF MAPPER; Bio-Rad, United States). Patterns were obtained by SfiI (Promega, United States) digestion using self-made reagents produced by a slight modification of a previously described protocol (50). The running conditions were 200 V for 27 h at 14°C with switch times of 1 s (initial) and 35 s (final). After ethidium bromide staining, the gels were photographed with a UV light source. According to Tenover criteria, a pattern designation was assigned if the electrophoretic profile differed by more than three bands (52).

Macrorestriction patterns were analyzed using the Bionumerics software (version 3.0; Applied Maths, Kortrijk, Belgium) to generate a similarity dendrogram. A similarity matrix was created by using the band-based Dice similarity coefficient, and the unweighted pair group method was used to cluster strains (16).

Physical and chemical analyses.

Water temperature and the residual free chlorine content (colorimetric DPD method; Microquant; Merck, Darmstadt, Germany) were determined at the time of sample collection. Standard techniques were used to measure oxidizability (2) and water hardness (method 2040; IRSA-CNR, Rome, Italy). Concentrations of calcium, magnesium, iron, manganese, copper, and zinc were determined with a flame atomic absorption spectrophotometer (model 5000; Perkin-Elmer, Wellesley, MA) by using acidified samples (1% HNO3) concentrated by boiling (12).

Risk factors.

A detailed standardized questionnaire was developed to evaluate risk factors possibly associated with colonization. The first part of this questionnaire collected information on the building, including the number of floors, the number of hotel rooms, the number of bathrooms, the building age, the type of water supply, and the disinfection systems used. The second part collected information on the heating system (central or independent, electric or gas heater), the distance of the sample site from the water distribution point, the existence of a tank and its volume, the age of the system, the frequency of service, and the existence and characteristics of a softening system and a water recycling system. The normal water temperature was also recorded.

Statistical analysis.

All statistical calculations were made with SPSS/pc (SPSS Inc., Chicago, IL). Logarithmic transformations were used in statistical analyses to normalize the nonnormal distributions, and the results were expressed as geometric means. The bacteriological data were converted into log10 (x + 1). When possible, variables were categorized into dichotomous variables. The results were analyzed by correlation analysis, by a t test, by a one-way analysis of variance, and by a χ2 test. Odds ratios (OR) and 95% confidence intervals (CI) were calculated to assess categorical risk variables associated with microbial contamination. Variables that were significant as determined by the univariate analysis were entered into a multiple logistic regression model. By using conditional logistic regression models, independent predictors of colonization were established. Variables were retained in the model if the likelihood ratio test was significant (P < 0.05).


Table Table11 shows the general characteristics of the water examined in terms of water supply and distribution system. The typical features of Italian hotels include limited numbers of rooms and floors and old age of the building (up to 800 years). The water supply is mainly groundwater, and all water is disinfected by chlorine in different chemical forms irrespective of the water origin. A clear preponderance of metal plumbing, gas water heaters, and water recirculation systems was recorded.

Main characteristics of the building and water supplies in the 40 hotels examined

A total of 30 hotels (75.0%) and 72 water samples (60.5%) were contaminated by at least one Legionella species, and there were no significant differences for different geographic areas (Table (Table2).2). For 18 hotels, all 44 water samples were positive, while for the other hotels some sites were positive and some were negative (21 of 38 sites were positive). Pseudomonas spp. were isolated from 18 hotels (45.0%) and 33 (27.7%) water samples; no sample was positive in Bari, whereas 63.6% of the samples in Modena were positive (P < 0.001).

Distribution of Legionella and Pseudomonas isolates in the areas examineda

L. pneumophila was the most frequently isolated species (87% of the isolates). Among the positive hotels, nine were colonized by L. pneumophila serogroup 1 (20 positive samples and 5 negative samples) and 15 were colonized by L. pneumophila serogroups 2 to 14 (31 positive samples and 10 negative samples). Mixed Legionella cultures were obtained for the remaining six hotels. In two hotels three sites were colonized by L. pneumophila (all serogroups), and two sites were colonized by either serogroup 1 or serogroups 2 to 14. In two hotels five sites were colonized by serogroup 1 plus unknown species, and four sites were colonized by serogroup 1. In two hotels six sites were colonized by serogroups 2 to 14 plus unknown species, one site was colonized by serogroups 2 to 14, and one site was negative. As shown in Table Table3,3, the mean number of legionellae for positive samples was 1.9 × 103 CFU liter−1 (geometric mean), and 19.4% of these samples contained ≥104 CFU liter−1. L. pneumophila serogroup 1 was isolated from 33 points, mainly at levels of ≥103 CFU liter−1. L. pneumophila serogroup 1 correlated negatively with L. pneumophila serogroups 2 to 14 (r = −0.309, P < 0.001) and positively with the unknown species (r = 0.205, P < 0.05). The geometric mean level of Pseudomonas spp. was 176 CFU liter−1, and the values ranged from 2 to 105 CFU liter−1.

Characteristics of Legionella contamination in the hot water samples examined

Twenty-two L. pneumophila isolates collected from 15 different hotels in three cities (Naples, Modena, and Bologna) were characterized further, divided into four different serogroups (serogroups 1, 3, 6, and 9), and subsequently typed by the PFGE method. The similarity dendrogram derived from SfiI macrorestriction analysis is shown in Fig. Fig.1.1. The patterns obtained contained 8 to 12 fragments. Isolates with patterns that differed by no more than three bands had similarity coefficients of ≥0.80, and this cutoff was used to define pulsotypes. Sixteen different pulsotypes were identified, including 11 pulsotypes that contained a single isolate. No genetic similarities were found among isolates collected in different hotels. With regard to the correlation between serogroup and genotype, all the serogroups appeared to be particularly heterogeneous, with representatives found in multiple pulsotypes. It is interesting that pulsotypes 6 and 11 were represented by two isolates belonging to different serogroups (serogroups 3 and 6 and serogroups 6 and 9, respectively).

FIG. 1.
Dendrogram showing similarities between 22 environmental isolates of L. pneumophila based on PFGE results. Isolates with Dice similarity coefficients of ≥0.80 are considered identical. Solid boxes indicate the serogroups (sg) for isolates. The ...

No difference in building and water characteristics was observed between colonized and Legionella-free hotels, except for a lower zinc concentration in the water (geometric means, 158.2 and 498.5 μg liter−1[P < 0.001]). Similarly, colonized points did not differ from uncontaminated water, except for zinc concentration and total hardness (geometric means, 20.8 and 31.9°F [P < 0.001]) due to lower levels of both calcium and magnesium. Pseudomonas-positive hotels were older than Pseudomonas-free structures (53.1 versus 14.6 years [P < 0.001]), and the free chlorine concentrations were lower (10.1 versus 69.0 μg liter−1 [P < 0.05]). The total microbial counts for Pseudomonas-positive sites were also higher than the total microbial counts for Pseudomonas-free structures at both 36 and 22°C (596 versus 91 CFU ml−1 [P < 0.001] and 142 versus 35 CFU ml−1 [P < 0.05, respectively]).

In contrast, there were differences according to Legionella species and serogroups (Table (Table4).4). The zinc, calcium, and magnesium concentrations and the total hardness were significantly lower in L. pneumophila serogroup 1-colonized samples than in serogroup 2 to 14-colonized samples. The copper concentrations and oxidizability were lower in serogroup 2 to 14-contaminated samples than in uncontaminated samples. Trends toward greater hotel age for serogroup 2 to 14-colonized water and higher free chlorine concentrations in serogroup 1-contaminated water were also observed, but the differences did not reach statistical significance. The manganese concentration was significantly higher in serogroup 1-contaminated water than in serogroup 2 to 14-contaminated water, as determined by the Student t test (3.46 and 2.03 μg liter−1, respectively [P < 0.05]). When a nonparametric test (Kruskal-Wallis) was applied, the results did not change.

Variables (geometric means) associated with Legionella contamination as determined by analysis of variance

The risk factors for microbial contamination were analyzed by means of univariate logistic, regression on dichotomous variables (Table (Table5).5). A hotel age of >20 years was positively associated with a risk for Legionella contamination (OR = 2.70, P < 0.01), whereas an operating temperature of >60°C significantly inhibited Legionella. Furthermore, a sample temperature of >55°C, a total hardness of >20°F, an oxidizability of >3.0 mg liter−1 of O2, a copper concentration of >50 μg liter−1, and a zinc concentration of >200 μg liter−1 were protective. The results were similar when we introduced into the model positive/negative hotels instead of positive/negative sites, but the statistical significance disappeared for copper concentration and hardness (data not shown).

Association of water and building parameters with Legionella and Pseudomonas contamination (univariate regression analysis)

A higher free chlorine concentration and lower hardness were positively associated with L. pneumophila serogroup 1 colonization, whereas the opposite trend was observed in samples colonized by L. pneumophila serogroups 2 to 14 (Table (Table5).5). The results did not substantially change when we omitted samples from which serogroup 1 and/or serogroups 2 to 14 were isolated in combination with other species or serogroups (data not shown). The 11 unknown Legionella strains were negatively associated with total hardness (OR = 0.021, 95% CI = 0.003 to 0.18, P < 0001) and with zinc concentration (OR = 0.29, 95% CI = 0.08 to 1.03, P < 0.05) and were never found when the system temperature was >60°C and the sample temperature was >55°C. The only factor negatively associated with Pseudomonas contamination was the free chlorine level.

When the same parameters (excluding operating temperature) were introduced into a multivariate conditional logistic regression model (Table (Table6),6), a hotel age of >20 years remained independently associated with an increased risk for Legionella spp. colonization, whereas water oxidizability of >3 mg liter−1 O2 and total hardness of >20°F were protective. Chlorine levels of >100 μg liter−1 and hardness of ≤20°F were predictive of L. pneumophila serogroup 1 colonization, whereas greater hotel age and greater hardness were predictive of L. pneumophila serogroup 2 to 14 colonization. When we introduced positive hotels versus negative hotels into the model as a dependent variable, only hotel age remained significantly associated with an increased risk of colonization (OR = 8.00, 95% CI = 1.42 to 44.92, P < 0.05). Pseudomonas contamination remained negatively associated with the free chlorine concentration (OR = 0.26, 95% CI = 0.09 to 0.74, P < 0.05).

Multiple logistic regression of building and water characteristics associated with Legionella contamination


Sporadic cases and outbreaks of travel-associated Legionnaires' disease are common in Italy, which tops the list of European countries for the number of reported cases (45). Hotels are the most frequent sources of cases, probably because large buildings provide a more hospitable environment than small facilities because the more extensive piping network provides a larger surface with variable temperatures and biofilm accumulation (58). Our investigation confirmed that hot water systems of Italian hotels are strongly colonized by Legionella spp., as 75.0% of the buildings examined and 60.5% of the water sites sampled were colonized; Pseudomonas spp. are also common colonizers. Genotyping data obtained for 22 isolates revealed high genetic heterogeneity; all of the buildings were colonized by different strains, and in two hotels in Modena different pulsotypes were recovered at the same time. The frequencies of colonization, however, were similar in the different geographic areas, which included towns located in northern, central, and southern Italy, despite the fact that reports of Legionnaires' disease have occurred mainly in northern Italy (48). This suggests that diagnosis and/or notification may be underestimated in southern Italian regions, although recently in the Puglia region, where Bari is located, three travel-associated clusters were reported (47).

Unfortunately, we could not document the temporal variability of colonization by repeatedly sampling the same sites in each hotel. In studies in which sampling was repeated in each building over a period of 1 year, the workers found differences between seasons in terms of counts but not in terms of serogroup distribution, which remained fairly constant over time, at least in Italian private healthcare facilities (34). Indeed, in our study contaminating concentrations were not associated with the possible risk factors, suggesting that levels may be less important than types. Nevertheless, in our hotels the concentrations were noteworthy; 62.5% of the concentrations exceeded 103 CFU liter−1, a threshold for considering preventive measures according to Italian guidelines. During environmental investigations of water at Italian locations where clusters of travel-associated Legionnaires' disease occurred, the Legionella concentration was >104 CFU liter−1 in 52% of the buildings examined (47). In our study, the concentrations in 35% of the hotels exceeded this level, and more than 90% of the isolates belonged to L. pneumophila, suggesting that the risk for cases in contaminated Italian hotels is worrisome.

If we consider that cultural methods generally underestimate the presence of Legionella, the real level of hotel contamination could be greater. Recently, a real-time LightCycler PCR assay has been proposed for investigation of Legionella contamination in potable water systems (35, 57). Compared to cultures, this method is much more rapid, has higher sensitivity, and detects even nonculturable legionellae. However, possible inclusion of nonviable cells and nonlegionella DNA and shielding of legionellae sequestered within protozoans from the PCR are limitations of the PCR assay, which at present does not detect non-L. pneumophila species. Cultural methods remain the “gold standard” for associating environmental isolates with clinical isolates and for estimating the virulence of the Legionella spp. detected.

Another aspect which deserves special attention is the high frequency of L. pneumophila serogroup 1, which was isolated from 45.8% of the contaminated sites and from 32.5% of the hotels examined, with differences according to geographic area. Furthermore, the levels of serogroup 1 were higher than the levels of the other isolates. In our previous study of domestic contamination in the same cities, L. pneumophila serogroup 1 was found only in 18.2% of positive hot water samples, and residents living in Legionella-positive homes did not have a significantly higher risk of disease than residents living in Legionella-free buildings (10).

Worldwide, L. pneumophila serogroup 1 is the most common agent of Legionnaires' disease, accounting for about 80 to 90% of the reported cases (6, 26, 59) and approximately 70% of the European travel-associated cases (44). In contrast, L. pneumophila serogroups 2 to 14, although accounting for more than 50% of the isolates obtained from man-made aquatic systems, account for only 15 to 20% of community cases. The discrepancy between environmental isolates and clinical expression of disease has recently been observed by Doleans et al. (17), who suggested that there are differences in virulence rather than greater abundance in water distribution systems.

In Italy, as in other European countries, approximately 20% of the total cases and 25% of the community cases of Legionnaires' disease occur in travelers, and the main sources are hotels (30, 43). This proportion suggests that the risk for travelers is particularly high, as people do not normally spend 20 to 25% of their time traveling. The high frequency of L. pneumophila serogroup 1 in the hotels examined could at least partially explain the elevated number of travel-associated cases in Italy. It would be interesting to place serogroup 1 strains into monoclonal antibody subgroups, as travel-associated cases were found to be caused predominantly by monoclonal antibody 3/1-positive L. pneumophila serogroup 1 (26). In any case, there should be a major effort to avoid or reduce L. pneumophila serogroup 1 colonization in accommodation sites. In addition, 8 of 14 L. pneumophila serogroup 2 to 14 isolates belonged to serogroup 6, a serogroup recently involved in outbreaks of Legionnaires' disease and Pontiac fever in the United States (7).

In our study of the risk factors for Legionella contamination, we did not observe major differences between colonized and Legionella-free hotels, at least for the parameters measured, and hotel age was the only risk factor which remained positively associated with contamination after a multiple logistic regression was applied. In Italy, the existence of very old buildings is a further reason explaining the high frequency of travel-associated cases. Conversely, we could not find a clear association with heating system age, like that documented at the domestic level, most probably due to inaccurate reporting of this information. Hotel contamination was not associated with pipe material, water origin, type of chlorine used for water disinfection, type of heater, tank capacity, number of floors and/or rooms, or distance between the hot water reservoir and the outlet sites sampled. The role of copper in the water in reducing the risk of Legionella contamination was less evident than the role found at the domestic level, probably due to the limited number of water samples with copper concentrations of >50 μg liter−1 (11 samples). The zinc concentration was particularly high, probably due to plumbing corrosion, but we did not find any influence of zinc on Legionella contamination after applying a multivariate analysis. All these data confirm that Legionella colonization in complex water distribution systems is a deep-rooted phenomenon and that the organisms are able to persist and increase with time independent of the various water and system characteristics and of the cleanliness and regular preventive maintenance measures (54).

When the role of water temperature was examined, protection against Legionella colonization was observed only when the system temperature was >60°C and, to a lesser extent, when the outlet temperature was >55°C. L. pneumophila was not isolated from water samples at temperatures of >55°C in Finnish apartment buildings (60). The problem is interesting as the majority of countries set limits for hot water temperature at the storage site of around 50°C (in Italy, 48 ± 5°C is mandatory temperature for public buildings) to avoid scalding (27, 29). To control Legionella, hotels should set the storage temperature at 60°C and the delivery temperature at 55°C, but this apparently simple solution has two drawbacks: the increased cost of heating and, even more relevant from a public health point of view, the need to install appropriate water mixers at outlets to protect people from hot water burns.

Interestingly, the temperature effect was less relevant and statistically not significant for L. pneumophila serogroup 1 contamination, suggesting that serogroup 1 is more resistant to higher temperatures. Furthermore, L. pneumophila serogroup 1 was not influenced by hotel age or oxidizability, and the protective effect of copper was limited and not significant.

These are not the only differences that characterize L. pneumophila serogroup 1 compared with the other legionellae. L. pneumophila serogroup 1 appeared able to survive when the free chlorine concentration was higher (>100 μg liter−1), conditions which inhibited both L. pneumophila serogroups 2 to 14 and Pseudomonas spp. In our study, the free chlorine concentration in water in Bari was higher than that in the other cities examined (geometric mean, 207 μg liter−1; range, 100 to 500 μg liter−1), and the hotels in Bari had the highest frequency of L. pneumophila serogroup 1 isolates (45% of samples) and the lowest levels of Pseudomonas contamination. In contrast, in Modena, where the free chlorine concentration was low (mean, 0.76 μg liter−1; range, 0 to 100 μg liter−1), only four samples (18.1%) were contaminated by L. pneumophila serogroup 1, but Pseudomonas was found particularly frequently (Table (Table2).2). This peculiar situation was also observed at the domestic level, where L. pneumophila serogroup 1-contaminated water had a free chlorine level of 78.0 μg liter−1, compared with 2.6 μg liter1 in water contaminated by serogroups 2 to 14 (10). Our experience with water decontamination in a large hospital showed that shock chlorination with sodium chlorine substantially reduced the number of legionellae immediately after treatment, but the proportion of samples contaminated by L. pneumophila serogroup 1 increased from 2 to 30% (unpublished data).

Many microorganisms of aquatic origin, such as Pseudomonas, produce bacteriocins that may actually be inhibitory to Legionella spp. (9). Thus, chlorine levels that eliminate these microorganisms could increase the population of indigenous legionellae. Furthermore, Legionella spp. are more resistant to chlorine than other bacteria because they can enter a viable but nonculturable state, as observed using PCR methods (42), can be protected by amoebae, and/or can survive in pipe biofilms (23, 38). In an experimental study of microbial changes during and after application of various disinfection treatments, amoebae resisted all the treatments and probably acted as reservoirs for L. pneumophila, allowing quick recolonization of the system once the treatments were interrupted (53). As multiplication in protozoans can differ depending on the Legionella species (21, 40), it is conceivable that L. pneumophila serogroup 1 is more able than other serogroups to survive within protozoans or biofilms under stressful environmental conditions, such as high temperature and high chlorine levels. These observations should be noted by people responsible for water disinfection at the municipal and/or building level. Further studies are needed to establish the cutoff value for the residual free chlorine concentration not associated with survival of virulent legionellae, such as serogroup 1.

We found that total hardness protected L. pneumophila serogroup 1 and unknown species. As hotels frequently use water softening systems, the high levels of L. pneumophila serogroup 1 colonization of hotel water may also depend on this technical device. Hardness is generally considered a risk factor for Legionella colonization, and in our investigation this was confirmed for L. pneumophila serogroups 2 to 14. A negative correlation with hardness and chemical oxygen demand has been described in other studies, but Legionella species were not examined (33). As a lower number of bacteria were found in dental unit water systems using soft water than in dental unit water systems using hard water (56), we suggest that less microbial competition or modification of biofilm formation and/or composition in soft water might represent a risk factor for L. pneumophila serogroup 1 colonization more than hardness per se. Finally, higher manganese and lower zinc concentrations were typical features of water colonized by either L. pneumophila serogroup 1 or unknown species compared to water colonized by L. pneumophila serogroups 2 to 14 (Table (Table4).4). Different distributions of Legionella species and serogroups according to heater type, water temperature, and chlorine and manganese concentrations were also observed at the domestic level.

The reasons for these trends are currently unknown, but our data suggest that Legionella species and serogroups may have different ecological niches and/or that their ability to survive in man-made water environments varies with the environmental conditions. In particular, L. pneumophila serogroup 1 shows a special ability to colonize “cleaner” water systems and to survive under more stressful conditions, such as higher temperature and chlorine levels, which are not consistent with the survival of other serogroups.

In conclusion, Italian hotels, particularly those located in old buildings, represent a major source of risk for Legionnaires' disease due to the high frequency of Legionella contamination, high germ concentration, and major L. pneumophila serogroup 1 colonization. Our results do not suggest specific new measures to control Legionella contamination except for the protective role of temperatures of >60°C. However, the variability of contaminating species seems to depend on different environmental risk factors, such as chlorine content and hardness, which should be controlled to reduce at least the most virulent L. pneumophila serogroups as the risk of disease may substantially change.


We are sincerely grateful to Guerrino Predieri for assistance in performing trace element analyses and to Alberta Stenico for assistance in construction of the dendrogram.

This work was funded by the Italian Ministry of University and Scientific Research (MIUR) (funds PRIN year 2000 and 2002).


1. Abu Kwaik, Y., L. Y. Gao, B. J. Stone, C. Venkataraman, and O. S. Harb. 1998. Invasion of protozoa by Legionella pneumophila and its role in bacterial ecology and pathogenesis. Appl. Environ. Microbiol. 64:3127-3133. [PMC free article] [PubMed]
2. Anonymous. 1993. Directive ISO 8467 of 3 June 1993. Water quality—determination of permanganate index. International Organization for Standardization, Geneva, Switzerland.
3. Anonymous. 1998. Directive ISO 11731 of 30 April 1998. Water quality—detection and enumeration of Legionella. Congress of European Countries Biologists Association, Abano Terme, Italy.
4. Anonymous. 2000. Guidelines for prevention and control of legionellosis. Superior Institute of Health and Department of Health, Rome, Italy.
5. Atlas, R. M. 1999. Legionella: from environmental habitats to disease, pathology, detection and control. Environ. Microbiol. 4:283-293. [PubMed]
6. Aurell, H., J. Etienne, F. Forey, M. Reyrolle, P. Girardo, P. Farge, B. Decludt, C. Campese, F. Vandenesch, and S. Jarraud. 2003. Legionella pneumophila serogroup 1 strain Paris: endemic distribution throughout France. J. Clin. Microbiol. 41:3320-3322. [PMC free article] [PubMed]
7. Benin, A. L., R. F. Benson, K. E. Arnold, A. E. Fiore, P. G. Cook, L. K. Williams, B. Fields, and R. E. Besser. 2002. An outbreak of travel-associated Legionnaires disease and Pontiac fever: the need for enhanced surveillance of travel-associated legionellosis in the United States. J. Infect. Dis. 185: 237-243. [PubMed]
8. Borella, P., A. Bargellini, S. Pergolizzi, G. Aggazzotti, C. Curti, P. Nizzero, G. Stancanelli, R. Vaiani, G. Gesu, and R. Mazzuconi. 1999. Surveillance of legionellosis within a hospital in northern Italy: May 1998 to September 1999. Eur. Surveill. 4:118-120. [PubMed]
9. Borella, P., E. Guerrieri, I. Marchesi, M. Bondi, and P. Messi. 2005. Water ecology of Legionella and protozoan: environmental and public health perspectives. Biotechnol. Annu. Rev. 11:355-380. [PubMed]
10. Borella, P., M. T. Montagna, V. Romano-Spica, S. Stampi, G. Stancanelli, M. Triassi, R. Neglia, I. Marchesi, G. Fantuzzi, D. Tatò, C. Napoli, G. Quaranta, P. Laurenti, E. Leoni, G. De Luca, C. Ossi, M. Moro, and G. Ribera D'Alcalà. 2004. Legionella infection risk from domestic hot water. Emerg. Infect. Dis. 10:457-464. [PMC free article] [PubMed]
11. Borella, P., M. T. Montagna, V. Romano-Spica, S. Stampi, G. Stancanelli, M. Triassi, A. Bargellini, P. Giacobazzi, F. Vercilli, S. Scaltriti, I. Marchesi, C. Napoli, D. Tatò, G. Spilotros, N. Paglionico, G. Quaranta, M. Branca, M. Tumbarello, P. Laurenti, U. Moscato, E. Capoluongo, G. De Luca, P. P. Legnani, E. Leoni, R. Sacchetti, F. Zanetti, M. Moro, C. Ossi, L. Lopalco, R. Santarpia, V. Conturso, G. Ribera D'Alcalà, and S. Montegrosso. 2003. Environmental diffusion of Legionella spp. and legionellosis frequency among patients with pneumonia: preliminary results of a multicentric Italian survey. Ann. Ig. 15:493-503. [PubMed]
12. Borella, P., M. T. Montagna, V. Romano-Spica, S. Stampi, G. Stancanelli, M. Triassi, I. Marchesi, A. Bargellini, R. Neglia, N. Paglionico, G. Spilotros, U. Moscato, G. Casati, P. P. Legnani, R. Sacchetti, C. Ossi, M. Moro, and G. Ribera. 2003. Relationship between mineral content of domestic hot water and microbial contamination. J. Trace Elem. Med. Biol. 17:37-43. [PubMed]
13. Colbourne, J. S., P. J. Dennis, R. M. Trew, G. Berry, and G. Vesey. 1988. Legionella and public water supplies. Water Sci. Technol. 20:5-10.
14. De Luca, G., S. Stampi, L. Lezzi, and F. Zanetti. 1999. Effect of heat and acid decontamination treatments on the recovery of Legionella pneumophila from drinking water using two selective media. New Microbiol. 22:203-208. [PubMed]
15. Decludt, B., C. Campese, M. Lacoste, D. Che, S. Jarraud, and J. Etienne. 2004. Clusters of travel associated Legionnaires' disease in France, September 2001-August 2003. Eur. Surveill. 9:11-13. [PubMed]
16. Dice, L. R. 1945. Measures of the amount of ecological association between species. J. Ecol. 26:297-302.
17. Doleans, A., H. Aurell, M. Reyrolle, G. Lina, J. Freney, F. Vandenesch, J. Etienne, and S. Jarraud. 2004. Clinical and environmental distributions of Legionella strains in France are different. J. Clin. Microbiol. 42:458-460. [PMC free article] [PubMed]
18. Emtiazi, F., T. Shwartz, S. M. Marten, P. Krolla-Sidenstein, and U. Obst. 2004. Investigation of natural biofilms formed during the production of drinking water from surface water embankment filtration. Water Res. 38:1197-1206. [PubMed]
19. EWGLI. 2002. European guidelines for control and prevention of travel associated Legionnaires' disease. Part 2. Definitions and procedures for reporting and responding to cases of travel associated Legionnaires' disease, p. 15-20. PHLS, London, United Kingdom.
20. Fields, B. S. 1997. Legionellae and Legionnaires' disease, p. 666-675. In C. J. Hurst, G. R. Knudsen, M. J. McInerney, L. D. Stetzenbach, and M. V. Walter (ed.), Manual of environmental microbiology, vol. 1. American Society for Microbiology, Washington, D.C.
21. Fields, B. S., R. F. Benson, and R. E. Besser. 2002. Legionella and Legionnaires' disease: 25 years of investigation. Clin. Microbiol. Rev. 15:506-526. [PMC free article] [PubMed]
22. Fliermans, C. B., W. B. Cherry, L. H. Orrison, S. J. Smith, D. L. Tison, and D. H. Pope. 1981. Ecological distribution of Legionella pneumophila. Appl. Environ. Microbiol. 41:9-16. [PMC free article] [PubMed]
23. Guerrieri, E., M. Bondi, F. Foschieri, C. Ciancio, I. Marchesi, and P. Messi. 2003. Role of Acanthamoeba polyphaga in the microbial ecology of Legionella pneumophila and other aquatic bacteria, p. 33-39. In Proceedings of the XVI Congress of European Countries Biologists Association. Congress of European Countries Biologists Association, Abano Terme, Italy.
24. Hambidge, A. 2001. Reviewing efficacy of alternative water treatment techniques. Health Estate 55:23-25. [PubMed]
25. Harb, O. S., L. Y. Gao, and Y. Abu Kwaik. 2000. From protozoa to mammalian cells: a new paradigm in the life cycle of intracellular bacterial pathogens. Environ. Microbiol. 2:251-265. [PubMed]
26. Helbig, J. H., S. Bernander, M. Castellani Pastoris, J. Etienne, V. Gaia, S. Lauwers, D. Lindsay, P. C. Luck, T. Marques, S. Mentula, M. F. Peeters, C. Pelaz, M. Struelens, S. A. Uldum, G. Wewalka, and T. G. Harrison. 2002. Pan-European study on culture-proven Legionnaires' disease: distribution of Legionella pneumophila serogroups and monoclonal subgroups. Eur. J. Clin. Microbiol. Infect. Dis. 21:710-716. [PubMed]
27. Hockey, R. 2002. Safe hot tap water and the risk of scalds and Legionella infection. Injury Prev. 8:170. [PMC free article] [PubMed]
28. Hutchinson, E. J., C. Joseph, and C. L. Bartlett. 1996. EWGLI: a European surveillance scheme for travel associated Legionnaires' disease. Eur. Surveill. 1:37-39. [PubMed]
29. Jaye, C., J. C. Simpson, and J. D. Langley. 2001. Barriers to safe hot tap water: results from a national study of New Zealand plumbers. Injury Prev. 7:302-306. [PMC free article] [PubMed]
30. Joseph, C. A., and European Working Group for Legionella Infections. 2004. Legionnaires' disease in Europe 2000-2002. Epidemiol. Infect. 132:417-424. [PMC free article] [PubMed]
31. Kool, J. L., D. Bergmire-Sweat, J. C. Butler, E. W. Brown, D. J. Peabody, D. S. Massi, J. C. Carpenter, J. M. Pruckler, R. F. Benson, and B. S. Fields. 1999. Hospital characteristics associated with colonization of water systems by Legionella and risk of nosocomial Legionnaires' disease: a cohort study of 15 hospitals. Infect. Control Hosp. Epidemiol. 20:798-805. [PubMed]
32. Kuiper, M. W., B. A. Wullings, A. D. Akkermans, R. R. Beumer, and D. van der Kooij. 2004. Intracellular proliferation of Legionella pneumophila in Hartmannella vermiformis in aquatic biofilms grown on plasticized polyvinyl chloride. Appl. Environ. Microbiol. 70:6826-6833. [PMC free article] [PubMed]
33. Kusnetsov, J., E. Torvinen, O. Perola, T. Nousiainen, and M. L. Katila. 2003. Colonization of hospital water systems by legionellae, mycobacteria and other heterotrophic bacteria potentially hazardous to risk group patients. APMIS 111:546-556. [PubMed]
34. Legnani, P. P., E. Leoni, and N. Corradini. 2002. Legionella contamination of hospital water supplies: monitoring of private healthcare facilities in Bologna, Italy. J. Hosp. Infect. 50:220-223. [PubMed]
35. Levi, K., J. Smedley, and K. J. Towner. 2003. Evaluation of a real-time PCR hybridization assay for rapid detection of Legionella pneumophila in hospital and environmental water samples. Clin. Microbiol. Infect. 9:754-758. [PubMed]
36. Luck, P. C., I. Leupold, M. Hlawitschka, J. H. Helbig, I. Carmienke, L. Jatzwauk, and T. Guderitz. 1993. Prevalence of Legionella species, serogroups, and monoclonal subgroups in hot water systems in south-eastern Germany. Zentrbl. Hyg. Umweltmed. 193:450-460. [PubMed]
37. Lye, D., G. Shay Fout, S. R. Crout, R. Danielson, C. L. Thio, and C. M. Paszko-Kolva. 1997. Survey of ground, surface, and potable waters for the presence of Legionella species by EnviroAmpR PCR Legionella kit, culture, and immunofluorescent staining. Water Res. 31:287-293.
38. Momba, M. N. B., R. Kfir, S. N. Venter, and T. E. Cloete. 2000. An overview of biofilm formation in distribution systems and its impact on the deterioration of water quality. Water SA (Pretoria) 26:59-66.
39. Murga, R., T. S. Forster, E. Brown, J. M. Pruckler, B. S. Fields, and R. M. Donlan. 2001. The role of biofilms in the survival of Legionella pneumophila in a model potable water system. Microbiology 147:3121-3126. [PubMed]
40. Neumeister, B., S. Schoniger, M. Faigle, M. Eichner, and K. Dietz. 1997. Multiplication of different Legionella species in mono Mac 6 cells and in Acanthamoeba castellanii. Appl. Environ. Microbiol. 63:1219-1224. [PMC free article] [PubMed]
41. Ortiz-Roque, C. M., and T. C. Hazen. 1987. Abundance and distribution of Legionellaceae in Puerto Rican waters. Appl. Environ. Microbiol. 53: 2231-2236. [PMC free article] [PubMed]
42. Palmer, C. J., G. F. Bonilla, B. Roll, C. Paszko-Kolva, L. R. Sangermano, and R. S. Fujioka. 1995. Detection of Legionella species in reclaimed water and air with the EnviroAmp Legionella PCR kit and direct fluorescent antibody staining. Appl. Environ. Microbiol. 61:407-412. [PMC free article] [PubMed]
43. Passi, C., R. Maddaluno, and M. C. Pastoris. 1990. Incidence of Legionella pneumophila infection in tourists. Ital. Public Health 104:183-188. [PubMed]
44. Ricketts, K., and C. Joseph. 2004. Travel associated Legionnaires' disease in Europe: 2002. Eur. Surveill. 9:6-9. [PubMed]
45. Ricketts, K., and C. Joseph. 2004. Travel associated Legionnaires' disease in Europe: 2003. Eur. Surveill. 9:5-6.
46. Rogers, J., A. B. Dowsett, P. J. Dennis, J. V. Lee, and C. W. Keevil. 1994. Influence of temperature and plumbing material selection on biofilm formation and growth of Legionella pneumophila in a model potable water system containing complex microbial flora. Appl. Environ. Microbiol. 60:1585-1592. [PMC free article] [PubMed]
47. Rota, M. C., M. G. Caporali, and M. Massari. 2004. European guidelines for control and prevention of travel associated Legionnaires' disease: the Italian experience. Eur. Surveill. 9:10-11. [PubMed]
48. Rota, M. C., M. L. Ricci, M. G. Caporali, and S. Salmaso. 2004. La legionellosi in Italia nel 2003. Not. Ist. Super Sanit. 17:8-13.
49. Rowbotham, T. J. 1980. Preliminary report on the pathogenicity of Legionella pneumophila for freshwater and soil amoebae. J. Clin. Pathol. 33:1179-1183. [PMC free article] [PubMed]
50. Schoonmaker, D., T. Heimberger, and G. Birkhead. 1992. Comparison of ribotyping and restriction enzyme analysis using pulsed-field gel electrophoresis for distinguishing Legionella pneumophila isolates obtained during a nosocomial outbreak. J. Clin. Microbiol. 30:1491-1498. [PMC free article] [PubMed]
51. Steinert, M., U. Hentschel, and J. Hacker. 2002. Legionella pneumophila: an aquatic microbe goes astray. FEMS Microbiol. Rev. 26:149-162. [PubMed]
52. Tenover, F. C., R. D. Arbeit, R. V. Goering, P. A. Mickelsen, B. E. Murray, D. H. Persing, and B. Swaminathan. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233-2239. [PMC free article] [PubMed]
53. Thomas, V., T. Bouchez, V. Nicolas, S. Robert, J. F. Loret, and Y. Levi. 2004. Amoebae in domestic water systems: resistance to disinfection treatments and implication in Legionella persistence. J. Appl. Microbiol. 97:950-963. [PubMed]
54. Vickers, R. M. V. L. Yu, S. S. Hanna, P. Muraca, W. Diven, N. Carmen, and F. B. Taylor. 1987. Determinants of Legionella pneumophila contamination of water distribution systems: 15-hospital prospective study. Infect. Control 8:357-363. [PubMed]
55. Wadowsky, R. M., and R. B. Yee. 1985. Effect of non-Legionellaceae bacteria on the multiplication of Legionella pneumophila in potable water. Appl. Environ. Microbiol. 49:1206-1210. [PMC free article] [PubMed]
56. Walker, J. T., D. J. Bradshaw, A. M. Bennett, M. R. Fulford, M. V. Martin, and P. D. Marsh. 2000. Microbial biofilm formation and contamination of dental-unit water systems in general dental practice. Appl. Environ. Microbiol. 66:3363-3367. [PMC free article] [PubMed]
57. Wellinghausen, N., C. Frost, and R. Marre. 2001. Detection of legionellae in hospital water samples by quantitative real-rime LightCycler PCR. Appl. Environ. Microbiol. 67:3985-3993. [PMC free article] [PubMed]
58. Yu, V. L. 2002. Legionella surveillance: political and social implications—a little knowledge is a dangerous thing. J. Infect. Dis. 185:259-261. [PubMed]
59. Yu, V. L., J. F. Plouffe, M. C. Pastoris, J. E. Stout, M. Schousboe, A. Widmer, J. Summersgill, T. File, C. M. Heath, D. L. Paterson, and A. Chereshsky. 2002. Distribution of Legionella species and serogroups isolated by culture in patients with sporadic community-acquired legionellosis: an international collaborative survey. J. Infect. Dis. 186:127-128. [PubMed]
60. Zacheus, O. M., and P. J. Martikainen. 1994. Occurrence of legionellae in hot water distribution systems of Finnish apartment buildings. Can. J. Microbiol. 40:993-999. [PubMed]
61. Zacheus, O. M., and P. J. Martikainen. 1995. Occurrence of heterotrophic bacteria and fungi in cold and hot water distribution systems using water of different quality. Can. J. Microbiol. 41:1088-1094.

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...