However, because water-borne transmission is one of the major routes of acquiring infection, it must also be recognized that infection rates for either enteric parasite may vary by geographic location. One study in southern Ontario demonstrated an association between giardiasis and rural areas and this has been corroborated by subsequent GIS spatial scan statistics investigation of clusters of giardiasis in this area [14]. Although a previously study had shown significant associations of giardiasis rates with manure application on agricultural land and livestock density this was not born out using spatial statistics scanning methods [15].

If adequate denominator data is available then incidence rates may be calculated from population-based datasets. The rates we determined in this study should be viewed as the minimum incidence because multiple stool sampling is not a standard practice in our region, and many patients may have had these diseases without having a stool sample submitted for parasitological testing. In addition, certain groups that seek medical attention less frequently may be underrepresented using a population-based laboratory surveillance approach (i.e., shut-in elderly, intravenous drug users, those with mental illness etc.), but these limitations would also be a reality of other approaches. However, even considering these limitations, Giardia and Cryptosporidium result in a significant disease burden in Calgary. We found that infections with Giardia sp. and/or Cryptosporidium sp. were common in our region.

Epidemiological analysis of regional stool parasitological testing data over a two-year period demonstrated that these two parasites were the most commonly identified pathogenic enteric protozoa in the Calgary Health Region, which is similar to the rates previously reported in the United States [1,15,24]. Population-based studies in the United States have indicated rates of Giardia sp. infections ranging from 0.98 to 42.3 cases per 100,000 per year with an average of 9.5 cases per 100,000 per year [1]. However, true population based rates for Cryptosporidum sp. based on laboratory surveillance data have been difficult to determine in the United States because most diagnostic laboratories do not routinely perform stool parasitological procedures that would allow detection of this parasite. A recent survey indicates that most laboratories only analyze stool parasitological samples for the presence of Cryptosporidium sp., Cyclspora sp. and Microsporidium sp. on specific request of the physician [25]. Based on laboratory surveillance data from clinical microbiology laboratories, there were a total of 443 cases of cryptosporidiosis reported to the CDC in the United States in 2003 [26] which is much lower than that expected from population-based estimates of infection rates. Estimates for the Cryptosporidium sp. infection rates for the general population (i.e., 1 – 400 cases per 100,000) in the United States have been based on US-wide water monitoring data [27]. The population-based rate of giardiasis and cryptosporidiosis cases in Calgary is similar to the national average rate per 100,000 reported from the United States. However, the rate of Cryptosporidium sp. infections found in our region is well above the proposed acceptable annual risk level of 1 case of infection per 10,000 [13]. Substantially more Cryptosporidium sp. infections were found in our region during the study period. To our knowledge studies conducted to date from Canada have also not been adequately designed to report incidence rates to allow comparison. The results of our study demonstrate that Giardia sp. and Cryptosporidium sp. infections are common with the former occurring at a rate of more than 3 times more frequently.

We observed a number of similar and contrasting features among infections due to Giardia and Cryptosporidium sp. Giardia sp. infections occurred more commonly in late summer in our region and may be related to water exposure and the increased use of backcountry recreational areas. On the other hand, there was no apparent yearly seasonal occurrence of sporadic (i.e. non-outbreak related) Cryptosporidium sp. infections in our region (Figure (Figure3).3). Although few studies of the seasonal incidence of intestinal parasites have been done, prior reports from specific geographic locations in the United States suggest that Cryptosporidium sp. infection is most prevalent in the spring, while no seasonality was found in Giardia sp. infections [28]. Infection with either parasite occurs when cysts are ingested via contaminated hands, food, and/or water or through person-to-person contact. Individual cases as well as outbreaks have therefore been associated with improper food handling, exposure to contaminated water (i.e. municipal sources including swimming pools, surface and groundwater including those found in beaver ponds and springs), travel to less developed countries or close contact with a case (i.e. families, day care centers) [12,29-33]. Specific risk factors that have recently been associated with sporadic cryptosporidiosis among immunocompetent persons in the United States include international travel, contact with cattle, contact with persons >2 to 11 years of age with diarrhea and freshwater swimming [34]. Temporal investigation of fecal shedding of Cryptosporidium parvum oocysts in cow-calf herds has also shown that most fecal shedding of this parasites by infected cattle is limited to calves 1 to 4 months old so that the risk of parasitic watershed contamination is greatest in the spring during calving season [35]. Enteric parasitic infection with either Giardia sp. or Cryptosporidium sp. may also be transmitted through sexual contact and immunocompromised persons (i.e. acquired immunodeficiency syndrome) are particularly at risk of developing severe persistent infection [36]. Differences in the seasonality of intestinal parasite prevalence between disparate geographic locations may therefore be dependent on a variety of factors including the composition of the population and their coming into contact with various contaminated water sources (i.e. urban versus rural areas).

Another contrasting feature among these infections was that the age distribution was different. Giardia sp. infections, although most commonly infecting children, occurred across a broad age range (Figure (Figure2).2). In contrast, Cryptosporidum sp. infections occurred primarily in young children. The main commonality amongst these two infections was that they both tended to more commonly affect males as compared to females [11,37]. Although this study did not investigate the risk factors associated with disease acquisition in various age groups, the demographics of infection is similar to prior studies [11,13,34,37]. Young children tend to have a higher rate of infection with Giardia sp. and/or Cryptosporidium sp. because of their exposure to contaminated municipal and rural water sources such as well water and swimming pools and contact with cases in communal living situations including day care centers. Epidemiological investigation into the risk factors for enteric infection due to one or both of these parasites is the focus of future studies in our region.

Our observation that an outbreak of Cryptosporidium sp. occurred in the third year of the study demonstrates the importance of their detection by conducting population-based studies over extended periods (years). We have previously raised the theoretical potential for bias in population-based studies designed to assess the burden and risk factors for disease if they are conducted in the setting of an outbreak [38]. If this study had only been conducted over a one-year period in 2001 we would have made erroneous inferences on the occurrence and demographic risk factors for endemic disease as evidenced by the dramatically different rate in that year and that outbreak patients were younger. In our opinion it is important that "baseline" or endemic disease be differentiated from outbreak cases in the description of the epidemiology of an infectious disease. We urge authors of other studies to plot cases over time prior to detect outbreaks prior to description.