Symptoms. The term symptom in this study is defined as any change in normal body function induced by direct or indirect action of parasites.  Direct action includes invasiveness and tissue damage due to parasite feeding or migration.  Indirect action results from parasite metabolic byproducts and toxic secretions.  Symptoms are in two categories:  1) gastrointestinal, including (in order of observed frequency) flatulence, diarrhea, bloating, abdominal cramping, constipation, malabsorption / maldigestion, bloody or odorous stool, irritable bowel, mucus, and leaky gut, and 2) extra-intestinal (systemic), including (in order of observed frequency) fatigue, nervous/sensory disorders, pain, skin disorders, allergies, nausea, muscle weakness/pain, immune deficiencies, headache, fever/ night sweats, insomnia, and weight changes.  Most infected patients with parasitic symptoms experienced 1 – 4 gastrointestinal and/or extra-intestinal symptoms; they are simply called symptoms. 

    Of  the 826 patients with single infections, 584 (70%) experienced overt symptoms and 242 (30%) had none (Table 2).  Infections with pathogenic protozoa included 0 – 31% asymptomatic infections.  Protozoans regards as non-pathogenic were associated with symptoms in 69 – 100% of the cases.  Ascaris lumbricoides produced no symptoms in one of 12 cases.

    Multiple infections.  Ninety patients (10% of cases) were concurrently infected with 2 – 4 species of parasites.  Among these, 21 patients experienced no symptoms.  These 21 cases involved 19 infections with B. hominis, 7 with C. parvum, 6 with Endolimax nana, 5 with E. histolytica/E. dispar, 4 with Giardia lamblia, 3 with Entamoeba coli, 2 with Chilomastix mesnili, Cyclospora cayetanensis, and Iodamoeba butschilii, and 1 with Retortomonas intestinalis.  The remaining 69 patients with concurrent infections were symptomatic.  These 69 cases involved 58 infections with B. hominis, 22 with C. parvum, 19 with E. coli, 18 with E. histolytica/E. dispar, 16 with E. nana, 5 with C. cayetansis and G. lamblia, 2 with Ascaris lumbricoides, and 1 with C. mesnili, Dientamoeba fragilis, Entamoeba hartmanni, I. Butschlii, R. intestinalis, and Taenia sp. 

    Seasonality.  Monthly seasonal prevalence of single and multiple infections gradually increased from a minimum in February to a maximum between August and October, then decreased in December.  Prevalence of infection with Blastocystis hominis was lowest in May and highest in September and November.  Peak prevalences were observed in C. parvum in the spring (March).  Oscillations in seasonal frequencies were not dramatic in other parasite species and most were represented in all seasons (Table 1). 

    Charcot-Leyden crystals.  These crystals were found in 34 specimens of which 21 (62%) were infected with B. hominis, C. parvum, E. histolytica/E. dispar, and G. lamblia.  No parasites were detected in the remaining 13 (38%) species (Table 3).

DISCUSSION

    This study population was demographically similar to the 644 patient population studied during the summer of 1996 under the same circumstances by Parasitology Center Inc.¹ In the present study, overall infection prevalence rates were comparable throughout the country and did not vary much between the southwest, the west coast, the Midwest, and the east coast.  Only the larger sample sizes are considered.  Patient age was between 0 and 80 years, approximately twice as many females (1,945) than males (951) were tested, relatively more cases (550 of 916, 60%) had a history of foreign travel than non-cases (970 of 1980, 49%) within the last five (P< 0.001, by Fisher’s exact test relative risk [RR] = 1.36, confidence interval [CI] = 1.22 – 1.52.  Relatively more infected (110 of 916, 12%) than uninfected (198 of 1980, 10%) patients lived with infected household contacts (P = 0.117, not significant, by Fisher’s exact test).

    Our new methods of parasite detection, adopted since 1996, reflected prevalence rates considered closer to true prevalences compared with standard methods used.⁶ By 1998, 3,373 (32.6%) of 10, 358 specimens examined at Parasitology Center, Inc. were infected with parasites.⁶ An almost identical prevalence of 32% is reported in this study (Table 1).  This prevalence is markedly higher than reported prevalences in the United States of 20% (from 216,275 stool specimens) and 19.7% (from 178,786 stool specimens) reported by state diagnostic laboratories in 1987.² The markedly higher prevalence in our study (P<0.001 by Fisher’s exact test, RR = 1.9, CI = 184 – 1.99) suggests real increases in prevalence but does not exclude the possibility of differences in test populations.  The results of the latter report also differ significantly from our finding in the composition of the component parasites species found, e.g., B. hominis was diagnosed in only 2.6% of the specimens examined compared with 23% (or 72% of all infected cases) in our study.  In a 1984 study of 2,360 patients in the United States, prevalences of 20.6% for all parasitic species and 12.2% for cases of B. hominis infections were reported with B. hominis constituting 59% of all infections.³ The latter figure is much closer to our current finding of a B. hominis prevalence of 72% among all parasitic infections (Table 1).  In 1995, overall B. hominis prevalences of 20 – 30% and greater than 15% were also reported from an unspecified number of patients.⁹

    The prevalence of B. hominis reported herein (23%) is one of the highest ever reported in the United States and may be epidemiologically significant.  Increasing prevalences are noted in more recent years.  This prevalence is closest to that reported for Argentina (25%)¹⁰ and Switzerland (16.7 – 19.0%)¹¹ but considerably lower than those in other studies from Argentina (43%)¹² and Chile (61.8%)¹³

    The second most prevalent parasite found was C. parvum (Table 1).  Prevalences reported in surveys from North America (0.6 – 4.3%) and Europe (1 – 2%) are significantly lower than those reported for Asia, Australia, Africa, Central America, and South America (3 – 20%).¹⁴  Cryptosporidium parvum appears to be underdiagnosed in the western hemisphere; its seroprevalence in Europe and North America is usually between 25% and 35%.¹⁴  In a recent survey of 279 children from three clinics along the Texas-Mexico border 96 children (70.2%) were found infected with C. parvum.¹⁵  Children living in a large non-border urban area were less frequently infected , drank more bottled water, and came from households with higher income.¹⁵  Cryptosporidium oocysts were observed in 27% of the drinking water samples taken from 66 surface water treatment plants in 14 states and one Canadian province.¹⁶

    The E. histolytica/E. dispar prevalence (Table 1) is markedly higher than the prevalence of 0.9% reported in a large survey in 1987 in the United States, but lower than the estimated prevalence of 4% in the United States.¹⁷ In developing countries with poor sanitation, the prevalence may reach as high as 50%.¹⁷  prevalences quoted for E. histolytica infections are clearly misleading since more than 90% of these infections are due to E. dispar.^17,18  In the Philippines, a polymerase chain reaction survey of 1,872 patients detected 137 stools (7.3%) containing E. dispar and 18 stools (1.0%) containing E. histolytica.¹⁹  The importance of developing a simple and inexpensive way of distinguishing the two species to obtain information on true prevalence, pathogenecity, and treatment can not be overlooked.

    The remaining parasites recovered in this study were of minor importance and their overall prevalence was comparable or somewhat lower than those reported in other surveys.  The very low prevalence of C. cayetensis agrees with other findings, suggesting under diagnosis in indigenous populations in the United States.²⁰

    Symptoms in 826 singly infected patients did not always agree with the purported pathogenecity of the parasites involved (Table 2).  Approximately one third of B. hominis infections were not associated with symptoms.  Asymptomatic infections with B. hominis varied between 30% and 60% in various populations in the United States.^1,3,9,21,22  In Canada, B. hominis is usually asymptomatic.⁵  It is not known if the degree of pathogenicity of B. hominis is related to the distinct immunologic, serologic, and genetic identity of the demes constituting that species.^23,24  The epidemiologic significance of these findings and the B. hominis species complex question remain to be resolved.  We regard B. hominis as a species complex usually showing pathogenicity based on our findings (Table 2) and those reported by others; see the review by Garcia.²⁵ This parasite will cause asymptomatic subclinical infections similar to most pathogens. 

    Asymptomatic periods in the intermittent and recurrent infectious cycle of C. cayetensis are evident and may be attenuated with long-term suppressive therapy.  During the asymptomatic periods, C. cayetensis is often absent (undetectable) in stool specimens.  Asymptomatic infections with C. parvum are directly related to the immune status of the host²⁶ and have been previously reported in immuno-competent persons.¹⁴ The asymptomatic cases of E. histolytica/E. dispar are attributed to E. dispar infections.  These observations also apply to the 21 asymptomatic patients of the 90 (23%) multiply infected cases.

    Six species of presumably non-pathogenic protozoa singly infecting 72 patients were associated with symptoms (Table 2).  It is our experience that a host body, particularly if immune-compromised, will not be indifferent to the presence of foreign organisms irrespective of their purported non-pathogenic status.  Until recently, B. hominis was considered to be a harmless yeast.  The literature is beginning to show awareness of the pathogenic potential of such “harmless” organisms such as Dientamoeba fragilis,^25,26 E. coli,^1,27,28 and E. hartmanni.^1,25,29  The non-pathogenic status of these organisms is questioned in light of our findings.

    Seasonal studies of human parasite infections are rarely conducted in the United States.  An increase in the prevalence of C. parvum (comparable to ours) in the spring observed in patients from New Orleans³⁰ and elsewhere¹⁴ was attributed to the warm wet spring weather.  However, our overall seasonal data and that of B. hominis show the highest prevalence between August and October.  The latter dates do not essentially negate possible spring/early summer prevalence peaks.  Exposure, experiencing symptoms, seeking medical help, and testing may explain this time lag.

    Charcot-Leyden crystals are breakdown products of eosinphils found usually in feces and occasionally in sputum and body tissues.³¹  The crystals have been traditionally associated with E. histolytica infections,³² but have been more recently found in patients infected with Ancylostoma spp., A. lumbricoides, Isospora belli, and Trichuris trichura.²⁵  In the present study, it was also found in patients singly infected with B. hominis, C. parvum, and G. lamblia.  The 13 patients with crystals but with no detectable infections were probably infected with a cyclic parasite such as E. histolytica.