Risks of Recreational Exposure to Waterborne Pathogens Among Persons With HIV/AIDS in Baltimore, Maryland

Abstract

Objectives. We assessed the prevalence of recreational activities in the waterways of Baltimore, MD, and the risk of exposure to Cryptosporidium among persons with HIV/AIDS.

Methods. We studied patients at the Johns Hopkins Moore Outpatient AIDS Clinic. We conducted oral interviews with a convenience sample of 157 HIV/AIDS patients to ascertain the sites used for recreational water contact within Baltimore waters and assess risk behaviors.

Results. Approximately 48% of respondents reported participating in recreational water activities (fishing, crabbing, boating, and swimming). Men and women were almost equally likely to engage in recreational water activities (53.3% versus 51.3%). Approximately 67% (105 of 157) ate their own catch or that of friends or family members, and a majority (61%, or 46 of 75) of respondents who reported recreational water contact reported consumption of their own catch.

Conclusions. Baltimoreans with HIV/AIDS are engaging in recreational water activities in urban waters that may expose them to waterborne pathogens and recreational water illnesses. Susceptible persons, such as patients with HIV/AIDS, should be cautioned regarding potential microbial risks from recreational water contact with surface waters.

Persons with HIV/AIDS are at high risk for increased morbidity and mortality associated with a range of opportunistic infections, some of which are caused by Cryptosporidium. Cryptosporidium species are of particular public health and medical importance because they are prevalent in surface waters of the United States,^1–7 are efficiently transmitted via water,⁸ and can be consumed in foods contaminated by fecal matter.^9–11 Exposures to Cryptosporidium are common in the US population,¹² and past studies have demonstrated that Cryptosporidium infections significantly contribute to illness and mortality in persons with HIV/AIDS.^13–15 In the 1980s, Cryptosporidium was identified as a major opportunistic pathogen.^16–21 Infection continues to be frequently diagnosed in persons with HIV/AIDS.^22–27 Before the advent of highly active antiretroviral therapy, Cryptosporidium was a relatively common opportunistic infection even in developed countries.^28,29

Cryptosporidiosis manifests as an acute gastroenteritis, accompanied by cramps, anorexia, vomiting, abdominal pains, fever, and chills²⁹ and by histological presentation of gastrointestinal mucosal injury.^30,31 Persons with AIDS who become infected with this parasite are at increased risk of developing chronic and often life-threatening diarrhea, biliary tract diseases, pancreatitis, colitis, and chronic asymptomatic infection and recurrence. These developments are especially likely in those who are severely immunosuppressed (CD4 counts < 150 cells/mL).^29,32–35 Infection is diagnosed by the presence of oocysts in unpreserved or preserved stools.³⁶ Histological and ultrastructural examination of biopsy material for different Cryptosporidium life stages, detection of Cryptosporidium DNA and antigens, and identification of species through molecular techniques can also aid in diagnosis.^36–38

Cryptosporidium species are enteric protozoan organisms and are prevalent in US watersheds, especially in urban waters.^1,6,39 These parasites have natural hosts in domestic and wild animals such as cattle (especially newborn calves), horses, fish, and birds.^5,40–42 These parasites cause cryptosporidiosis by infecting and damaging the cells of the small intestine and other organs.^13,41 For persons with HIV/AIDS, increased risk for infection by Cryptosporidium has been related to sexual practices such as engaging in sexual intercourse within the past 2 years, having multiple partners during that time, and engaging in anal intercourse.⁴³ Use of spas and saunas has also been identified as a risk factor.⁴³

In the United States, Cryptosporidium is the most commonly identified pathogen in cases of recreationally acquired gastroenteritis⁴⁴; the majority of those affected are children. Increased risk of cryptosporidiosis in persons with HIV/AIDS has been associated with swimming.^45,46 US residents make an estimated 360 million annual visits to recreational water venues such as swimming pools, spas, and lakes; swimming is the second most popular physical activity in the country and the most popular among children.⁴⁷

Recreational swimming, even in highly chlorinated water, carries a high risk of exposure to enteric pathogens, including Cryptosporidium, Norovirus, Shigella, Escherichia coli, and Giardia.⁴⁸ Cryptosporidiosis and some other enteric illnesses are seasonal, with spikes in occurrence in the summer months from contact with recreational water venues.⁴⁹ Extreme precipitation⁵⁰ and high ambient temperatures⁵¹ can also affect patterns of disease outbreaks. Because not all infections with Cryptosporidium lead to apparent illness or symptoms, infected persons may unknowingly transmit these pathogens to others, such as household members and other recreationists.^12,52 Cryptosporidiosis from swimming, wading, and splashing is prevalent in the United States.^44,46,53,54

Risks from the presence of pathogens in waterways include (1) waterborne gastroenteritis and other recreational water illnesses in anglers and other recreationists^44,55–59; (2) transmission of pathogens to humans from caught seafood acting as fomites, or surface carriers⁶⁰; (3) food-borne gastroenteritis from consumption of raw or improperly cooked fish and shellfish^61,62; and (4) hand-to-mouth transmission of pathogens while eating, drinking, or smoking during activities such as fishing and crabbing.⁷

Recreational water activities in the Baltimore, Maryland, area take place in Jones Falls and Baltimore Harbor. These and other waterways are used for angling, crabbing, swimming, kayaking, and boating (including paddle boating).^7,63 In addition, Baltimore-area residents often catch and consume fish and crabs from the Baltimore Harbor and local waterways, many of which are already highly contaminated by persistent chemicals such as mercury and polychlorinated biphenyls.⁶⁴ These activities are known to increase risks of exposure to waterborne pathogens through direct contact with contaminated waters or through contact with or handling and consumption of caught seafood (fish, crabs, oysters).^7,65,66

To investigate the potential contribution of recreational water contact to Cryptosporidium exposures among persons with HIV/AIDS, we carried out a cross-sectional study at the Johns Hopkins Moore Outpatient AIDS Clinic. The Baltimore metropolitan area has a high prevalence rate of HIV/AIDS among both men and women,⁶⁷ and its population makes intensive recreational use of a contaminated watershed. In addition, laboratory experiments have indicated that crabs can become superficially contaminated by Cryptosporidium and transfer the pathogen to hands.⁶⁸ Local anglers are at risk from Cryptosporidium on wild-caught fish.⁷

Our objective was to address the risks of exposure to Cryptosporidium for an urban subpopulation, persons with HIV/AIDS, as a result of recreational contact with Baltimore waterways. We also assessed the patterns and locations of recreational water activities in Baltimore waters.

METHODS

The Hopkins AIDS Care Program and Moore Outpatient AIDS Clinic serve persons from the entire Baltimore metropolitan region; as of December 2008, approximately 3000 patients were being actively followed. Of these 31.3% were women, 78.2% were African American, 20.1% were non-Hispanic White, and 1.6% were Hispanic. Fifty percent of these patients were intravenous drug users, 22% were men who have sex with men, and 25% were heterosexual. These demographics are reflective of the AIDS epidemic in Baltimore and central Maryland.

Study Design

We identified a convenience sampling of patients that represented the typical clientele of the Moore clinic; our overall sample target was 150 patients. Our study was conducted between July 10 and September 17, 2007. Consent was obtained from respondents. The study was facilitated by R. D. M., the director of the clinic, as well as Vivian Zhao, the clinic administrator. Patients were ineligible if they worked with farm animals, had contact with animals from a petting zoo, or had previously completed the survey. No personal identifiers were collected during the course of the study. In total, 157 patients consented and fulfilled the criteria to participate in the study.

We used an oral interview format validated in a pilot project.⁶⁹ The survey was administered in the lobby of the clinic and in a downstairs office between 9 am and 2 pm. Participants were asked to respond to questions about their place of residence, age, gender, ethnicity, locations of recreational water activities (specifically crabbing), patterns of activity, consumption of wild-caught fish and crabs from Baltimore waterways, and hand-washing habits during recreational water activities. A Batimore County watershed map was used to facilitate location of sites used for recretional water activities (Figure 1). They also reported current and recent recreational water activity by month and frequency and responded to questions about potential risk factors for Cryptosporidium transmission such as contact with diapered children, consumption of untreated water, travel overseas, pet ownership, and wild-animal contact.

FIGURE 1.

Map of Baltimore County watershed used during survey to identify sites of recreational water activity: 2007.

Source. Baltimore County Environmental Protection and Resource Management Watershed Management Program.⁷³

Most responses were qualitative (yes, no, not applicable). We used a categorical frequency scale to quantify monthly recreational activity (e.g., 0–4 times/mo). The term recent referred to any event occurring with the previous month. Respondents were compensated with a $10 store gift card upon completion of the survey.

Responses to survey questions were checked off or handwritten on the survey by C. C. M. and 2 other survey administrators. Respondents were also encouraged to comment on the survey and to ask any questions they thought were relevant.

Statistical Analyses

Data were entered into Microsoft Excel 2003⁷⁰ and analyzed with STATA version 9.2.⁷¹ We made statistical comparisons with the Fisher exact test and χ² test. Statistical significance was set at .05. We used the 2-sided t test for means to assess difference in age by categories. We assessed group differences with the binomial test of proportions and used nonmissing data for estimations.

RESULTS

Participants were predominantly African American and evenly distributed between men and women (Table 1). The participation rate was more than 90% among clinic patients who were approached. Demographic characteristics and risk exposure behaviors of the study population are presented in Table 1 and Table 2. Because of small numbers, racial/ethnic groups other than African American and non-Hispanic White were collapsed into 1 category, other (Hispanic, Asian, biracial, and African). There were 75 women, 80 men, and 1 transgendered person. One respondent did not provide gender information.

TABLE 1.

Race/Ethnicity and Gender Comparisons Between Study Participants (n = 157) and Clinic Patients (n = 1830) of the Johns Hopkins Moore Outpatient AIDS Clinic: Baltimore, MD, 2007

	Participants	Clinic Patientsa
Race/ethnicity, %
African American	86.6	74
Non-Hispanic White	7	23
Biracial	3.2	…
Hispanic	1.3	2
Other	0.6	1
Missing	1.3	…
Gender, %
Men	51.3	65
Women	48	35
Transgender	0.7	…
Age, y, median (range)	47 (27–75)	41 (19–88)

Note. Ellipses indicate data were not available or not collected.

^aData were from 2003 to 2007.

TABLE 2.

Participant Characteristics and Risk Behaviors, by Gender and Race/Ethnicity: Baltimore, MD, 2007

	Gender		Race/Ethnicity
	Men (n = 80)	Women (n = 75)	African American (n = 136)	Non-Hispanic White (n = 11)	Other (n = 8)
Age, y, mean ±SD (95% CI)	48.2 ± 7.7 (46.5, 49.9)	46.1 ± 8.9 (43.9, 48.1)	47.1 ± 7.8 (45.8, 48.5)	44 ± 8.3 (37.6, 50.9)	51 ± 14.9 (38.5, 63.5)
Age, y, median (range)	49 (27–68)	45 (27–75)	47 (27–75)	48 (30–53)	50 (27–68)
Recreational water contact, No. (%)
Yes	41 (51.3)	40 (53.3)	71 (52.2)	5 (45.5)	5 (62.5)
No	39 (48.7)	35 (46.7)	65 (47.8)	6 (54.5)	3 (37.5)
Consumption of self-caught fish and crabs, No. (%)
Yes	24 (30)	22 (29.3)	41 (30.1)	3 (27.3)	2 (25)
No	3 (3.8)	3 (4)	4 (2.9)	2 (18.2)	0 (0)
Not applicablea	52 (65)	50 (66.7)	91 (66.9)	6 (54.5)	6 (75)
Missing	1 (1.25)	0 (0)	0 (0)	0 (0)	0 (0)
Consumption of others' wild-caught fish and crabs
Yes	46 (57.5)	52 (69.3)	88 (64.7)	6 (54.5)	4 (50)
No	32 (40)	22 (29.3)	46 (33.8)	5 (45.5)	4 (50)
Missing	2 (2.5)	1 (1.4)	2 (1.5)	0 (0)	0 (0)
Handwashing during recreational activity, No. (%)
Yes	31 (38.8)	25 (33.3)	50 (36.8)	4 (36.4)	2 (25)
No	6 (7.5)	6 (8)	10 (7.4)	2 (18.2)	0 (0)
Not applicablea	41 (51.3)	40 (53.3)	72 (52.9)	5 (45.5)	5 (62.5)
Missing	2 (2.5)	4 (5.3)	4 (2.9)	0 (0)	1 (12.5)
Pets in home, No. (%)
Yes	29 (36.3)	29 (38.7)	51 (37.5)	5 (45.5)	2 (25)
No	49 (61.3)	44 (58.7)	84 (61.8)	4 (36.4)	6 (75)
Missing	2 (2.4)	2 (2.6)	1 (0.7)	2 (18.1)	0 (0)
Consumption of well water, No. (%)
Yes	4 (5)	4 (5.3)	7 (5.2)	1 (9.1)	0 (0)
No	74 (92.5)	70 (93.3)	128 (94.1)	9 (81.8)	8 (100)
Missing	2 (2.5)	1 (1.33)	1 (0.7)	1 (9.1)	0 (0)
Travel outside United States, No. (%)
Yes	1 (0.8)	1 (1.3)	1 (0.7)	0 (0)	1 (12.5)
No	77 (96.3)	73 (97.3)	134 (98.5)	10 (90.9)	7 (87.5)
Missing	2 (2.5)	1 (1.33)	1 (0.7)	1 (9.1)	0 (0)
Changing of baby diapers, No. (%)
Yes	13 (16.3)	26 (34.7)	34 (25)	2 (18.2)	3 (37.5)
No	65 (81.3)	47 (62.7)	101 (74.3)	7 (63.6)	5 (62.5)
Missing	2 (2.4)	2 (2.6)	1 (0.7)	2 (18.2)	0 (0)
Contact with wild animals, No. (%)
Yes	4 (5)	3 (4)	6 (4.4)	1 (9.1)	0 (0)
No	73 (91.3)	70 (93.3)	128 (94.1)	8 (72.7)	8 (100)
Missing	3 (3.7)	2 (2.7)	2 (1.5)	2 (18.2)	0 (0)

Note. CI = confidence interval. Totals across categories may not sum to total study population because of missing data.

^aParticipant did not engage in this activity.

The mean age of the 152 participants whose age was reported was 47.2 ±8.4 years (range = 27–75 years). The median age was 47 years, with the age distribution slightly positively skewed (0.35). There was no difference in age between respondents who engaged in recreational water activities and those who did not (P = .32). Ninety-five percent of the respondents were 61 years or younger. Almost half of the respondents (47.8%, or 75 persons) reported some form of recreational water activity, and 82 participants (52.2%) reported none. Similar numbers of men and women engaged in recreational water activity (39 and 35, respectively), and similar numbers did not (41 and 40).

The most common locations for reported recreational water activity were the Baltimore Harbor (Canton, Fells Point), Patapsco River (Middle Branch Park, Harbor Hospital), and Gunpowder River (Figure 2). These sites are centrally located within the Baltimore County watershed (Figure 1).⁷² Respondents gave multiple reasons for engaging in fishing and crabbing. The majority (n = 75) reported that they fished or crabbed for fun, recreation, and relaxation (68%) as well as for food or to supplement diets (58.9%). Other reported reasons were sharing caught fish and crabs (45%) and spending time with family or children (2.7%).

FIGURE 2.

Recreational water sites used by patients surveyed at the Johns Hopkins Moore Outpatient AIDS Clinic, Baltimore, MD: 2007.

Note. The number reporting activities includes multiple sites and activities.

Among risk factors associated with Cryptosporidium infection and transmission, the most commonly reported were changing baby diapers (39 respondents, or 24.8%) and pets in the home (58, or 36.9%). Eight respondents (5.1%) reported well water consumption, 2 (1.3%) reported travel outside the United States, and 7 (4.5%) reported contact with wild animals (wild birds, raccoons, or deer). The distribution of these risk factors by race/ethnicity and gender are shown in Table 2.

We estimated contact with wild-caught fish and crabs by asking about consumption of fish and crabs caught by others. Approximately 67% of participants (105 of 157) reported consumption of their own catch or wild-caught fish and crabs from friends or family members. A majority (46 of 75, or 61%) of respondents who reported recreational water contact ate their own catch. Consumption of fish and crabs caught in the wild by either friends or family members was reported by approximately 80% (59 of 74) of those who engaged in recreational water activity and only 48% (38 of 79) of those who did not.

Respondents also provided data on their frequency of recreational water activities (Table 3), with the majority reporting taking part in activities 1 to 4 times a month.

TABLE 3.

Study Participants' Frequency of Participating in Recreational Water Activities, by Location: Baltimore, MD, 2007

Activity	Patapsco River	Baltimore Harbor	Gunpowder River	Back River
Fishing, times/mo
1–4	21	8	11	7
5–8	4	0	0	0
≥ 9	1	1	0	0
Missing	5	1	0	1
Not applicablea	44	65	64	67
Total	75	75	75	75
Crabbing
1–4	15	9	6	4
5–8	2	0	0	0
≥ 9	1	1	0	0
Missing	3	1	0	1
Not applicablea	54	64	69	70
Total	75	75	75	75
Boating
1–4	7	26	4	2
5–8	2	0	0	0
≥ 9	0	1	0	0
Missing	0	5	0	0
Not applicablea	66	43	71	73
Total	75	75	75	75
Swimming
1–4	5	6	14	2
5–8	1	1	0	0
≥ 9	0	0	0	0
Missing	1	1	2	0
Not applicablea	68	67	59	63
Total	75	75	75	65

Note. These 4 sites were the most frequently used by the 75 patients reporting recreational water contact.

^aParticipant did not engage in this activity at this site.

DISCUSSION

We found a high level of self-reported participation in recreational water activity among HIV/AIDS patients at an urban Baltimore outpatient clinic. Respondents reported activities such as fishing, crabbing, boating, and swimming, as well as consumption of their own catch and wild-caught fish and crabs supplied by friends or family members. Our study sample was representative of the larger clinic population; therefore, our findings could be generalized to the larger Moore Clinic patient population, despite our small sample size. These findings are relevant to public health in light of recent findings of viable Cryptosporidium in many of the waters commonly used by our respondents and others in the region for recreational water activities.^7,67,73

To our knowledge, ours is the first study to describe recreational water contact among persons with HIV/AIDS. Another strength of our study was an excellent participation rate of more than 90% of clinic patients who were approached for participation.

Limitations

Our study had several limitations. Respondents were asked to recall recreational activities, including locations, activities, frequency, and specific months of participation, a method of data collection with a potential for recall bias. Some participants could not describe or identify the specific sites they used for recreational activities. Efforts were made to fill in these data gaps with a Baltimore County watershed map during the interview process.

Interviews took place in a clinical setting where participants often had limited time to complete surveys. Patients were free to leave at any point during the interview to go to their appointments; most returned to complete their surveys. Some incomplete data collection and nonresponses resulted from time restraints on both patients and interviewers (who could not always review all questions or double-check responses).

Language barriers may have prevented enrollment and participation of patients in the study. However, fewer than 5 patients were identified as having a language barrier. Some respondents may have completed multiple surveys during the course of the study period, because no personal identifiers were collected. However, efforts were made to limit participation to persons who reported not previously completing our survey, and we used an exemption form to exclude these persons.

Conclusions

We found that persons with HIV/AIDS, a subset of the urban population, frequently take part in recreational water activities in the Baltimore area. Unfortunately, this watershed is contaminated by Cryptosporidium and other enteric pathogens of concern for both healthy and immunocompromised individuals.

This population is engaging in fishing and crabbing not only for fun, sport, or recreation but also for diet supplementation and subsistence. Contact with fish and crabs during preparation and handling may expose individuals to Cryptosporidium via contamination of hands, preparation and cooking surfaces, and equipment. In addition, exposure may occur from consumption of improperly cooked fish and crabs. These recreational sites lack hand-washing facilities, a concern because food and beverages are commonly consumed there.

Pathogens such as Cryptosporidium are not routinely monitored in surface waters and are not included in existing fish advisories. As a result, the general public, including immunocompromised and other highly susceptible persons, are not aware of the risks of exposure and infection posed by pathogens. This lack of communication is a major public health failure, not limited to but of particular concern for persons who are immunocompromised. Regulatory agencies, watershed associations, and others involved in recreational water programs should inform themselves about the effect of infectious pathogens in recreational waters and develop routine surveillance, monitoring, and notification systems to ensure the safety of urban waters.

Physicians and other caregivers for persons with HIV/AIDS should incorporate questions about recreational water activities, foreign travel, and other activities with potential for pathogen exposure during routine intake procedures and follow-up of patients with acute or chronic gastroenteritis. In addition, patients who are at high risk for noncompliance with antiretroviral therapy regimens or who are not receiving these therapies should be counseled on the risk factors for exposure and infection by Cryptosporidium.

Future national health surveys should include questions about potential sources of exposure to Cryptosporidium, as well as about residence, foreign travel, and recreational water activities (including chlorinated water venues such as swimming pools) to accurately characterize the populations at risk.