Microsporidiosis and Malnutrition in Children with Persistent Diarrhea, Uganda

Microsporidiosis by using immuno? uorescence microscopy, with con? rma-
and Malnutrition tion and genotyping subsequently determined by PCR–re-
striction fragment length polymorphism analysis (10). HIV
in Children status was determined by using established methods, and
children positive for HIV were referred to the Mulago Hos-
with Persistent pital Pediatric Infectious Disease Clinic for further care.
The study population and results of primary analysis are
Diarrhea, Uganda described in more detail elsewhere (6).
A complete set of anthropometric measures (age,
weight, height, weight-for-age z-score, height-for-age z-
Siobhan M. Mor, James K. Tumwine,
score, and weight-for-height z-score) was available for
Elena N. Naumova, Grace Ndeezi,
224 children. Wasting was twice as likely in children with
and Saul Tzipori
microsporidiosis than in children without the infection
(Table 1). Microsporidiosis was strongly associated with
We show that the microsporidian fungus Enterocyto-
HIV and concurrent cryptosporidiosis. These infections
zoon bieneusi is associated with lower rates of weight gain
likely compound the poor nutritional status of children
in children in Uganda with persistent diarrhea. This relation-
with microsporidiosis, although this assessment is limited
ship remained after controlling for HIV and concurrent cryp-
tosporidiosis. Children with microsporidiosis were predicted
by sample size (Figure 1).
to weigh 1.3 kg less than children without microsporidiosis
Linear regression models were used to describe weight
at 5 years of age.
and height gain in study children. Weight and height were
treated as continuous dependent variables. When regressed
on age, the slopes of the curves represent the rate of weight
Enterocytozoon bieneusi is an important cause of persis- and height gain, respectively. All variables were trans-
tent diarrhea, intestinal malabsorption, and wasting in formed to the natural logarithmic scale.
HIV-positive adults. Mucosal damage associated with mi-
For initial data exploration, several simple models
crosporidiosis is more extensive than that related to other were applied with children strati? ed according to HIV
opportunistic intestinal infections (1,2) and leads to sub-
and Cryptosporidium spp. status. Within each strata, slope
stantial malabsorption of carbohydrates, fat, and essential parameters were compared by using t tests to identify dif-
nutrients (2–5). Although microsporidiosis is common in ferences in growth rates between E. bieneusi–positive and
children <5 years of age, particularly those who live in de-
–negative children. Growth rates were reduced in children
veloping countries (6,7) or who are HIV positive (6,8,9), with microsporidiosis across all HIV and Cryptosporid-
the effects of infection on nutritional health of these vulner-
ium strata (Table 2). In HIV-positive children, the rate
able populations are not well documented. We reexamined of weight gain was lower in children with microsporidi-
anthropometric data of children in Uganda with persistent osis than in those without microsporidiosis (model 1a vs.
diarrhea (6) and used regression analysis to determine model 1b), and some evidence showed that this was also
whether there is an association between microsporidiosis true for HIV-negative children (model 1c vs. model 1d).
and reduced growth rates.
In children concurrently infected with Cryptosporidium
spp., rates of weight (model 1e vs. model 1f) and height
The Study
gain (model 2e vs. model 2f) were lower in children with
A total of 243 children <60 months of age with persis-
microsporidiosis.
tent diarrhea (>14 days) were enrolled at Mulago Hospital
Adjusted growth rate estimates were obtained by ? t-
in Kampala, Uganda, from November 2002 through May ting a multiple linear regression model that controlled for
2003. After informed consent was obtained, demographic, the effect of sex, HIV status, and concurrent cryptosporidi-
anthropometric, and clinical information was collected osis. The independent variable of interest was an interac-
from each child. E. bieneusi spores were detected in stool tion term between E. bieneusi and age, which re? ected the
specimens by using a nested PCR with E. bieneusi–spe-
difference in the growth rates of children with and without
ci? c primers (7). Cryptosporidium oocysts were detected microsporidiosis. Interaction terms between E. bieneusi,
Author af? liations: Tufts Cummings School of Veterinary Medicine,
Cryptosporidium spp., HIV, and age were also explored but
North Grafton, Massachusetts, USA (S.M. Mor, S. Tzipori); Tufts
were excluded from the ? nal model because they did not
University School of Medicine, Boston, Massachusetts, USA (S.M.
improve model ? t. When we simultaneously adjusted for
Mor, E.N. Naumova); and Makerere University Medical School,
sex, HIV status, and concurrent cryptosporidiosis, rate of
Kampala, Uganda (J.K. Tumwine, G. Ndeezi)
weight gain remained signi? cantly lower in children with
microsporidiosis (p = 0.014). However, rate of height gain
DOI: 10.3201/eid1501.071536

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 15, No. 1, January 2009
49

---

DISPATCHES
Table 1. Clinical features of 224 children with persistent diarrhea with and without cryptosporidisis, Uganda*
Microsporidiosis
Feature Total
Yes No
Crude OR (95% CI)
p value†
No. patients
224
68
156


Age category, mo





<6, no. (%)
32 (14.3)
5 (7.4)
27 (17.3)
1.0

7–12, no. (%)
108 (48.2)
31 (45.6)
77 (49.4)
2.2 (0.8–6.2)
0.137
13–24, no. (%)
69 (30.8)
24 (35.3)
45 (28.8)
2.9 (1.0–8.4)
0.048
>25, no. (%)
14 (6.3)
8 (11.8)
6 (3.8)
7.2 (1.7–29.9)
0.004
Female sex, no. (%)
89 (39.7)
29 (42.6)
60 (38.5)
0.8 (0.5–1.5)
0.556
Nutritional
status‡


Mean WHZ (SD)
–1.44 (1.79)
–1.76 (1.83)
–1.30 (1.76)

0.077
Mean WAZ (SD)
–2.61 (1.41)
–2.76 (1.60)
–2.55 (1.32)

0.356
Mean HAZ (SD)
–2.16 (1.77)
–2.10 (1.90)
–2.19 (1.73)

0.750
Wasted, no. (%)
94 (42.0)
37 (54.4)
57 (36.5)
2.1 (1.2–3.7)
0.013
Underweight, no. (%)
148 (66.1)
47 (69.1)
101 (64.7)
1.2 (0.7–2.2)
0.525
Stunted, no. (%)
121 (54.0)
39 (57.4)
82 (52.6)
1.2 (0.7–2.2)
0.508
Concurrent cryptosporidiosis, no. (%)
63 (28.1)
55 (80.9)
8 (5.1)
78.3 (30.8–199.1)
<0.001
HIV+, no. (%)
77 (34.4)
58 (85.3)
19 (12.2)
41.8 (18.3–95.4)
<0.001
HIV+ and cryptosporidiosis, no. (%)
54 (24.1)
51 (75.0)
3 (1.9)
153.0 (43.1–543.5)
<0.001
*OR, odds ratio; CI, confidence interval.
†Continuous variables compared by using 2-sided t test. Categorical variables were compared by using Pearson 2 test.
‡WHZ, weight-for-height z-score; WAZ, weight-for-age z-score; HAZ, height-for-age z-score. Children were considered wasted, underweight, or stunted if
WHZ, WAZ, or HAZ were <–2.0, respectively.
was not signi? cantly different between children with and needed to document decelerating linear growth. In contrast,
without microsporidiosis (p = 0.151). Predicted weight-for-
weight-for-age re? ects chronic and acute nutritional states.
age growth curves are shown in Figure 2, which also dis-
Because all study participants had a history of persistent
plays reference curves for healthy Ugandan children (11). diarrhea and stool was generally collected on the day of
The growth trajectory of children with microsporidiosis hospitalization, nosocomial acquisition of microsporidiosis
was such that by age 5, these children were predicted to was unlikely.
weigh ?1.3 kg less than children without microsporidiosis.
There are several limitations to this analysis. The caus-
This ? nding exceeded the predicted difference in weight in al role of E. bieneusi in childhood malnutrition cannot be
children with and without HIV (0.74 kg) at the same age.
inferred because of the retrospective and cross-sectional
study design. Although intestinal infection in children may
Conclusions
impair absorption of nutrients (as documented in adults),
Given consistent reports of severe wasting and mal-
nutrition in HIV-positive adults with microsporidiosis, it
Weight-for-height
is conceivable that E. bieneusi infection early in life may
6
Weight-for-age
result in malnutrition. Two cross-sectional studies attempt-
Height-for-age
4
ed to correlate microsporidiosis with poor anthropometric
2
status but did not ? nd a signi? cant association (7,12). How-
ever, the dichotomous method used for these assessments
0
-
score
z
is sensitive to the choice of cut-off values. Although chil-
– 2
dren who fall >2 SDs below the reference growth curves
– 4
are conventionally categorized as malnourished, this cut-
– 6
off does not denote a biologically meaningful distinction
between healthy and malnourished children.
Microsporidiosis
+
+
+
+
Using regression analysis, we present evidence that Cryptosporidiosis –
+
–
+
HIV
–
–
+
+
microsporidiosis is associated with growth faltering in
n
6
4
7
51
children in Uganda. In this approach, anthropometric Figure 1. Nutritional status of children in Uganda with microsporidiosis
data were treated as continuous variables, thus avoid-
with or without HIV or concurrent cryptosporidiosis. The infection
ing arbitrary categorization of children into malnourished status and number of children in each infection category are shown.
and normally nourished groups. Lack of evidence for an Outliers are indicated with dots (1.5–3 interquartile ranges [IQRs])
or the asterisk (>3 IQR). The dashed horizontal line indicates
independent effect of microsporidiosis on linear growth conventional cut-off for malnutrition (z-score <2), horizontal lines in
might be explained by the fact that these children were cur-
each column indicate the median score, and error bars indicate the
rently infected with E. bieneusi, and longer intervals are highest and lowest z-scores excluding outliers.
50
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 15, No. 1, January 2009

---

Microsporidiosis and Malnutrition in Children
Table 2. Rate of weight (model 1) and height gain (model 2) in children with and without microsporidiosis when stratified by HIV
and Cryptosporidium infection, Uganda*
Infection status
Measurement at age:
Model
No.
Micro
Crypto
HIV
12 mo
36 mo
60 mo
R2
Rate (95% CI)
Difference†
p value‡
1a
58
+
+
6.63
8.46
9.48
0.27
0.222 (0.13–0.32)
0.147
0.016
1b
19
–
+
6.69
10.03
12.11
0.73
0.369 (0.25–0.49)
1c
10
+
–
7.60
8.77
9.37
0.28
0.130 ( 0.04–0.30)
0.177
0.056
1d
137
–
–
6.92
9.70
11.35
0.33
0.307 (0.23–0.38)
1e
55
+
+
6.65
7.98
8.69
0.17
0.166 (0.07–0.27)
0.338
<0.001
1f
8
–
+
6.50
11.30
14.62
0.82
0.504 (0.27–0.74)
1g
13
+
–
7.40
9.50
10.67
0.51
0.227 (0.08–0.37)
0.081
0.336
1h
148
–
–
6.91
9.69
11.34
0.39
0.308 (0.25–0.37)
2a
58
+
+
69.55
82.83
89.84
0.63
0.159 (0.13–0.19)
0.035
0.123
2b
19
–
+
70.46
87.20
96.28
0.82
0.194 (0.15–0.24)
2c
10
+
–
71.25
81.47
86.70
0.71
0.122 (0.06–0.19)
0.040
0.206
2d
137
–
–
69.17
82.64
89.77
0.55
0.162 (0.14–0.19)
2e
55
+
+
69.49
82.03
88.61
0.59
0.151 (0.12–0.19)
0.085
0.009
2f
8
–
+
68.75
89.10
100.51
0.81
0.236 (0.12–0.35)
2g
13
+
–
71.49
82.73
88.55
0.75
0.133 (0.08–0.18)
0.034
0.240
2h
148
–
–
69.41
83.38
90.81
0.61
0.167 (0.15–0.19)
*For ease of interpretation, predicted measurements at various ages are shown in addition to the rate of growth in ln(kg or cm)/ln(months). Micro,
microsporidiosis; Crypto, cryptosporidiosis; CI, confidence interval.
†Rate of growth of children with microsporidiosis minus rate of growth of children without microsporidiosis in ln(kg or cm)/ln(months).
‡By t test comparing rate of growth of children with and without microsporidiosis.
malnourished children may also have immune defects that of reduced weight gain is followed by catch-up growth.
predispose them to E. bieneusi infection. Because the PCR Because microsporidiosis is highly prevalent in children
was speci? c for E. bieneusi, we cannot rule out that some in developing countries, the ? nding that the infection has
children were infected with other microsporidian species.
However, E. bieneusi is the more common of 2 species
16
known to cause intestinal microsporidiosis (13,14). In pre-
vious studies at Mulago Hospital, 16.8% of children with
14
acute diarrhea and 16.8% of children without diarrhea had
microsporidiosis (7). Because all children in the current
12
study had persistent diarrhea, direct comparison between
groups was possible without the need to control for diarrhea
10
status. However, we cannot comment on the effect of acute
or subclinical infection on nutritional health. Residual con-
8
founding may exist through sociodemographic factors not
accounted for in the analysis. Sociodemographic data were
Weight, kg
6
limited to the accessibility of safe drinking water and type
of sanitary facility in the household, neither of which were
4
Healthy children
associated with microsporidiosis (data not shown). Finally,
HIV– children without microsporidiosis
use of cross-sectional data is a major limitation because
2
HIV+ children without microsporidiosis
measurements obtained at a single point in time do not cap-
HIV– children with microsporidiosis
HIV+ children with microsporidiosis
ture individual growth trends. Such data make it dif? cult to
0
assess the effect of a particular episode of illness on growth
0
12
24
36
48
60
attainment (15). To this extent, longitudinal anthropometric
Age, mo
assessment is the only means of detecting growth faltering
that results from
Figure 2. Weight-for-age growth curves of study children (as
E. bieneusi infection in childhood.
modeled by multiple linear regression) and reference populations
Although our results suggest an association between in Uganda (11). Curves represent the median weight-for-age,
reduced weight gain and microsporidiosis, further studies averaged between boys and girls and controlling for concurrent
are required to determine the role of E. bieneusi in child-
Cryptosporidium spp. infection. The difference, 95% con? dence
hood malnutrition. Longitudinal studies enabling compari-
interval, and signi?
cance of the interaction term between
son of preinfection and postinfection weights in individual Enterocytozoon bieneusi and age re? ect the difference in growth
rates of children with and without microsporidiosis in ln(kg)/ln(age).
children are needed to establish the direction of causation. R2 = 0.42, difference = –0.133, 95% con? dence interval –0.23 to
It will be particularly useful to identify whether the period –0.03, p = 0.009.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 15, No. 1, January 2009
51

---

DISPATCHES
a lasting effect on growth would highlight the importance 7. Tumwine JK, Kekitiinwa A, Nabukeera N, Akiyoshi DE, Buckholt
of nutritional rehabilitation and provide impetus to develop
MA, Tzipori S. Enterocytozoon bieneusi among children with diar-
therapeutics suitable for use in young children.
rhea attending Mulago Hospital in Uganda. Am J Trop Med Hyg.
2002;67:299–303.
8. Leelayoova S, Vithayasai N, Watanaveeradej V, Chotpitayasunondh
Acknowledgments
T, Therapong V, Naaglor T, et al. Intestinal microsporidiosis in HIV-
We thank the members of the Initiative for the Forecast-
infected children with acute and chronic diarrhea. Southeast Asian J
Trop Med Public Health. 2001;32:33–7.
ing and Modeling of Infectious Diseases at Tufts University and 9. Wanachiwanawin D, Chokephaibulkit K, Lertlaituan P, On-
anonymous reviewers for providing constructive comments on
grotchanakun J, Chinabut P, Thakerngpol K. Intestinal microsporidi-
the manuscript.
osis in HIV-infected children with diarrhea. Southeast Asian J Trop
Med Public Health. 2002;33:241–5.
This study was supported by grant R21A1068474 from the 10. Tumwine JK, Kekitiinwa A, Nabukeera N, Akiyoshi DE, Rich SM,
National Institutes for Allergy and Infectious Diseases (S.M.M
Widmer G, et al. Cryptosporidium parvum in children with diar-
rhea in Mulago Hospital, Kampala, Uganda. Am J Trop Med Hyg.
and S.T.) and by the University of Sydney Traveling Scholarship
2003;68:710–5.
Fund (S.M.M.).
11. Cortinovis I, Vella V, Ndiku N, Milani S. Weight, height and
arm circumference of children under 5 in the district of Mba-
Dr Mor is a veterinarian and doctoral candidate at the Cum-
rara, south-west Uganda. Ann Hum Biol. 1997;24:557–68. DOI:
mings School of Veterinary Medicine at Tufts University. She is
10.1080/03014469700005322
also an instructor at the School of Medicine at the same univer-
12. Leelayoova S, Subrungruang I, Rangsin R, Chavalitshewinkoon-
sity. Her research interests include the clinical epidemiology of
Petmitr P, Worapong J, Naaglor T, et al. Transmission of Enterocyto-
zoon bieneusi genotype a in a Thai orphanage. Am J Trop Med Hyg.
E. bieneusi, Cryptosporidium spp., other tropical infections, and
2005;73:104–7.
zoonotic diseases of global health importance.
13. Samie A, Obi CL, Tzipori S, Weiss LM, Guerrant RL. Microsporidi-
osis in South Africa: PCR detection in stool samples of HIV-positive
and HIV-negative individuals and school children in Vhembe district,
References
Limpopo Province. Trans R Soc Trop Med Hyg. 2007;101:547–54.
DOI: 10.1016/j.trstmh.2007.02.005
1. Schmidt W, Schneider T, Heise W, Schulzke JD, Weinke T, Igna-
14. Espern A, Morio F, Miegeville M, Illa H, Abdoulaye M, Meysson-
tius R, et al. Mucosal abnormalities in microsporidiosis. AIDS.
nier V, et al. Molecular study of microsporidiosis due to Entero-
1997;11:1589–94. DOI: 10.1097/00002030-199713000-00007
cytozoon bieneusi and Encephalitozoon intestinalis among human
2. Kotler DP, Francisco A, Clayton F, Scholes JV, Orenstein JM. Small
immunode? ciency virus–infected patients from two geographi-
intestinal injury and parasitic diseases in AIDS. Ann Intern Med.
cal areas: Niamey, Niger, and Hanoi, Vietnam. J Clin Microbiol.
1990;113:444–9.
2007;45:2999–3002. DOI: 10.1128/JCM.00684-07
3. Molina JM, Sarfati C, Beauvais B, Lemann M, Lesourd A, Ferchal F,
15. McMurray C. Cross-sectional anthropometry: what can it tell us
et al. Intestinal microsporidiosis in human immunode? ciency virus-
about the health of young children? Health Transit Rev. 1996;6:147–
infected patients with chronic unexplained diarrhea: prevalence and
68.
clinical and biologic features. J Infect Dis. 1993;167:217–21.
4. Lambl BB, Federman M, Pleskow D, Wanke CA. Malabsorption Address for correspondence: Siobhan M. Mor, Division of Infectious
and wasting in AIDS patients with microsporidia and pathogen-neg-
Diseases, Tufts Cummings School of Veterinary Medicine, Bldg 20, 200
ative diarrhea. AIDS. 1996;10:739–44. DOI: 10.1097/00002030-
199606001-00007
Westborough Rd, North Grafton, MA 01536, USA; email: siobhan.mor@
5. Asmuth DM, DeGirolami PC, Federman M, Ezratty CR, Pleskow
tufts.edu
DK, Desai G, et al. Clinical features of microsporidiosis in patients
with AIDS. Clin Infect Dis. 1994;18:819–25.
6. Tumwine JK, Kekitiinwa A, Bakeera-Kitaka S, Ndeezi G, Downing
The opinions expressed by authors contributing to this journal do
R, Feng X, et al. Cryptosporidiosis and microsporidiosis in Ugandan
not necessarily re? ect the opinions of the Centers for Disease Con-
children with persistent diarrhea with and without concurrent infec-
trol and Prevention or the institutions with which the authors are
tion with the human immunode? ciency virus. Am J Trop Med Hyg.
af? liated.
2005;73:921–5.
52
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 15, No. 1, January 2009

---