JOURNAL OF TROPICAL PEDIATRICS, VOL. 59, NO.

6, 2013

Toddler’s diarrhea: Is it an under-recognized entity in

developing countries?
by Ujjal Poddar, Jaya Agarwal, Surender Kumar Yachha, and Anshu Srivastava
Department of Pediatric Gastroenterology, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India

Correspondence: Ujjal Poddar, Department of Pediatric Gastroenterology, SGPGIMS, Lucknow-226014, India. Tel: þ91 522
2494418. Fax: þ91 522 2668017. E-mail <[email protected]>.

Downloaded from https://academic.oup.com/tropej/article-abstract/59/6/470/1659951 by guest on 09 July 2019

Summary
Aim: As there is no report of toddler’s diarrhea (TD) from the developing world, we have analyzed our
experience of 191 children (<5 years) with chronic diarrhea over 7 years.
Methods: Clinical details, investigations and outcome were retrieved and recorded in a proforma. TD
was ascribed in those who had normal growth and no evidence of malabsorption or enteric infections.
Results: The etiology of chronic diarrhea was TD 16%, celiac disease (CD) 37%, cow’s milk protein allergy
(CMPA) 35% and others 12%. The mean age of 30 children with TD was 2.7  1.2 years (22 boys). While
comparing TD with CD and CMPA, weight z score (>2.2) and hemoglobin (>10 or 10.9 g/dL for CD and
CMPA, respectively) were independent variables to discriminate TD from CD and CMPA.
Conclusions: TD is common in developing countries like India too. Anthropometry and hemoglobin
can differentiate it from CD and CMPA.

Key words: butter, celiac disease, cow’s milk protein allergy, non-specific diarrhea

Introduction especially in developing countries like India, where

Toddler’s diarrhea (TD) or chronic non-specific diar- this entity has not been reported before.
rhea of childhood is the commonest cause of chronic In this study, we have analyzed our experience of
diarrhea without failure to thrive in preschool chil- 191 children aged <5 years with chronic diarrhea to
dren of the Western world [1, 2]. On the other hand, find out the prevalence of TD in our setup.
in developing countries, malnutrition, gastrointes-
tinal infections and infestations are believed to be Methods
the common problems than TD [3–6]. However, the This study was carried out in the Department of
picture is changing over the years. Many diseases like Pediatric Gastroenterology at the Sanjay Gandhi
Crohn’s disease [7], cow’s milk protein allergy Postgraduate Institute of Medical Sciences,
(CMPA) [8], celiac disease (CD) [9–11], non-alcoholic Lucknow, India. We analyzed our experience of
fatty liver disease and others, which were believed to chronic diarrhea (>3 stools/day of >4 weeks’ dur-
be the diseases of the West, have been reported from ation) in children aged <5 years from January 2004
India. As India is in the transition phase from de- to December 2010. In our department, we follow
veloping to developed world, we have started getting a standard protocol for investigation of chronic
diseases related to affluence. So far, there is no report diarrhea cases. Clinical details, including anthropom-
of TD from India or from any developing country. etry, were retrieved and recorded in a proforma.
TD was first described as a separate clinical entity Children were segregated into two groups, those
by Cohlan in 1956 [12], but in 1966, Davidson with history of bloody diarrhea and those without a
and Wasserman [13] gave the consistent diagnostic history of bloody diarrhea, for the purpose of
criteria, further characterizing this entity. investigation.
Traditionally, TD is diagnosed clinically as chronic As per our investigation protocol, complete blood
diarrhea without failure to thrive and malabsorption count and stool routine examination (for ova,
[14], but recently, questions have been raised about parasites and cysts) were done in all children. Serum
this clinical diagnosis, as it potentially stops the clin- immunoglobulin A–anti-endomysial antibody, eso-
ician from thinking about the possible causes of loose phagogastroduodenoscopy (OGD) and duodenal
stools in this subset of patients [15]. Hence, common biopsy were done in children without bloody diarrhea,
causes of chronic diarrhea in this age-group need to and proctosigmoidoscopy and rectal biopsy with or
be ruled out before putting a label of TD on a child, without OGD and duodenal biopsy in children with

ß The Author [2013]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 470
doi:10.1093/tropej/fmt055 Advance Access published on 27 June 2013
U. PODDAR ET AL.

bloody diarrhea. Specific investigations like serum im- 3 months and then 3 monthly), effect of dietary ma-
munoglobulins, T-cell functions and others were done nipulation was assessed in terms of stool frequency
on an as-and-when-required basis. (no diarrhea means 3 stools/day). Informed consent
CD was diagnosed on the basis of modified was taken from parents before endoscopy and intes-
European Society of Pediatric Gastroenterology, tinal biopsy.
Hepatology and Nutritional (ESPGHAN) criteria
[16], i.e. characteristic histopathologic changes in Statistics
duodenal biopsy and unequivocal clinical response Statistical analysis was done using SPSS 17.0 soft-
to gluten-free diet. CMPA was diagnosed on the ware (IBM, USA). Continuous variables were ex-
basis of previously described criteria [8, 17], i.e. endo- pressed as mean  standard deviation. The standard
scopic duodenal biopsy showing partial villous atro- deviation score (z score) was calculated for weight

Downloaded from https://academic.oup.com/tropej/article-abstract/59/6/470/1659951 by guest on 09 July 2019

phy with excessive eosinophilic infiltrates in lamina and height by using software Epi Info (version
propria or sigmoidoscopy showing aphthous ulcers 3.5.1; CDC, Atlanta, GA). Univariate and multivari-
and rectal biopsy showing focal eosinophilia (>6 eo- ate analyses were done for comparison of TD cases
sinophils/ high power field) in lamina propria, clin- with CD and CMPA cases.
ical and histological response to milk withdrawal and
histological deterioration within 48 hours of milk ex- Results
posure (after 6–8 weeks of milk-free period). During the study period, 191 children with chronic
Diagnosis of TD was ascribed in those who had diarrhea in the age-group of <5 years were seen.
normal pattern of growth and development, and no Their mean age was 2.66  1.30 years and male-to-
evidence of malabsorption or enteric infections in the female ratio was 129:62. Etiological spectrum is
age-group of 6 months to 5 years [13, 14]. shown in Fig. 1. Of these, 30 (16%) were diagnosed
As per the treating physician’s preference, a subset to have TD, 71 (37%) CD, 67 (35%) CMPA, 14 (7%)
of children with TD (n ¼ 16) were treated with sup- giardiasis and others 9 (5%; immunodeficiency in 6,
plemented fat diet (butter 2 g/kg/day), whereas others Crohn’s disease in 2 and intestinal lymphangiectasia
were given reassurance and dietary advice (to avoid in 1). Of the six immunodeficient children, three had
excess fruit juice and to have a balanced diet with congenital (common variable) and three had
adequate fat and fiber). On follow-up (monthly for acquired (HIV) immune deficiency.

Crohn's disease: 2(1%)

Immuno deficiency: 6(3%) Lymphangiectasia: 1(0.5%)

Giardia: 14(7%)
35%

16%

CMPA: 67
Toddler diarrhea: 30

Celiac disease: 71

37%
FIG. 1. Etiology of chronic diarrhea in children aged <5 years (n ¼ 191).

Journal of Tropical Pediatrics Vol. 59, No. 6 471

 U. PODDAR ET AL.

Of 30 children with TD, 22 were boys, with mean were found to be significant. We have constructed
age of 2.7  1.2 (range, 1–5) years. The mean dur- ROC (receiver operating characteristics) curves of
ation of diarrhea was 17  11 (range, 6–36) months. these two parameters for determining cutoff values
The age of onset of diarrhea was 15.5  9.6 (range, and found area under ROC to be 0.87 for weight
6–36) months. None had clinical features of malab- z score (Fig. 2A) and 0.88 for hemoglobin
sorption like significant pallor, edema or vitamin (Fig. 2B). We found that in a given case, weight
deficiencies. Anti-endomysial antibody test was z score of >2.28 had an odds ratio of 99.6 (95%
done in 29 and was negative in all. OGD and duo- confidence interval: 12.5–789.8) and hemoglobin
denal biopsy was done in all 30 cases and was normal >10 g/dL had an odds ratio of 92.11 (95% confidence
in all. In 22 cases, besides OGD, proctosigmoido- interval: 11.6–727.5) for predicting a case to be TD in
scopy and rectal biopsy was also done, which was comparison with that of CD. Among the clinical

Downloaded from https://academic.oup.com/tropej/article-abstract/59/6/470/1659951 by guest on 09 July 2019

found to be normal in all. Mild anemia (hemoglobin symptoms, abdominal distension as a complaint
between 10 and 11 g/dL) was detected in five chil- was seen more in children with CD (42/71) than in
dren, and all five had nutritional anemia (predomin- children with TD (1/30) (P ¼ 0.0001).
antly milk-fed), which was corrected with oral iron While comparing TD with CMPA, on univariate
therapy, and none had hypoalbuminemia. analysis, children with CMPA were significantly
The comparison of TD with two other common younger with shorter duration of symptoms and
causes of chronic diarrhea in this age-group, i.e. had significantly lower hemoglobin, albumin,
CD and CMPA, is given in Tables 1 and 2, respect- weight z scores and height z scores. However, on
ively. On univariate analysis, children with CD were multivariate analysis, of those six parameters, low
significantly older, had significantly lower hemoglo- weight z score and low hemoglobin were found to
bin and albumin values along with lower weight and be independent predictors for CMPA. We have con-
height z scores in comparison with children with TD. structed ROC curves for these two parameters for
However, on multivariate analysis, of those five par- determining cutoff values and found area under
ameters, only low weight z score and low hemoglobin ROC to be 0.73 for weight z score (Fig. 3A) and

TABLE 1
Comparison between toddler diarrhea and celiac disease

Parameters Toddler diarrhoea Celiac disease P value P value

(n ¼ 30) (n ¼ 71) (univariate analysis) (multivariate analysis)

Mean age of presentation(years) 2.7  1.2 3.4  1.1 0.004 0.220

Male:female 22:8 40:31 0.072 –
Duration of symptoms(months) 17  11 19.7  14 0.353 –
Mean weight z score (SD) 1.13  1.03 3.6  2.0 <0.0001 0.0001
Mean height z score (SD) 0.08  1.45 2.9  1.3 <0.0001 0.200
Mean (SD) hemoglobin value (g/dL) 11.4  1.06 9.0  1.7 <0.0001 <0.0001
Mean (SD) albumin value (g/dL) 4.11  0.46 3.5  0.8 0.001 0.520

TABLE 2
Comparison between toddler diarrhea and cow’s milk protein allergy

Parameters Toddler diarrhoea Cow milk allergy P value P value

(n ¼ 30) (n ¼ 67) (univariate (multivariate
analysis) analysis)

Mean (SD) age of presentation 2.7  1.2 1.7  0.8 0.0002 0.85
(years)
Male:female 22:8 46:21 0.810 –
Mean (SD) duration of symptoms 17  11 9.5  9.1 0.0008 0.24
(months)
Mean (SD) weight z score 1.13  1.03 2.4  1.6 <0.0001 0.04
Mean (SD) height z score 0.08  1.45 1.1  1.6 <0.0024 0.67
Mean (SD) hemoglobin value 11.4  1.06 10.1  1.5 <0.0001 0.001
(g/dL)
Mean(SD) albumin value (g/dL) 4.11  0.46 3.7  0.6 0.026 0.196

472 Journal of Tropical Pediatrics Vol. 59, No. 6

U. PODDAR ET AL.

0.73 for hemoglobin (Fig. 3B). We found that in a routine counseling and dietary advice (avoiding
given case, weight z score of >2.24 had an odds fruit juice and other high-osmotic drinks). During
ratio of 12.3 (95% CI: 2.7–56.2) and hemoglobin the follow-up, diarrhea subsided in 13 of 16 (80%)
>10.9 g/dL had an odds ratio of 13 (95% CI: children with added butter and 10 of 14 (71%) chil-
3.9–42.9) for predicting a case to be TD in com- dren with routine dietary advice (P ¼ ns).
parison with CMPA. Of all symptoms, blood in
stools as a complaint was seen in 26 of 67 (39%) Discussion
CMPA children and none in the TD group
(P ¼ 0.0001). This is first time we have documented that even in
The mean duration of follow-up was 4.8  4.7 developing countries like India, TD is not so uncom-
months (range, 1–16 months). Added fat (butter mon in preschool children. In fact, TD turned out to
be the third most common cause of chronic diarrhea

Downloaded from https://academic.oup.com/tropej/article-abstract/59/6/470/1659951 by guest on 09 July 2019

2 g/kg/day) was tried in 16 children as a dietary inter-
vention, and the remaining 14 were treated with after CD and CMPA in less than 5 years of age in

FIG. 2. (A) ROC curve for weight z score between TD and CD. (B) ROC curve for hemoglobin between TD
and CD.

FIG. 3. (A) ROC curve for weight z score between TD and CMPA. (B) ROC curve for hemoglobin between TD
and CMPA.

Journal of Tropical Pediatrics Vol. 59, No. 6 473

 U. PODDAR ET AL.

India. It has been shown that of all the causes of active role of definition, a recent article by Powell
chronic diarrhea in children, TD occurs most fre- and Jenkins [15] from the UK has raised an import-
quently in affluent societies [2]. In developed coun- ant issue of labeling a child with TD without any
tries, the diet of children has changed dramatically investigation, as it potentially stops the clinician
over the years. The change in food habits parallels to from thinking about other possible causes of chronic
the increase in affluence. Children of this era con- diarrhea. The concerns of Powell and Jenkins [15] are
sume more of carbohydrate-rich fluid (fruit juice, more important in developing countries, where this
squash, etc.) with less of fat and fiber [1] and that entity has not yet been described. Hence, despite ful-
can lead to TD. filling Rome II criteria, we went ahead and did rele-
It seems the spectrum of chronic diarrhea in India vant investigation to make the diagnosis of TD more
is changing with the improvement of economic objective. In our study, we have shown that weight

Downloaded from https://academic.oup.com/tropej/article-abstract/59/6/470/1659951 by guest on 09 July 2019

status. In the 1970s and 1980s, gastrointestinal infec- z score and hemoglobin level can easily differentiate
tions and infestations used to be the commonest TD from CD and CMPA. Hence, in a typical case of
cause (48–51%) [3, 4]. In the 1990s and 2000s, CD TD, even in developing countries, there is no need to
emerged as a frontrunner, especially in North India search for an organic disease if the weight z score
[9–11], and the latest entries into this list are CMPA is >2.2 and hemoglobin is >10 (for CD) or >10.9
[8] and Crohn’s disease [7]. So far, nobody has paid (for CMPA).
attention to TD as a distinct entity in India. With the
improvement in economic status and personal hy-
giene, the rates of gastrointestinal infections and in- Conclusions
festations are going down and diseases, known for TD is the third most common cause (16%) of chronic
long as Western diseases, have started increasing. diarrhea after CD and CMPA in children aged <5
Although we have not done a detailed analysis of years in India. Absence of failure to thrive and
socioeconomic status, as our hospital caters to the normal hemoglobin differentiate TD from other
middle and upper classes of the society, it is likely causes like CD and CMPA. Besides counseling of
that the dietary habits of this section of the society parents about benign self-limiting nature of the con-
resemble that of the Western society and that may be dition, added butter in the diet helps in rapid recov-
the reason for getting so many cases of TD in our ery in the majority of children with TD.
study.
In the pathogenesis of TD, disturbed intestinal mo- References
tility along with intake of excess fruit juice and less
1. Hoekstra JH. Toddler diarrhea: more a nutritional dis-
fat play an important role. Normally, small intestinal order than a disease. Arch Dis Child 1998;79:2–5.
motility is governed by the electrical activity called 2. Kneepkens CM, Hoekstra JH. Chronic nonspecific
migratory motor complex (MMC), and it has two diarrhea of childhood: pathophysiology and manage-
patterns: rapid movement in fasting state and slow ment. Pediatr Clin North Am 1996;43:375–90.
movement in postprandial state. In a classical study, 3. Gupte SP, Mehta S. Chronic diarrhea-an etiological
Fenton et al. [18] showed that intraduodenal instilla- study. Indian Pediatr 1970;7:625–27.
tion of glucose failed to interrupt the MMC in all 4. Walia BNS, Gupte SP, Mehta S, et al. Chronic diarrhea
eight children with TD but disrupted MMC in all in north Indian children. Indian J Med Res 1971;59:
four healthy children. The most powerful food that 1448–53.
has the capability of slowing down intestinal motility 5. Rastogi A, Malhotra V, Uppal B, et al. Aetiology of
chronic diarrhea in tropical children. Trop
is fat [19, 20]. In a study of 44 children with TD,
Gastroenterol 1999;20:45–9.
Cohen et al. [21] documented low fat in the diet 6. Mishra OP, Dhawan T, Singla PN, et al. Endoscopic
(<27% to total calories) in 29 (66%) children. All and histopathological evaluation of pre-school chil-
these children were put on diets containing at least dren with chronic diarrhea. J Trop Pediatr 2001;47:
4 g/kg/day of fat, and a dramatic response was seen 77–80.
in 82% of cases (remission of diarrhea within 3–14 7. Avinash B, Dutta AK, Chacko A. Pediatric inflamma-
days). We also tried added fat in the form of butter in tory bowel disease in south India. Indian Pediatr 2009;
almost half of our TD children and got similar results 46:639–40.
(80%). Hence, in addition to counseling about the 8. Poddar U, Yachha SK, Krishanani N, et al. Cow’s
benign and self-limiting course of the condition, milk protein allergy: an entity for recognition in
added butter to diet can be tried in TD. developing countries. J Gastroenterol Hepatol 2010;
25:178–82.
As per Rome II criteria [14], TD is renamed as
9. Yachha SK, Misra S, Malik A, et al. Spectrum of mal-
‘functional diarrhea’ and diagnostic criteria include absorption syndrome in north Indian children. Indian J
age of onset between 6 and 36 months and chronic Gastroenterol 1993;12:120–5.
diarrhea (>3 stools/day) of 4 weeks’ duration with no 10. Mohindra S, Yachha SK, Srivastava A, et al. Celiac
failure to thrive. All our patients with TD fulfilled disease in Indian children: assessment of clinical, nutri-
Rome II criteria. Although new diagnostic criteria of tional and pathological characteristics. J Health Popul
functional gastrointestinal disorders laid emphasis on Nutr 2001;19:204–8.

474 Journal of Tropical Pediatrics Vol. 59, No. 6

U. PODDAR ET AL.

11. Poddar U, Thapa BR, Nain CK, et al. Celiac disease in 17. Iyngkaran N, Robinson MJ, Prathap K, et al. Cow’s
India: are they true cases of celiac disease? J Pediatr milk-protein sensitive enteropathy: combined clinical
Gastroenterol Nutr 2002;35:508–12. and histopathological criteria for diagnosis. Arch Dis
12. Cohlan SQ. Chronic nonspecific diarrhea in infants Child 1978;53:20–6.
and children treated with diiodohydroxyquinoline. 18. Fenton TR, Harries JT, Milla PJ. Disordered small in-
Pediatrics 1956;18:424–32. testinal motility: a rational basis for toddlers’ diarrhea.
13. Davidson M, Wasserman R. The irritable colon of Gut 1983;24:897–903.
childhood (chronic non-specific diarrhea syndrome). 19. Rees WD, Malagelada JR, Miller JL, et al. Human
J Pediatr 1966;69:1027–38. interdigestive and postprandial gastrointestinal motor
14. Hyman PE, Milla PJ, Benninga MA, et al. Childhood and gastrointestinal hormone patterns. Dig Dis Sci
functional gastrointestinal disorders: neonate/toddler. 1982;27:321–9.
Gastroenterology 2006;130:1519–26. 20. Brown NJ, Read NW, Richardson A, et al.
15. Powell CVE, Jenkins HR. Toddler diarrhea: is it a Characteristics of lipid substances activating the ileal

Downloaded from https://academic.oup.com/tropej/article-abstract/59/6/470/1659951 by guest on 09 July 2019

useful diagnostic label? Arch Dis Child 2012;97:84–6. break in the rat. Gut 1990;31:1126–9.
16. Walker-Smith JA, Guandalini S, Schmitz J, et al. 21. Cohen SA, Hendricks KM, Mathis RK, et al. Chronic
Revised criteria for diagnosis of celiac disease. Arch non-specific diarrhea: dietary relationships. Pediatrics
Dis Child 1990;65:909–11. 1979;64:402–7.

Journal of Tropical Pediatrics Vol. 59, No. 6 475