Indian J Med Res 104, July 1996, pp 103-114

Epidemiology & management of persistent diarrhoea jn children of

developing countries

M.K. Bhan, N. Bhandari, S. Bhatnagar, & R. Bahl

ICMR Advance Centre for Diarrhoeal Disease Research, Department of Pediatrics

All India Instih:tte of Medical Science,S', New Delhi

Accepted June 24,1996

Diarrhoea that begins acutely but lasts longer than two weeks is defined to be persistent. Revised

estimates in developing countries including India showed that acute diarrhoea actOunts for 35 per

cent, dysentery 20 per cent and non-dysenteric persistent diarrhoea (PD) for 45 per cent of total

diarrhoeal deaths. PD a Iso often changes marginal malnutrition to more severe forms. Factors that

inere21se the risk of acute diarrhoea becoming persistent have been identified in India and other

developing countries. These include antecedent malnutrition, micronutrient def:iciency particularly

for zinc and vitamin A, transient impairment in cell mediated immunity, infection with entero

aggregative E.\'cherichia coli and cryptosporidium, sequential infection with different pathogens and

lack (if exclusive breast feeding during the initial four months of life particularly use of bovine milk.

Several issues regarding the management of pe~sistent diarrhoea in hospitalized children in India

have been resolved. Diets providing modest amounts of milk mixed with cereals are well tolerated.

In those who fail on such diets providing cari)ohydrate as a mixture of cereals and glucose or sucrose

hasten recovery. The role of antimicrobial agents and individual micronutrients in PD is currently

being investigated. A management algorithm appropriate for India and other developing countries has

been developed and found to substantially reduce case fatality in hospital settings to about 2-3 per

cent. Recent epidemiological and clinical research related to persistent diarrhoea is also reviewed.

Key words lmpaired immunity -incidence -lactQse intolerance -mortality micronutrient deficiency -persistent diarrhoea

(WHO) has now defined persistent diarrhoea as anepisode that begins acutely and lasts for at least 14

days I. The use of 14 days cut-off to partition acute

and persistent episodes seems justified. In India for

instance, in a longitudinal study the case fatalityrates were similar for diarrhoeal episodes of I and 2wk duration (0.61 and 0.8% respectively) and this

increased to 14 per cent as the duration exceeded t~oweeks7. The term persistent diarrhoea does not in-

clude specific disorders, such as hereditary syn-dromes, celiac disease or surgical conditions and

these are not considered in this reviewl.

Although tht~ majority of episodes of acute diar-rhoea are self limiting, a small proportion last sev-eral weeksl. These episodes persistent diarrhoea are

important because of a substantially higher casefatality rate th~Ln in acute diarrhoea, impact on nu-

tritional status and the fact that these adverse out-

comes may not be prevented by oral rehydration

therapy alonel-.'.

The distribu1ion of the duration of diarrhoeal epi-sodes is a con1inuum8. As such, delineating a sub-

group of diarrhoeaJ episodes as persistent by anycut-off is arbitrar:r'. The World Health Organization

103

104 INDIAN J MED IlliS, JUL y ] 996

Table I. Cause specific mortality rate* among children under five

years of age'

Type of diarrhoea No. of deaths Cause specific

mortaJity rate

5.6

2.1

2.8

4.9

5.6

8.5

Acute watery 8

Acute dysentery 3

Dysenteric PD 4

All dysentery 7

Non-dysentric PD 8

All persistent diarrhoea 12

* (number of deaths in a year/mid population) x 100

PD, Persistent diarrho~a

Superscript number indicates the reference no.

Incidence of persistent diarrhoea

Estimates indicate that 3-20 per cent of episodesof acute diarrhoea become persistentl. The incidenceof persistent diarrhoea obtained in severallongitudi-

nal studies varied considerably from 7-150 episodes

per l-oo child years5,9-14. In most settings the inci-dence of persistent diarrhoea: peaked between 7months and 2 yr of age; in countries Witll very high

diarrhoea attack rates, tile incidence was as high evenin the first six montlls of lifeI2.J4,J5. The disease inci-

dence declined rapidly after the fourth year. This

high variability in attack rates across countries maybe due to differences in disease definitions afld inten-

sity of household surveillance, apart from true geo-

graphical differences.

Mortality and growth faltering related to persis-tent diarrhoea

The colltribution of dehydration to persistent diar-rhoea associated deaths is uncertain but availabledata indicate that it is much less important .than ill

acute watery diarrhoea7,J4,!9.

Role of enteric pathogens in persistent diarrhoea

Several theoretical models for the role of micro-bial agents in the genesis of persistent diarrhoea canbe conceived8,2°. Firstly, an acute enteric iIlfection

may result ill non illfectious complications e.g., lac-tose illtolerance which leads to prolonged symptoms.

Secondly, persistent diarrhoea may occur as a resultof sequel1tial acute infectiolls with differel1t patho-

gens. Lastly, prolongation of diarrhoea may resultfrom persistence of intestinal infection; the failure to

eliminate the organisms from the il1testinal tract maybe related to the characteristics of the pathogen, to

host factors or to both.

Across various studies, one or more pathogellswere detected in half to two-thirds of children with

persistel1t diarrhoea during the initial few days ofillness2,5,13,14,21.22. The interpretation of these isola-tions is difficult because the pathogens were also

commonly excreted by asymptomatic children. Ingeneral, the pathogens excreted during persistentdiarrhoea were the same as those reported in acutediarrhoeal episodes with the exception of rotavirus,

whic\1 was often detected ill acute episodes, but rarelyill those that became persistel1t. To idel1tify el1teric

pathogens which may have a predilection for causing

prolonged diarrhoea, the approach commollly adoptedhas been to compare their initial excretion rates in

Recent research has led to a revision of the earlierestimates of the contribution of different types ofdiarrhoea to diarrhoea related mo1iality. In an inter-national study that compared clinical patterns of di-arrhoea among diarrhoea related deaths in India,

Bangladesh, Brazil and Senegal, a similar patternwas observed; acute diarrhoea accounted for 35 percent (range 25 to 46%), dysentery 20 percent (range

8 to 24%) and non-dysenteric persistent diarrhoea 45

per cent (range 23 to 62%) of the total diarrhoeal

deaths'4.16. The dysenteric deaths included those re-1ated to acute as well as persistent illness. In Incliathe cause specific mortaiity rates (per 1,000) in chil-

dren aged less than five years for acute watery diar-

rhoea5.6, dysentery4.9 and non dysenteric persistentdiarrhoea5.6 were similar (Table 1)7. While efforts to

promote oral rehydration therapy and antibiotic treat-ment of dysentery must be sustained, persistent diar-

rhoea needs greater attention than at present.

Antecedent malnutrition, lack of breast feeding,and associated systemic infection increase the risk of

death during persistent diarrhoea2.7.14.17.'9. In a com-

munity based study in north India, nearly a third offatal cases of persistent diarrhoea had associated

pneumonia7.

In Bangladesh, malnourished children had a 68fold higher risk of death from persistent diarrhoeathan those who were better nourished and residing inthe same communities'7,

106 INDIAN J MED RES, JULY 1996

Malnourished children are more likely to havesubclinical deficiency of individual micronutrientsand the deficiency of some of these may explain, at

least partly, the increased risk of persistent diarrhoeain the undernourished. In a recent Indian study, chil-

dren 6-36 months of age attending a communityclinic with acute diarrhoea were supplemented withzinc. gluconate or placebo during the enrolIment epi-sode and for a 6 months period thereafter; zinc supple-

mentation resulted in a significant reduction in the

average duration of the acute enrolIment episodesand in the proportion that became prolongedJS. Overthe ensuing 6 months the incidence of persistent

diarrhoea was reduced by 49 per cent in zinc supple-mented children older than] 1 months but there wasno such effect in those below this ageJ6. Zinc supple-

mentation was previously shown to reduce the dura-tion and severity of acute diarrhoea by othersJ7.J8.

A positive association between vitamin A defi-

ciency and diarrhoeal morbidity has been reported in

only some studies39.4o. In Brazil, routine administra-tion ofvitamin A (200,000 IV) at 4 monthly intervals

to children aged 1-5 yr reduced the incidence of

severe diarrhoea by 20 per cent and diarrhoea preva-lence by 23 per cent; the incidence of persistent

diarrhoea was not reported41. Micronutrients such aszinc and possibly vitamin A reduce the severity and

duration of diarrhoea either by a rapid and effectiverepair of the intestinal epithelium following an acuteenteric infection due to its role in the regulation ofcell division or by enhancing the immune re-

sponse42.43.

Recent morbidity and persistent diarrhoea

It has been seen that children are at an increased

risk of persistent diarrhoea following an episode of

measles or acute diarrhoea but not acute lower res-piratory tract infection49.so.The basis for an increasedrisk of persistent diarrhoea over the 2 to 3 months

period following an acute diarrhoeal episode is un-clear. A possible explanation may be that children

who develop persistent diarrhoea are a subgroup witha high overall diarrhoea burdensl; Alternatively, an

acute diarrhoeal episode may induce alterations ill

the intestinal epithelium or in immune responsive-ness that renders the subsequent diarrhoeal episodesto be more severe or prolonged. Acute diarrhoea is

characterized by excessive faecal zinc losses42. It hasbeen postulated that the impaired epithelial cell re-newal and immuno incompetence associated with

subclinical zinc deficiency may be the basis for theincreased susceptability to persistent diarrhoea fol-

lowing acute diarrhoea. These hypotheses need to beexamined in future studies.

Type of feeding and persistent diarrhoea

It is important to identify feeding practices that

increase the risk of persistent diarrhoea as these are

potentially modifiable.

Factors that increase the overall risk of diarrhoeaare also likely to cause more persistent diarrhoea. III, 131 ' I

1~ Partially breast-fed

D Not br~ast.,.fed9

7

51

**~II!...

c.I>

:;::;Oc.I~

3

Immuno incompetence, particularly decreased cellmediated immunity is one of the likely underlying

mechanisms for the increased incidence of persistentdiarrhoea among malnourished children3J,34,40,44. I nPeru and Bangladesh, however, reduced delayed hy-

persensitivity responses (DHR) to several skin anti-gens were found to be associated with an increasedrisk ofpersistent diarrhoea, even after controlling for

nutritional status31,34. The impairment in DHR wasoften transient. Transient cell mediated immunosup-

pression may be the result of acute viral infections45,46.Anergy has also been reported following bacterial

infections such as tuberculosis and pneumonia47.48.Diminished cell mediated immunity could also berelated to decline in micronutrient status42.

J -~- .'--- .c JPersistent diarrhoea Hosp)talization

Fig. I. Relative risk of persistent and severe diarrhoea amonginfants under 3 months ofage by feeding mode* (Ref) 8)*at the ageof I wk; **reference group: exclusively breast-fed.

107BHAN et at : PERSISTENT DIARRHOEA IN CHILDREN

Brazil, breast feeding was protective against persis-tent diarrhoea even after controllincr for socio-eco-bnomic status18. In early infancy, use of milk supple-ments in addition to breast feeding increased the risk

of persistent diarrhoea by three fold. The risk was

even greater among the non breast-fed (Fig. I ).

Once an acute diarrhoeal illness has occurred, the

mode of feeding during the acute phase may influ-ence the severity and duration of symptoms. Breast

feeding during acute diarrhoea reduces the severity

and duration of the episode52.

A highly contentious issue is whetl1er the use of

bovine milk or infant formula during acute diarrhoeaincreases the episode duration and severity. Several

factors may influence the response to milk feeding

during acute diarrhoea; the source of milk, amount,type of processing, other foods consumed during theillness, severity of the infection and factors related to

the host53. The lack of consistency in the findings on

this issue across studies is therefore understandable.There is evidence that feeding of non human milk as

the sole or predominant nutrient source during acutediarrhoea may increase the episode severity alid du-

ration54. In many of the studies where milk intakeduring acute diarrhoea was moderate, or when milk

was offered as a part of a mixed diet containingcereals, a significant increase in the episode duration

or "Severity was not observed55,56. Lactose malabsorp-tion is dose dependent and it is likely that milk inmoderate amounts particularly when mixed with

other foods is well. tolerated by most children with

acute diarrhoea.

In an Indian study, tile risk of persistence of acutediarrhoea was more with bovine milk than with an

infant formula57. This seems plausible as bovine milkhas a higher lactose content and osmolarity and thesensitization capability of milk protein in the infantformula is reduced due to spray drying during the

manufacturing process.Fermented products of milk such as yoghurt have

been shown to be better acGepted by lactose intoler-ant subjects and by children with persistent diar-rhoea58: In a recently completed trial however, the

feeding of yoghurt instead of milk to malnourisl\edchildren during r.cute diarrhoea was not associatedwith a reduction in the episode duration or the risk

ofpersistent diarrhoea (Singh et al unpublished data).

When viewed together, the reported data indicate

that the use of rnixed diets including low to moderate

amounts of milk during acute diarrhoea promoteweight gain without increasing the episode duration.

Type oforal rehydration salts (ORS) solution and

persistent diarrhoeaRice-based ORS was shown to substantially re-

duce the stool output in cholera in comparison tostandard ORS, but its effect on stool output andepisode duration in non cholera acute diarrhoea has

been relatively smaI16°. In a study from Bangladesh,the use of rice ORS during acute diarrhoea was re-

ported to have reduced the proportion of episodes

that became persistent61.In recent clinical trials, where feeding was care-

fully standardized, there was no significant differ-ence in the outcome of episodes treated with rice-based ORS than with standard ORS61. Thus, in the

presence of optimal feeding, rice-based ORS duringacute diarrhoea is unlikely to substantially reduce the

risk of development of persistent diarrhoea.

In a recent multi-centre WHO sponsored clinical

trial includ ing 1 ndia, the effect of a reduced osmolar-ity ORS (224 mmol/litre) on tl1e stool output andepisode duration was examined in comparison to the

standard WHO ORS (311 mmol/litre). Stool output

was reduced by 30 percent and the average episodeduration by 22 per cent in the children tr.eated withreduced osmolarity ORS62. Data on the proportion of

episodes that became persistent were not repolied.

There are few community-based reports on the riskof development of persistent diarrhoea in relation toantibiotic use during the first week of diarrhoeal ill-

ness. In a study from India, antibiotic use during the

initial days ofillness was equally common in episodes

that were eventually classified as acute or persistent57.In anotl1er study, initial antibiotic treatment was sig-nificantly less common in the episodes that becamepersistent, raising the possibility of a beneficial ef-

fectl2. Neveliheless it is important to recognize that

unwarranted antibiotic treatment for acute .diarrhoea

could lead to Clostridium difficile associated pseudomembranous colitis and hasten the emergence of an-

tibiotic resistant pathogens.

Specific causes and their role in the pathogenesis

The common causes of persistent diarrhoea as

seen in developing countries are persistent infection

108INDIAN J MED RES, JUL y 1996

with one or more enteric pathogens, seqliential en-

teric infection, disaccharide and rarely monosaccha-ride malabsorption and dietary protein intoler-

ance5,]4.22,63-70, The hallmark of tIle disorder is persis-tent mucosal damage which may result from failure

to eliminate the causative agent or delayed and inef-

fective mllCOsal restoration. Protein energy malnutrj-tion and micronutrient deficiency in humans are

known to be associated with abnormalities in intes-tinal structure and function. Delayed regeneration ofthe epithelium with reduced crypt cell multiplicationand ineffective maturation of cells during their mi-

gration up the villi following an enteric infection hasbeen demonstrated in experimental malnlJtrition67,

Pathogenic bacteria cause mucosal damage and diar-rhoea through m llcosal effacement or invasion andaction of enterotoxins or cytotoxins; malnutrition inthe host prolongs the healing of1;l1-e injured mucosa,

tIle state ofmaJabsorption and diarrhoea. In a propor-

tion of cases, an immunological response to luminal,

bacterial or dietary antigens has also been proposedas the basis for gut mucosal damage in persistent

diarrhoea but the data are still inconclusive68.

Faltered growth in persistent diarrhoea is not onlythe result of nutrient malabsorption; inadequate di-

etary intake due to anorexia, a continuation offaulty

pre-illness feeding practices or as a response todiarrhoeal illness itself by the family or the physi-

cians, are all important.

Management

The initial step is to determine the appropriate

place for care. Patients with persistent diarrhoea re-

quire hospitalization in the presence of dehydration,associated systemic infection requiring intravenous

antibiotics or when anorexia is severe. Intravenousfluid therapy may be required initially when dehy-

dration is severe, to correct major electrolyte abnor-

malities or acidosis, and in extremely cachexic or

systemically infected infants who accept oral fluids

poorly.

C11t1dren with sOJ11-e but not severe dehydrationcan be effectively treated with oral rehydration salt

solutioJis. It is important to provide additional po-tassium supplements to those severely malnour-ished.

The role of dietary protein intolerance in persis-tent diarrhoea is highly contentious. Based on stud-ies in India and other developing countries, it is

unlikely to be of significance in tl1e pathogenesis of

persistent diarrhoea63-65.69.7o. Altllough, increased ill-testinal uptake of intact proteins has been demon-strated following acute gastroenteritis in early in-

fancy, the clinical significance of this phenomenon isnot established7°,

There is a substantial impairment of fat and pro-tein absorption during persistent diarrhoea; carbohy-

drate malabsorption is only moderately affected on

diets that are predominantly cereal based!J. Second-

ary malabsorption of sugars, particularly of lactose iscommon. In a proportion of cases, djsaccharides otherthan lactose are also malabsorbed to a clinically sig-

nificant degree.

Clinically significant monosaccha.ide malabsorp-tion is fortunately infrequent62,63.66. The causes of

carbohydrate malabsorption are acute or persistententeric infection and malnutrition by decreasing brushborder enzymes. Unabsorbed sugars are osmoticallyactive and induce water and electrotyte secretion intothe intestinal lumen.

When clinical dehydration is not associated, homeavailable fluids are appropriate for replacement of

ongoing stool losses.

Basis for dietary management

Several general principles have been establjshed

through recent studies in persistent diarrhoea.

In developjng countries, the need to use total in-

travenous nutritjon arises very rarely. Optimal oral

feeding, based on an appropriately constituted diet iswell tolerated and achieves recovery and catch up

growth in the vast majority of these pati-ents. Al-though there is some malabsorption of nutrients in

persistent diarrhoea, about 80-90 per cent of carbo-hydrates and 70- 75 per cent of fats and proteins are

actually absorbed from mjxed diets based on locallyavailable ingredients.

Breast feeding is safe and well tolerated during

diarrhoea. Although, a few predominantly breast-fedinfants with acute diarrhoea may continue to passstools with more tllan the usual frequency or stoolsof somewhat liquid consistency for more than two

weeks, physical growth is well maintained.

109BHAN et al : PERSISTENT DIARRHOEA IN CHILDREN

Should milk be withdrawn in persistent diar-

rhoea?Table III. Comparison of milk-based and milk-free diets

persistent diarrhoea73

Difference ir

median

(95% CI)

In the non breast-fed babies, an important issue is

whether milk should be totally eliminated or simply

reduced in amount during persistent diarrhoea.

Milk free

cereal

(n=46)

Milk

cereal

(n=47)

Median stool weight

in males'(g/kg/h)

0-48 h 6.23

(-0.26,0.73)

0.26

(-0.19.0.74)

0.91(-0.48. 2.4)

1.7

(0.9, 2.5)

1.5"

(0.6; 2.4)

130-120 h 1.6

(0.9.2.7)

% change in 2.8

weight at 120 h' (0.8.5.9) (-

Number with :

Stool output >200g/kg 3

in any 24 h period

Reappearance of 4

dehydration during study

Stool output >60g/kg 2

on day- 7

Weight on day-7

< rehydration weight

Treatment failures 17.2

by any of the above (%)

, Median (range)

" Odds ratio (95% CI)

Superscript no. indicate the reference no

(0.61.2.39)

2.30. 4.3)

Brown and colleagues reported increased stoolweights on diets ,predominantly based on whole milkas compared to lactose hydrolysed milk72. The milk

intakes were equivalent to about 6 g/kg/day or moreof lactose; few children in Indian communities takesuch large quantities of milk. The issue of whether

lower intakes equivalent to 2.5 g/kg lactose load per

day would also be poorly tolerated has been recently

examined in a clinical trial; preliminary analysisshowed significantly greater weight gain in the groupreceiving a mixture of cereals with milkin which the

latter provided 35 per cent of the total calories than

in the other group consuming isocaloric cereal-baseddiet without milk. There was only a modest 15 percent increase in the stool output in the milk group but

the treatment failure rates were similar (Table IlI)73.

Similarly, Bhutta et aJ14 reported lower stool out-

put and greater weight gain in persistent diarrhoeawith curds cereal mixtures than with lactose free soybased diets. Together, these studies make a case forreduction rather than total elimination of milk as theinitial step. A modest amount of milk in cereal dietsimproves their protein quality, trace elements and

mineral content. Further, the consistency and palat-ability ensures higher intakes of these diets than with

purely cereal-based diets. The possibility of occa-sional milk protein allergy is out weighed by thebenefits offered by adding modest amounts of milk

to cereal based diets.

0.67"

(0.24.1.8723.6

Table IV. Composition of diet A

Amount (g)Ingredients

12.5

40.

2.25

2.0

100

puffed rice

Milk

Sugar

Oil

Water to make

Specific recommendations for the initial diet

Energy density (cal/lOOg)

Per cent protein

Per cent carbohydrate

Per cent lactose

Per cent fat

Amino acid score

96

10.0

55.87

1.73

33.9

1.0

Once a child is ready for oral feeding after fewhours of stabilization, the choice of an initial diet

would be milk rice mixtures with added oi1, yieldingan energy density of about 85-95 kcal/l OOg with 30-35 per cent calories from milk. The diet provides the

ideal minimal 10 per cent energy from a proteinsource. The composition of one such diet is given in

Table IV. milk even in small quantities, rice sugar oil based

diets are appropriate. Egg is well tolerated and pro-

vides useful allimal proteill in such diets.In a small proportion of patients with very severe

diarrhoea where some clinicians feel reluctant to use

10 INDJAN J MED RES, JUL y 1996

Table V. Composition of diet B

Ingredients Amount (g)

13.50

11.0

3.50

3.50

100

Energy density (cal/IOOg)

Per cent protein

Per cent carbohydrate

Per cent fat

Amino acid score

95.22

9.51

56.9

33.29

1.00

and oligosaccharides and part or whole of the fats asri1edium chain triglycerides. Micronutrients and vita-

mins are already added. The diet B is based onsimilar principles and is at least as effective. Its

advantage is that the concentration of the individualsugars can be tailored to each individual child, it ischeap and can be easily prepared by mothers at homeand in small hospitals, the disadvantage is that vita-

mins and mrnerals need to be supplemented.

The role of antim icrobial agents against enteric

pathogens in terms of improved nutrient absorption,

decreased stool output or shortened illness durationis undecided. In a recent Indian study large doses of

nonabsorbable, broad spectrum antibiotic were ad-ministered based on the hypothesis that it woulderadicate aerobic bacterial overgrowth. There was noimprovement in purge rates or weight gain with mas-sive doses of oral gentamicin as compare to a pla-

cebo, despite clearance of stool pathogenic organ-isms including adherent E. coifs. It was conceivable

tl1at systemic antibiotics may be of greater benefit in

persistent diarrhoea. However, a recently concludeddouble blind field trial done to evaluate tl1e efficacy

of metronidazole given alone or in combination withnalidixic acid in comparison to a placebo has shownno significant clinical benefit (Behl et 01, unpub-lished data). Metronidazole was evaluated for itsaction against anaerobes. Other controlled trials with

co-trimoxazole have shown similar results. It ap-pears that when the mucosa is already severely dam-aged, whatever the initiating factors, nutritional sup-po11 is the key to its rapid repair.

Nearly 25 per cent of hospital ized patients show a

poor response to diet A. Useful criteria for defining

treatment failure are reappearance of dehydration at

any time, passage of 7 or more liquid stools in a dayat the end of7days treatment and weight loss or poor

weight gain despite an oral intake of at least 100 cals/kg/day over the previous three days. Poor oral intakeas a result of systemic infection is more often tile

cause of weight loss than true dietary failure in hos-

pitalized children. In milder cases that are managedat the household level, a common reason for poor

weight gain is the offering of only small quantitiesof thin foods to the child by the family.

The factors related to treatment failure on low

lactose (milk) cereal based diets are systemic infec-tion, severe carbohydrate intolerance involving notonly lactose but also otller disaccharides and starch.

Therefore, dietary modification should be made onlyafter effective treatment of associated systemic in-fection. The second line diet should be milk free with

substitution of part of the starch by sucrose or glu-cose (diet B; Table V). This mixture of sugars

achieves the right balance between dietary osmolar-

ity, digestibility and energy density. In such a diet,egg or chicken is a suitable protein source. Monosac-

charides as the only carbohydrates in the diet shouldbe used for the few patients who are treatment fail-

ures on diet B as it is difficult to provide sufficientenergy density with the permissible 2-3 per cent

glucose concentration; at higher concentrations os-motic diarrhoea may develop.

Until their role in the management of persistent

diarrhoea is well established, generous but safe

amounts of micronutrients equivalent to 2 times theRDA should be provided. These may include vitamin

A, zinc, iron, folate and when feasible others. Se-verely malnourished children should receive magne-sium 1-1.5 mi/kg body weight ofa 50 per cent solu-

tion, given IM for 2-3 days. Patients on a milk freediet should also receive calcium supplementation.

Several commercial diets are also available. Forreasons that are not fully explained, home based low

lactose or lactose free diets perform much better than

commercial soy based formulations74. Semi elemen-tal diets like Nutramigen or Progestimil are usefulbut expensive. They usually contain protein hydroly-sate or calcium caseinate, mixture ofdisaccharides~

IIIBHAN et al : PERSISTENT DIARRHOEA IN CHILDREN

An effective antishigella agent should be used in

the presence of blood or numerous pus cells in thestools. Treatment for giardiasis and amoebiasis is

indicated when a stool examination reveals tropho-zoites. Currently, there is no suitable tre~tment for

cryptosporidium infection.

Cholestyramine, which binds unconjugated bile

salts or bacterial toxins, has not proved to be useful.There is also little evidence of clinical benefit when

lactobacilli are administered to replace intestinal

m icroflora.

Systemic antibiotics are required to treat associ-

ELIGIBLE FOR HOSPIT'\LIZATION

YESNO

~

Stabi.Lized over6-24 hours

NOYES

IN'l'ENS IVE CAREUNIT'

TREATMElrT FAILURE

fig. 2. Practical algorithm tor thc trcal1nent or pcrsistcllt diarrhoea.

12 INDIAN J MED RES, JULY 1996

Therefore, it is the outpatient care of patients with

persistent diarrhoea that needs strengthening insteadof the current over focus on sophisticated treatmentof the few patients who actually get to hospitals.

References

ated pneumonia, septicemia, meningitis or urinary

tract iAfection in patients with persistent di~rrhoea.These infections are detected in nearly 60 per cent of

patients with associated severe malnutrition usuallyseen in a hospital setting, they are much less com-

mon when malnutrition is only of moderate severity,as is likely witll those being treated as outpatients.

The search for such infections should be vigorous,even in the absence of fever; useful inql.cators are

persistent anorexia, refusal of liquids an4breast n)ilk.and dehydration despite modest stool 19sses.

These recommendations are summarised il~ a treat-ment algorithm (Fig. 2). This algorithm was evalu-ated in an international study including th~\group atthe All India Institute ofMedical Sciences (AIIMS),New Delhi. The success rate for the evaluated 460

children w.ith persistent diarrhoea while on diet A

was 70 per cent (95% CI 65%, 75%) and it was 84

per cent (95%CI 76%,93%) for those evaluated whileon diet B. Weight gain was achieved in over 90 per

cent and associated illnesses requiring antibioticswere found in 61 per cent of the children. The chil-

.dren at greatest risk were the youllgest, those se-

verely '.llalnourished, with highest initial purgingrates and associated infection.

As diarrhoea is 90mmon in children residing in

poor communities where family feeding habits con-tribute to malnutrition, an interaction with healtllcare providers during the illness offers a good oppor-

tunity to improve tile nutrient intake through pur-

posefulnutritional counselling, Mixtures ofmilk andcereals or of cereals and legumes fortified with oilare well tolerated during acute and persistent diar-

rhoea. They have the required energy density and

palatability. About 30-40 per cent of calories can bederived from fat sources without any deleteriouseffect. Mothers nlust receive nutritional counsellingfrom health care providers tllat is practical and takes

into account the family views and realities and in-

cludes clear instructions or the frequency of meals,the amounts to be fed at each and tile solutions to

problems of the individual child and family. This iscurrently tile weakest link in the sick child -healthcare provider interaction.

The vast majority of patients with persistent diar-rhoea are unable to avail of hospital care due to

physical and situational constraints of the family.

I. World Health Organization. Diarrhoeal diseases control.Persistent diarrhoea in children. CDD/DDM/85.1. Geneva:World Health Organization.

2. Bern C. Martines J, de Zoysa I, Glass RI.The magnitude of the

global problem of diarrhoeal disease: a ten year update. Ell//WHO 1992; 70: 705-14.

3. Black RE, Lopez de Romana G, Brown KH, Bravo N, Bazalar

OG, Kanashiro HC.lncidenceandetiology ofinfantilediarrhoeaand major routes of transmission in Wuascar, Peru. Am J

Epidemiol1982; 115: 305-14 U

4. Black RE, Brown KH, Becker S. Effects of diarrhea associatedwith specific enteropathogens on the growth of children inrural Bangladesh. Pediatrics 1984; 73 : 799-805.

5. Bhan MK, Bhandari N, Sazawal S, Clemins J, Raj P, LevineMM et a[. Descriptive epidemiology of persistent dian.hoea

among young children in rural northern India. Bull WHO1989; 67: 281-8.

6. Claeson M, Merson MH. Global progress in the control of

diarrhoeal diseases. Pediatr Infect Dis J 1990; 9 : 345-55.

7. Bhandari N, Bhan MK, Sazawal S. Mortality associated withacute watery diarrhoea, dysentery and persistent diarrhoea inrllral north India. Acta Pediatr 1992; 381 Suppl : 3-6.

8. World l-lealth Organization. Report of a WI-IO Meeting.

Persistent diarrhoea in children in developing countries. WHO/CDD/88.27 Geneva: World Health Organization.

9. Huttly SR, Hoque BA, Aziz KMA Hasan KZ, Palmary MY,I{ahman MM et a[. Persistent diarrhoea in a rural area ofB~ngladesh: a community based longitudinal study. Int JEp.idemioI1989; 18: 964-9.

10. Schorling:)B, Wanke CA, Schorling SK, McAullife JF. deSo~ MA, Gueriant RL. A prospective study of persistentdiarrhoea among children in an urban Brazilian slum: Patternsof occurrence and etiologic agents. Am J Epidemiol1990; 132: 144-56.

II. Lanata CF, BJack RE, Gilman RH, Lazo F, Del Aguila R.

Epidemiologic, clinical and laboratory characteristics of acutevs persisti:nt diarrhoea in periurban Lima, Peru. ; Pedialr

Gastroenterol Nutr 1991; J 2 : 82-8.

12. Baqui AH, Black RE, Sack RB, Yunus MD, Siddique AK,

Chowdury HR. Epidemiological and clinical characteristicsofacute and persistent diarrhoea in rural Bangladeshi children.Acta Paediatr 1992; 81 Suppl 381 : 15-21.

13. l-lenry FJ, Udoy AS, Wanke CA, Aziz K. Epidemiology of

persistent diarrhoea and etiologic agents in Mirzapur,Bangladesh. Acta Paediatr r992; 381 Suppl : 27-31.

113BRAN et al : PERSISTENT DIARRHOEA IN CHILDREN

14. Black RE. Persistent diarrhea in, children of developingcountries. PediatrInfect DisJ 1993: 12: 751-61.

15. Cruz JR, Bartlett A V, Mendez H, Sibrian R. Epidemiology ofpersistent diarrhoea among Guatemalan rur'dl children. Acta

Paediatr 1992; 381 Suppl : 22-6.

16. Victora CG, Huttly SR. Puchs SC, Barros PC, Garemhe M,Leroyo T et al. International differences in clinical patterns ofdiarrhoeal deaths: a comparison of children from Brazil,Senegal, Bangladesh and India. J Diarrhoeal Dis Res 1993; 11

: 25-9.

17. Fauveau V, Henr)' FJ, Briend A, Yunus M, Chakraborty J.Persistent diarrhoea as a cause of childhood mortality in ruralBangladesh. Acta Paediatr 1992; 38! Suppl : 12-4.

18. World Health Organization. Programme for control ofdiarrhoeal diseases. Seventh Programme report 1988-1989.

WHOICDD/90.34.

19. Victora CG, Huttly SR, Fuchs SC, Nobre LC, Barros FC.Deaths due to dysentery, acute and persistent diarrhoea amongBrazilian infants. Acta Paediatr 1992; 381 Suppl : 7.1 I.

20. Penny ME. Commentary .Paediatr Infect Dis J 1993; 12: 762.

4.

21. Lanata CF, Black RE, Maurtua D, Gil A, Gabilonto A, Yi A etal. Etiologic agents in acute vs persistent diarrhoea in childrenunder three years of age in peri-urban Lima; Peru. Acta

Paediatr 1992; 381 Suppl : 32-8.

22. Baqui AH, Sack RB, Black RE, Haider K, Hossain A, AbimAR et al. Enteropathogens associated with acute and persistentdiarrhoea in Bangladeshi children under five years of age. J

"ifect Dis 1992; 166: 792-6.

23. Bhan MK, Khoshoo V, Sommerfelt H, Raj P, Sazawal S,Srivastava R. Enteroaggregative Esch. coli and Salmonellaassociated with non dysenteric persistent diarrhoea. Pediatr

Infect Dis J 1989; 8 : 499-502.

24. Bhan MK, Raj P, Levine fvlM, Kapes JB. Bhandari N,Srinivastava R et al. Enteroaggregative Escherichia coliassociated with persistent diarrhoea in a cohort of ruralchildren in India. J 1nfect Dis 1989; 159 : 1061-4.

25. Cravioto A, Tello A, Navarro A, Ruiz J, Villafar M, Urine P,Brown C et al. Association of Escherichia coli HEp-2 adherencepatterns with type and duration of diarrhoea. Lancet 1991 ; 337

: 262-7.26. Keusch GT, Thea DM, Kamenga M, Kapandak, Muala M. et

al. Persistent diarrhoea associated with AIDS. Acta Paediatr

1992; 381 Suppl : 45-8.

27. Tomkins A. Nutritional status and severity of diarrhoea amongpre school children in rural Nigeria. Lancet 1981; i: 860-2.

28. Bairagi R, Chowdhury MK, Kim PJ, Curlin GT, Gray RH. Theassociation between maln~trition and diarrhoea in rural

Bangladesh.lntJ Epidemiol 1987; 16: 477-81.

29. El Samani EPZ, Willett WC, Ware JH. Association ofmalnutrition and diarrhoea in children aged under five years

age. Am J Epidemiol 1988; 128: 93-105.

30. Sepulveda J, Willett W, Munoz A. Malnutrition and diarrhoea.

Am J Epidemio/ 1988; 127: 365-76.

31. Black RE, Lananta CF, Lazo F. Delayed cutaneoushypersensitivity: epidemiologic factors affecting and usefulnessin predicting diarrhoeal incidence in young Peruvian children.

Pediatr Infect Dis J 1989; 8 : 210-5.

3~. Bhandari N, Bhan MK. Sazawal S, Clemens 10, Bhatnagar S,Khoshoo V. Association of antecedent malnutrition withpersistent diarrhoea: a case control study. Br Med J 1989; 298

: 1284-7.

33. Schorling JB, McAuliffe JF, de Souza MA. Guerrant RL.Malnutrition is associated with increased diarrhoea incidenceand duration among children in an urban Brazilian slum. Int J

Epidemioll990; 19 : 728-35.

34. Baqui AH, Black RE. Sack RB, Chowdhury HR, Yunus M.Siddique AK. Malnutrition, cell mediated immune deficiencyand diarrhoea: a community based longitudinal study inBangladeshi children. Am J Epidemio/ 1993; 137: 355-65.

35. Sazawal S, Black RE, Bhan MK, Bh,andari N, Sinha A, JallaS. Effect of Zinc supplementation during acute diarrhoea on

the duration and severity of the episode -A community-based,double blind controlled trial. N Engl J Med 1995; 333: 839-44.

36. Sazawal S, Black RE, Bhan MK et al. Zinc supplementationreduces the incidence of persistent diarrhoea and dyscnteryamong low socio-economic children in India. J Nutr 1996;

126 : 443-50.37. Sachdev HPS, Mittal NK, Mittal SK, yadav HS. A controlled

trial on utilityoforal zinc supplementation in acute dehydratingdiarrhea in infants. J Pediatr Gastroenlerol Nutr 1988; 7 :

877-81.

38. Roy SK, Behrens RH, Haider R, Abramuzzaman SM,Mahadanabis D, Wahed MA et a/. Impact of zinc

supplementation on intestinal permeability in Bangladeshichildren with acute diarrhoea and persistent diarrhoeasyndrome. J Pediatr Gastroenterol Nutr 1992; l5 : 289-96.

39. Shahid N, SackDA, Rahman M. Alam AN, Rahman N. Riskfactors for persistent diarrhoea. Br Med J 1988: 297: 1036-8.

40. Mahalanabis D, Alam AN, Rahman N. Hasnat A. Prognosticindicators and risk factors for increased duration of acutediarrhoea and for persistent diarrhoea in children. lnt J

Epidemiol1991: 20: 1064-72.

41. Bareto ML, Santos LMP, Assis AMO, Araujo MPN. FarenzenaGO, Santos PAB et al. Effect ofvitamin A supplementation ondiarrhoea and acute lower respiratory tract infections in young

children in Brazil. Lancet 1994; 344 : 228-31.

42. Hambidge KM. Zinc and diarrhea. Acta Paediatr 1992; 38/

Suppl : 82-6.43. Semba DIL Vitamin A, immunity and infection. Clin liifecl

Dis 1994: 19 : 489-99.44: Schlesinger L, Stekel A. Impaired cellular immunity in

m~rasmic infants. Am J C/in Nutr 1974; 27: 615-20.

45. Bloomfield AL, Mateer JG. Changes in skin sensitiveness to

tuberculin during epidemic influenza. Am Rev TlIberculosi.\"

114 INDIAN J MED RES, JUL y 1996

Pediatr J 987; 33 : 177-80.

59. Gore SM, Fontaine 0, Pierce NF .Impact ofrice-based oral rehydrationsolution on stool output and duration of diarrhoea: meta-analysis ofthirteen clinical trials. Br Med 11992; 304 :287-91.

60. Molla AM, Bari A. Role of cereal-based oral rehydration therapy inpersistent diarrhoea in children. Acta Paediatr 1992; 381 Suppl : 104-7.

61. Fayad IM, Hashem M, Duggan C, Refat M, Bahin M, Foutaine ° et al.Comparative efficacy ofrice-based and glucose-based oral rehydrationsalts plus early reintroduction of food. Lancet 1993; 342: 772-5.

62. International Study Group on Reduced-osmolarity ORS solutions.Multicentre evaluation of reduced osmolarity oral rehydration saltssolution. Lancet 1995;345: 282-5.

63. Khoshoo V, Bhan MK. Associated factors of protracted diarrhoea.1ndian Pediatr 1991; 27: 559-69.

64. Bhan MK, Arora NK, Singh KD. Management of persistent diarrhoeaduring infancy in clinical practice. Indian Pediatr 1991;58: 769-74.

65. Khoshoo V, Bhan MK, Arora NK, Sood D, Kumar R, Stintzing G.Leucocyte migration inhibition in cow's milk protein intolerance.Acta PediatrScand 1986; 75: 308-12.

66. Khoshoo V, Bhatnagar S, Bhan MK. Monosaccharide intolerancecomplicating protracted diarrhoea in infants. 1 Pediatr Ga~troenter91Nutr 1989; 9: 131-4.

67. Guiraldes E, Hamilton JR. Effect of chronic malnutrition on intestinalstructure, epithelial renewal, and enzymes in suckling rats. Pediatr Res1981; 15: 930-4.

68. Sullivan PB, Marsh MN. Small intf,stinal mucosal histology in thesyndrome of persistent d!arrhoea and malnutrition: a review. ActaPaediatr 1992; 381 Suppl : 72-7.

69. Walker SJA, Nazar H, Manuel P, Jackson D, Phillips AD, Soeparto P .Protein intolerance as a cause ofpostenteritis diarrhoea. In: E. Lebenthal,editor. Chronic diarrhoea in children. New York: Raven, Press. 1984:407-423.

70. Snyder JD. Dietary protein sensitivity: is it an important risk factor forpersistent diarrhea? Acta Paediatr 1992; 381 Suppl : 78-81.

7}. Roy SK, Haider R, Akbar MS, Alam AN, Khatun M, Eechels R et al.Persistent diarrhoea -Clinical efficacy and nutrient absorption with arice based diet. Arch Dis Child 1990; 65: 294-7.

72. Penny ME, Paredes P, Brown KH. Clinical and nutritional consequencesof lactose feeding during persistent postenteritis diarrhoea. Pediatrics1989; 84: 835-44.

73. Bhatnagar S, Bhan MK, Singh K. Comparison of low lactose andisocaloric lactQ~e free cereal based diet in the treatment of persistentdiarrhoea. (in press),

74. BhuttaZA, MollaAM, [ssani Z, Badruddin S, Hendricks K, Snyder JD.Dietary management ofpersistent diarrhoea: comparison ofa traditionalrice-Ientil based diet with soy formula. Pediatrics 1991; 88 : 101 0-9.

75. Bhatnagar S, Bhan MK, Sazawal S, et al. Efficacy of massive dose oralgentamicin therapy in nonbloody persistent diarrhoea with associatedmalnutrition. J Pediatr Gastroenterol Nutr 1992; 15 : 117 -24.

Pulmonary J)is 1919; 3 : ] 66-8.

46. Starr S, Berkovich S. The depression of tuberculin

reactivity during chicken pox. Pediatrics 1964; 33 :769-72.

47. Kent OC, Schwatrz R. Active pulmonary tuberculosis

with negative tuberculin skin reactions. Am Rev Respir

DisI967;95:411-8.

48. Kauffman CA, Linnemann CC Jr, Schiff GM, PhairJP. Effect of viral and bacterial pneumonias on cellmediated immunity in humans. Infect Immun 1976;13:78-83.

49. Feachem RG, Koblinsky MA. Interventions for controlof diarrhoeal diseases among young children: measles

immunization. Bul! WHO 1983; 61 : 641-52.

50. Sazawal S, Bhan MK, Bhandari N, Clemens J,

Bhatnagar S. Evidence for recent diarrhoeal morbidityas a risk factor for persistentdiarrhoea: a case controlstudy.lnt J Epidem~ol]99]; 20: 540-5.

51. Lima AA, Fang G, Schorling JB de Albuquerque L,Mc Auliffe JF, Mota S et al. Persistent diarrhoea inNorth East Brazil: etiologies and interactions withmalnutrition. Acta Paediatr 381 Supp11992; : 39-44.

52. Khin Maung U, Nyunt-Nyunt Wai, Myo-Khon, Mu-Mu Khin, Tin U, Thane- Toe. Effects on clinical

"outcome of breast feeding during acute diarrhoea. BrMed J 1985; 290 : 587-9.

53. Lembcke JL, Brown KH. Effect of milk containingdiets on the severity and duration of childhooddiarrhoea. Acta Pediatr ] 992; 381 Suppl : 87-92.

54. Brown KH, Lake A. Appropriatc use of human and

non-human milks for the dietary management ofchildren with dian.hoea. J Diarrhoeal Dis Res ]99] ;9: ]68-85.

55; Lanata CF, Black RE, Creed Kanashiro H, Lazo F,Gallardo ML, Ve\.astegui 1-1 et al. Feeding duringacute diarrhoea as a risk factor for persistent diarrhoea.Acta Paediatr 1992; 381 Suppl : 98-103.

56. Chew F, Penna FJ,Peret Filho LA, Quan C,Lopes...

MC, MotaJAC etal. Is dilutIon ofcow's mIlk formulanecessary for dietary management of acute diarrhoeain infants aged less than 6 months? Lancet 1993; 341: ]94-7.

57. Sazawal S, Bhan MK, Bhandari N. Type of milkfeeding during acute d iarrhoeaand the risk ofpersistentdiarrhoea: a case control study. Acta Pa~diatr 1992;381 Suppl : 93-7.

58. Dewit 0, Boudraa G, Touhami M, Desjuex JF. Breath

hydrogen test and stool characteristics after ingestionof milk and yoghurt in malnourished children withchronic diarrhoea and lactose deficiency. J Trop

Dr f\1.K. Bllan, Additional Protessor of Paediatrics, Division of Gastroenterology, Hepatotogy and Nutrition

All India 111stitute of Medical Sciences, Ansari Nagar, New Delhi 110029Reprint requests