Gut 1993; 34: 358-364

Does the fatty acid profile of dietary fat influence itstrophic effect on the small intestinal mucosa?

A P Jenkins, R P H Thompson

AbstractTo compare the enterotrophic effects ofdifferent triglycerides, five groups of eight ratswere fed mixed diets giving 50% of calories asoils rich in either essential fatty acids (EFA),at-linolenic acid, fully saturated fatty acids,oleic acid, or medium chain fatty acids. After21-24 days there were no significant differ-ences between the groups in overall smallintestinal whole gut weight, mucosal weight, ormucosalDNA; overall mucosal protein showedslight variation (p<005) that was compatiblewith differences in food intake between thegroups. However, long chain triglycerides(LCT) and medium chain triglycerides (MCT)differed in their regional effects on cell prolif-eration; all four LCT rich diets increasedmucosal mass and cell proliferation maximallyin the mid small intestine, whileMCT had theirgreatest effect proximally. Subsequently, twogroups of eight rats were fed diets in whichEFA orMCT were given as twice daily boluses(29% of dietary calories) for 20 to 23 days andcompared with a third group of eight ratsreceiving a glucose rich, low fat diet. EFAand MCT boluses increased the overall para-meters of small intestinal mucosal mass andfor both oils the effects were now maximalin the mid small intestine. Thus differenttriglycerides have similar effects on overallsmall intestinal mucosal mass, but MCT differfrom LCT in their regional effects on mucosalcell proliferation when they are given in mixeddiets, although not when given as boluses.(Gut 1993; 34: 358-364)

administered. Thus diets rich in LCT stimulatemucosal proliferation maximally in the mid smallintestine, while glucose rich diets stimulate cellproliferation proximally.' These differences maybe explained by differences in the sites ofabsorption of glucose and lipid, with mucosalcell proliferation being most stimulated at theabsorption sites of luminal nutrients.' Thisimplies that it is essential to make measurementsof the whole small intestine when comparinghigh and low fat diets.

Earlier studies using mixed diets suggestedthat the relative enterotrophic effects of differentLCT may vary.9 1 However, the findings havebeen inconsistent between experiments, and ineach case measurements have been confined toshort segments of proximal jejunum or distalileum, or both, with no attempt made to measurechanges in the mid small intestine or to assess thedistribution of mucosal mass along the smallintestine. Other studies have more consistentlyfound that medium chain triglycerides (MCT),given either in mixed diets or by intraluminalinfusion, exert a reduced enterotrophic effectcompared with LCT,2' " but again there hasbeen little emphasis on the effects of the MCT onthe distribution ofmucosal mass along the lengthof the small intestine.To clarify the relative enterotrophic actions of

different triglycerides, five different oils weregiven initially in mixed diets and their effects onboth the overall mass of the small intestine andon the distribution of mass along the smallintestine were compared. Subsequently theenterotrophic effects of bolus doses of LCT andMCT were compared.

The presence offood within the lumen maintainssmall intestinal mucosal structure and function,and this process has been called luminalnutrition.' It has been suggested that dietarylong chain triglycerides (LCT) are particularlypotent enterotrophic agents2- but we haveshown that the enterotrophic effects ofan oil richin polyunsaturated fatty acids depend on howthe oil is administered.5 Thus, when the oil isgiven to rats as part of a mixed diet over thecourse of 24 hours, it does not increase theoverall mucosal mass of the small intestinerelative to the effects of a diet rich in glucosesubstituted for the fat,' while given in bolusdoses it increases overall small intestinal mucosalmass compared with isocaloric amounts of the oilor glucose in a mixed diet.67 This enhancedenterotrophic effect of bolus doses of LCT maybe caused by the release of an enterotrophicpeptide from the distal gut."LCT also affect the distribution of mucosal

mass and cell proliferation along the length of thesmall intestine, regardless of how they are

Methods

EXPERIMENT IFive groups of eight female Wistar Rats (180-215 g) were fed mixed powder diets giving 50%of total calories as lipid, 35% calories as glucose,and 15% calories as an amino acid mixture(Albumaid Complete, Scientific HospitalSupplies Ltd, Liverpool, United Kingdom). Alldiets were supplemented equally (2-24 g/100kcal) with a vitamin/mineral mixture (SpecialDiets Services Ltd, Essex, United Kingdom).The nature of the lipid included in the diet

varied with each group and the following oils andoil mixtures were used; Efamol (Scotia Pharma-ceuticals Ltd, Surrey, United Kingdom; EFA),a 50% Efamol; 50% linseed oil (Hilton BanksLtd, Surrey, United Kingdom) (w/w) mixture(LIN), a 98% coconut oil (Sigma Ltd, Dorset,United Kingdom): 2% Efamol (w/w) mixture(SAT), olive oil (Sigma Ltd; OA), and a98% MCT oil (Mead Johnson Nutritionals,

GastrointestinalLaboratory, The RayneInstitute, St Thomas'sHospital, LondonSEI 7EH.A P JenkinsR P H ThompsonCorrespondence to:Dr A P Jenkins.Accepted for publication4 August 1992

358

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from

Does thefatty acid profile ofdietaryfat influence its trophic effect on the small intestinal mucosa?

TABLE I Composition ofthe oils

Composition (% by weight)

Fatty acid EFA LIN SAT OA MCT

6:0 caproic acid - - - - 2-98:0 caprylic acid - - 8-9 - 66-6610:0 capric acid - - 5-9 - 23.512:0 lauric acid - - 47 0 - 4 914:0 myristic acid - - 17-616:0palmiticacid 5-5 5 8 8-9 9 5 0-118:0 stearic acid 2-0 2 5 2-0 2-0 0 0418:1 oleicacid 8 5 14-2 0-2 83-5 0-218:2 linoleic acid 75-0 45 5 1-5 4 0 1-518:3 y linolenic acid 9 0 4-5 0-2 - 0-218-3 a linolenic acid - 275 -20:0 arachidic acid - - 2-9 1-0Other - 4-9 -

EFA=Efamol, LIN=50% Efamol/50% linseed oil, SAT=98%coconut oil/2% Efamol, OA=olive oil, MCT=98% medium chaintriglyceride oil/2% Efamol.

Middlesex, United Kingdom): 2% Efamol (w/w)mixture (MCT). Efamol is rich in linoleic acid,linseed oil in a-linolenic acid, coconut oil in fullysaturated fatty acids, and olive oil in oleic acid;the final compositions of the oils and oil mixturesare shown in Table I.The animals were kept to a 12 hour light

(07.00 to 19.00 hours): 12 hour dark cycle at anambient temperature of 21-22°C. They werehoused individually in cages with wire bottomsto reduce coprophagia and were allowed tapwater ad libitum. The diets were delivered fromlow spillage glass containers'4 and all the animalswere allowed to feed from these ad libitum. Foodintake was measured on a daily basis.Two rats from each group were killed by

overdose ofinhaled halothane between 09.00 and14.00 hours on each offour successive days, 21 to24 days after the start of the experiment. Foodwas withdrawn at 23.00 hours the day beforesacrifice and the animals were given intra-peritoneal vincristine sulphate (Lilly France SA,Fegersheim, France; 1-0 mg/kg body weight)exactly 2 hours before being killed in order tomeasure crypt cell proliferation from theaccumulation ofvincristine arrested metaphases.The small intestine, including part of the

duodenum, was removed from the duodenalpapilla to the ileocaecal valve. Its length wasmeasured under lOg vertical tension and theintestine divided into three segments of equallength (proximal, middle, and distal). One cmsamples for subsequent counting of arrestedmetaphases were taken from the proximal end,midpoint, and distal end of the proximal seg-ment, and the midpoints and distal ends of themiddle and distal segments; that is at distances0%, 17%, 33%, 50%, 66%, 82%, and 100% alongthe length of the removed small intestine. Thesegments of small intestine were then flushed

TABLE II Composition ofthe diets used in experiment 2

Composition (% total calories)

EFA bolus MCT bolus Glucose diet

Efamol 29-0 - 1-2MCT - 29-0 -

Glucose 56-0 56-0 83-8Amino acids 15-0 15-0 15-0

EFA=Efamol, MCT=98% medium chain triglyceride oil/2%Efamol mixture.

with ice cold 09% sodium chloride solutionfollowed by air. Adherent fragments ofmesenterywere carefully dissected free and the segmentswere weighed to give an estimate of the wholegut weight. The mucosa was gently blotted withtissue paper, scraped off with a glass slide,weighed and frozen at - 20°C for subsequentdetermination of protein and DNA content.'5I6The 1 cm samples of small intestine were slit

open longitudinally and orientated mucosal sideuppermost on small pieces of card. They werefixed in Carnoy's fluid and stored in 70%aqueous ethanol (v/v). A portion of each samplewas bulk stained using the Feulgen reaction. The2 hour accumulation of vincristine arrestedmetaphases was counted in 10 randomly selected,microdissected crypts,'7 and the mean wascalculated.

In addition to the small intestine, the pancreaswas dissected free from its mesentery and its wetweight was determined.

EXPERIMENT 2Two groups of eight female Wistar rats (190-220g) were fed diets giving 48-2% of total calorieseither as twice daily intragastric 1-5 ml bolusdoses of LCTs (Efamol, EFA; 8-4 kcal/ml), ortwice daily 1-7 ml bolus doses of a 98% MCT oil:2% Efamol mixture (MCT; 7-6 kcal/ml). Aftertwo days, the doses of EFA and MCT wererespectively reduced to 0 9 ml and 1I0 ml twicedaily, giving 29% of total calories as fat. Theremaining calories were provided by a mixture ofglucose (56% total calories) and amino acids(15% total calories) given separately. A thirdgroup of eight rats (190-220 g) was fed a mixeddiet containing 1 2% Efamol, 15% amino acids,and 83 8% glucose. The final compositions of thethree diets are shown in Table II and all rats weregiven a total of 52 kcal/day. As in experiment 1,the three diets were supplemented equallywith vitamin/mineral mixture (2-24 g/100kcal).Two rats from each group were killed on

successive days, 20 to 23 days after the start ofthe experiment and between 09.00 and 14.00hours. On the day before sacrifice each animalwas given half its normal intake of diet includingoil.As described for experiment 1, the small

intestine was removed and divided into threeequal length segments. The whole gut weight,mucosal weight, and mucosal protein and DNAwere measured. The pancreas was also removedand weighed.

STATISTICAL METHODSBetween group comparisons of energy intake,weight gain, small intestinal length, smallintestinal mass parameters, and pancreas weightwere made using one way analysis of variance. Ifsignificant differences were found, individualcomparisons weremade using Student's unpairedt test. The effects of diet on 2 hour metaphaseaccumulation in experiment 1 were analysedusing a repeated measures analysis of variance."The results are expressed as mean (SEM), unlessotherwise indicated.

359

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from

30enkins, Thompson

Results

EXPERIMENT IAnalysis of variance showed that there was asmall but significant variation between the

TABLE iII Effect ofdiet on daily energy intake, rat weight gain, and small intestinal length(Values mean (SEM); n=8)

EFA LIN SAT OA MCT ANOVA

Daily energy intake(kcal/rat) 54-5(1-2) 55-7(1-1)t 56-3(0 9)t 51-7(1-1)* 54-0(1-2) p<005

Ratweightgain(g) 23-6(3 2) 23-5(4 2) 25 6(2 5) 20-1(2-4) 19 5(3 4) NSSmall intestinal length

(cm) 106-1(2-0) 106-8(1-6) 108-9(1-2) 107-6(2-1) 107-4(1-0) NS

Student's t tests: *versus tp<O05, *versus tp<O-Ol.

groups in their energy intake (Table III), andindividual comparisons with t tests showed signi-ficant differences between the intake of OA andLIN and of OA and SAT. However, there wereno significant differences between the groups inweight gain or in small intestinal length (TableIII).There were no significant differences between

the two groups in overall whole gut weight,overall mucosal weight, and overall mucosalDNA (Table IV). Analysis of variance, however,did show significant variation (p<005) inoverall levels ofmucosal protein, while individualcomparisons showed significant differencesbetween the mucosal protein levels for SAT and

TABLE IV Effect ofdiet on overall small intestinal whole gut weight, mucosal weight, and mucosal protein and DNA (Valuesmean (SEM); n=8)

EFA LIN SAT OA MCT ANOVA

Overall whole gut weight(mg/cm) 46-37(1-25) 47-23(1-39) 51-31(1-57) 45-61(1 15) 48-51(1-80) NS

Overall mucosal weight(mg/cm) 15-27 (0 63) 15-66 (0-68) 16-83 (0-91) 14-71(0 66) 16-44(1-19) NS

Overall mucosal protein(mg/cm) 2-31(0-13) 2-12 (0 08)t 2-69 (0-14)* 2-22 (0 09)t 2-38 (0-17) p<005

Overall mucosal DNA([tg/cm) 174-80 (9 40) 165-76 (13-93) 183-42 (13-05) 183-54 (12-21) 187-66 (13-02) NS

Student's t tests: *versus tp<0-05, *versus tp<0-01.

A

P M D

EFA

P M D

LIN

B

P MD PM D

EFA LIN

P M D P M D

SAT OA

L%

IL0oP% MDIMCTd

P M D P M D P M D

SAT OA MCT

Diet

Figure 1: Effect ofdiet on the distribution of(A) whole gut weight and (B) mucosal weigalong the small intestine. Values mean (SEM); n=8; P=proximal, M=middle, D=dissegment. EFA =Efamol; LIN=50% Efamol/50% linseed oil; SAT=98% coconut oil/=Efamol; OA=olive oil; MCT=medium chain triglyceride oil/2% Efamol.

LIN and SAT and OA. The individual intestinalsegment mass parameters are plotted in Figures1 and 2 to show their distribution along the smallintestine. The four diets rich in LCT gave asimilar distribution ofmass parameters along thesmall intestine, with peak levels attained in themiddle small intestinal segment. The diet rich inMCT, however, induced a different distributionof mass parameters; on this diet, mass wasgreatest in the proximal segment and declinedsuccessively in the middle and distal segments.The data for 2 hour metaphase accumulation

are plotted in Figure 3. The curves for the fourdiets rich in LCT were of similar shape, whilethat for MCT differed. Repeated measuresanalysis of variance confirmed that there wassignificant variation in the five slopes (p<0O01),but this was not present if the data forMCT wereexcluded from the analysis.

Table V shows that there was no significantdifference between the groups in pancreasweight, although there was a trend for MCT togive the greatest weight.

EXPERIMENT 2The rats tolerated the initial twice daily 1I7 mlbolus doses ofMCT poorly; shortly after dosingthey became subdued, cold to the touch, anddeveloped severe diarrhoea. After 2 days, there-fore, the dose of MCT was reduced to 1I0 mltwice daily giving 29% of total calories. This wasbetter tolerated, although there was still someinitial diarrhoea. The intake of the Efamol dosedanimals was reduced correspondingly.There were no significant differences in

weight gain between the three groups or in totalsmall intestinal length (Table VI), although

.ht small intestinal length seemed somewhat lesstal% after the glucose rich diet than after the other two

diets.

E0

0)3:

60 -

50 j

40

30 -

20 -

10 -

0-

25 -

0)

0)

4_

0

.)

en116

20 -

15 -

10 -

5 -

360

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from

Does thefatty acid profile ofdietaryfat influence its trophic effect on the small intestinal mucosa?

A

T

P MD PM D

LIN SAT

P MD PMCD

LIN SAT

Diet

P M D P MD

OA MCT

P M D P M D

OA MCT

---- EFA

A-LA LIN

-0--3- SAT

Overall small intestinal whole gut weight andmucosal weight were similar in the groupsreceiving bolus doses of Efamol and bolus dosesof MCT and were increased compared with thegroup fed the glucose rich diet (Table VII).Overall mucosal protein and DNA showed asimilar trend, although the changes forDNA didnot achieve statistical significance (p=008). Inthis experiment the distribution of mucosalmass along the small intestine was similar for theEfamol rich diet and MCT rich diet, with greaterlevels of mass in the middle than in the proximalsegment (Figs 4 and 5). As in earlier studies,5'7the glucose rich diet gave levels of mucosal massparameters that were greatest in the proximalsegment and declined in the subsequent twosegments.There was significant variation in the absolute

pancreatic weight and the pancreatic weightexpressed relative to body weight between thediets (Table VIII). Individual comparisons witht tests showed a significant difference betweenpancreatic weights for the two high fat diets,with MCT giving the higher value.

DiscussionThis study compared the enterotrophic actionsof different triglycerides. In the first experimentthe oils were given in mixed diets with theanimals being deliberately allowed to feed adlibitum. If the intake of all the groups had beenmatched to that of the group with the lowestintake (OA), the remaining groups, and especi-ally the SAT group, would have been effectivelyfood restricted. These animals, therefore, mighthave gorged their food, and, in turn, this mighthave produced a bolus effect that could have

ong created differences between the groups. It wasent- also to avoid a confounding variable that two of

the oils studied, SAT and MCT, containedsupplements of Efamol. This was because themajor constituents of these oils, coconut oil andMCT oil respectively, contain very low levels ofessential fatty acids. It was important to avoidessential fatty acid deficiency, since this itself cancause small intestinal mucosal atrophy. '9The first experiment provided no evidence

that the fatty acid composition of the ingested oilsignificantly affected overall small intestinalwhole gut weight or mucosal weight and anytrends between the groups were broadly inkeeping with differences in energy intake. Thus,whole gut weight and mucosal weight weregreatest in the group with highest energy intake,the SAT group, and least in the group with thelowest intake, the OA group. The variation in

-O mucosal protein just attained statistical signifi-11*. cance (p<0-05), but this was mainly because of

the high level for the SAT group, which mayi have reflected the high intake of this group.

Mucosal DNA, on the other hand, showed no.significant variation and no obvious trend inrelation to energy intake.

Although the LCT and MCT did not differ in-' their effects on the overall mass of the small100 intestine, the first experiment showed that their

effects on the distribution of mass along thelength of the small intestine did vary. Thedifferences may relate to differences in the

4 -

r-._

00.

Caco0C.

3 -

2 -

1 -

0 -

P M D

EFA

300 -

-5? 200 -C)

z00 00

o loo-C.)

0-

B

P M D

EFA

Figure 2: Effect ofdiet on the distribution of(A) mucosal protein and (B) mucosalDNA althe small intestine. Values mean (SEM); n=8; P=proximal, M=middle, D=distal segme(For other abbreviations see legend to Figure 1).

-zCnUl)a)Wm

al)4-c

0.

n

a)E

V

a)CA)I.a)

00z

0 20 40 60 80

Small intestinal distance (%)Figure 3=Effect ofdiet on 2 hour metaphase accumulation. Each point is the mean ofmeasurements from eight rats. (For abbreviations see legend to Figure 1. )

361

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from

3Jenkins, Thompson

TABLE V Effect ofdiet on pancreas wet weight. (Values mean (SEM); n=8)

EFA LIN SAT OA MCT ANOVA

Pancreas weight (mg) 882-33 (73-51) 775 45 (37-78) 820-24 (36-21) 814 79 (44 41) 940 55 (39 24) NSPancreas weight per 100 gbody weight (mg) 404 06 (31-75) 355 73 (20 67) 371-97 (13-55) 379 71 (18-72) 437-12 (18-08) NS

TABLE VI Effect ofdiet on rat weight gain and small intestinal length. (Values mecn=8)

EFA bolus MCT bolus Glucose diet

Rat weight gain (g) 8-1 (4 7) 9-4 (3 4) 9-6(2 5)Small intestinal length (cm) 108-9 (1-4) 110-8 (2 2) 105-8 (0 8)

EFA=Efamol, MCT=98% medium chain triglyceride oil/2% Efamol mixture.

TABLE VII Effect ofdiet on overall small intestinal whole gut weight, mucosal weixmucosal protein andDNA (Values mean (SEM), n=8)

EFA bolus MCTbolus Glucose diet

Overall whole gut weight(mg/cm) 57-81(1-65)* 60-34 (2 53)* 49 41 (1 04)t

Overall mucosal weight(mg/cm) 21 76(0 70)* 23-37(1 36)* 17 83(0-52)t

Overall mucosal protein(mg/cm) 3-02(0 10)* 3 19(019)* 2-60(0-13)t

Overall mucosal DNA(Itg/cm) 214 38 (6 89) 234-57 (16 32) 193-79 (10-91)

EFA=Efamol, MCT=98% medium chain triglyceride oil/2% Efamol mixture. Student's,* versus tp<005; * versus 1p<0 001.

80 -

60 -

40 -

20 -

A

O4-

..... . .

... ....... ...... ....... ...... ....... ...... ..... . . . a .... ...... ...@ @9Po" @@Z"Z Z-" @@@@@@ @@Z@*e @o0Z ****@ @ @ **@Z 9w@Z w"w @@wZ @@"w @@@z* wZ@w @@@@Z 999 *w@ZoZ Xww 6@9

--P M

-T-

...d.0....<...1...<

.D .

Efamolbolus

30 -

20

10

P M D P M

MCT bolus Glucose

B

P M D P M D P M

Efamol MCT bolus Glucosebolus

Figure 4=Effect ofdiet on the distribution of(A) whole gut weight and (B) mucosalalong the small intestine. Values mean (SEM); n= 8; P=proximal, M= middle, D-

segment; MCT= medium chain triglycerides.

in (SEM); absorption sites ofthese two classes oftriglycerideand their digestive products. There is good

ANOVA evidence that luminal nutrients have a directNS enterotrophic effect on the small intestinalNS mucosa, perhaps related to the 'workload' of

their absorption""22 and this effect seems to bemaximal at the site of absorption itself." MCTare more rapidly hydrolysed and absorbed in the

ght, and intestinal lumen than are LCT23 and, therefore,would be expected to stimulate cell proliferation

ANOVA more proximally in the small intestine.The results of this first experiment, therefore,

p<o0ool did not support earlier studies239 -1' whichp<0*001 suggested that there may be differences between

the enterotrophic effects of different LCT, andP<0 °S between LCT and MCT. Differences betweenNS the precise fatty acid composition of the oils used

in the current experiment and those of the earliert tests: studies may contribute to this discrepancy, but it

is also relevant that in most of those studies.39"'"the overall small intestinal mucosal mass anddistribution of mass along the small intestinewere not measured. Without these measures it istherefore difficult to interpret the significance ofdifferences in isolated short segments of smallintestine.

In the second experiment, the enterotrophiceffects of bolus doses of MCT were comparedwith those of LCT. Administration of LCT inbolus doses produces an enhanced enterotrophiceffect compared with the same daily dose of oilconsumed as a mixed diet,6 perhaps becausethere is increased delivery of lipid to the distalgut, where it may release an enterotrophicpeptide.6 It was anticipated that boluses ofMCTmight have had a less potent enterotrophic effectthan boluses of LCT, since MCT might have

D been so well absorbed proximally in the smallintestine that little would have reached the distalileum to enhance release of an enterotrophicpeptide. Contrary to these expectations, how-ever, the study showed that bolus doses ofMCThad an overall enterotrophic action at leastequivalent to that of isocaloric doses of LCT.Interestingly, administration of MCT in bolusdoses also affected the distribution of mucosalmass along the small intestine; after the bolusdoses of MCT mucosal mass was greatest in themiddle small intestinal segment, whereas thefirst experiment had shown that MCT in the

[ mixed diet most increased mucosal mass in themost proximal segment.The enterotrophic effects of the bolus doses of

MCT may have been related to the diarrhoea thatthey were observed to cause, for, as a result ofaccelerated transit, the oil might have travelledfurther along the small intestine before being

D absorbed, thus maximally enhancing cellproliferation in the mid rather than the proximalsmall intestine. The presence of unabsorbed

weight MCT in the distal small intestine might also have-distal released an enterotrophic peptide that enhanced

their overall trophic effect to the small intestine.

E

-cCD

._

0)

0a)

E-c0)

0)E

362

0'r

... ...

... ...

.. ...

... ...

.. ....

... ...

.. ...

... ...

... ....

.. ...

... ....

.. ....

... ...

"F"

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from

I)oes theJwttI ! acid profile of dietarv tat influience its trophic eJfect on the small itntestinial tzucosa? 363

5

A

4

0

22

0--------

P M D P M D P M D

Efamol MCT bolus Glucosebolus

2 0---

0

P M D P M D P M D

Efamol MCT bolus Glucosebolus

Figure S= kfJect of diet on the distribuoon oJ (A) mucosall protein, alnd B) mnucosal l)N\A alontgthe smalll intestine. 'alules mealn (SkAlI); n- 8; I'-proximal,ll?i middle, l) =distal .segJment;AlIC, l =nmdiumt chalin triglvceride.s.

It is noteworthy that MCT infused directly intothe human ileum cause a release of entero-glucagon,5 which has been proposed as anenterotrophic hormone.'The mechanism for the MCT induced

diarrhoea is not clear, but recent studies in therat suggest that MCT may stimulate greaterrelease of cholecystokinin than do LCT.'Cholecystokinin may itself stimulate intestinalmotility27 and the output of pancreaticobiliarysecretions into the lumen,25 both of which couldcontribute to the diarrhoea. Cholecystokinin isalso trophic to the pancreas25 and it is interestingthat the pancreatic wet weight was increased bythe boluses of MCT. Indeed, even as part of amixed diet (experiment 1, Table V), MCT

TAIBI.I VIII /fre?t of/ diet onl pancfrealtic wet weig-ht. X al uesxt meant i(kAlI); n1=8)

1(1-A ho/u(s MCi> ho/ils (;lCL(-{s' dliet .1ixoi) ',

IPancrcsas wighlt mg )42 59 30) 17* X(0)5*6 45 52 4 729-62 60)36 )t p<05-(*)30an0rasTeighC per 1() g body,weight Imgi 3()255X 1I1 )' 372 66 5 31)t 342 21 19)314 P<05)°

Il;A= bs];dl,ll MCT-r 95o, mediumti chainl rigiveeridie oil/2" 1.;tamol mlixmtre. SC det'sct ttest:*'versu tN5; * ersus(. tp5t( (10 t ve-suxs tNS.

showed a non-significant trend to increasepancreatic weight.The results of the current study are compatible

with our earlier findings.' The differencesbctwecn the effects of MCT and LCT in mixeddiets on the distribution of mucosal mass and cellproliferation along the small intestinc are con-sistent with our proposal that regional variationin mucosal proliferation may reflect differencesin the sites of absorption of ingested nutrients.Although surprising, the enterotrophic effects ofbolus doses of MCT also do not contradict ourearlier results. As discussed above, it is possiblethat accelerated intestinal transit accounts for theeffects of MCT boluses on both regional andoverall levels of small intestinal mucosal mass.From a clinical viewpoint, it is disappointing

that none of the oils in the current study has beenshown to have an especially potent enterotrophicaction. Such an action could be useful in thetreatment of the short bowel syndrome, forexample. It is still possible, however, that thereare other oils with enhanced enterotrophicproperties, and this has been suggested recentlyfor both short chain triglycerides " and fish oil."'This possibility is currently under investigation.

We thank N Curtis-Davis and Miss M-I Ileffernan for technicalassistance and N Taub f'r statistical advice. APJ was a NutritionalResearch Foundation/British Digestive Foundation RescarchFellow. We are grateful to the 'Trustees of' St Thomas's Ilospitalfor their continuing sttpport. Efamol was a gift from ScotiaPharmaceuticals Ltd and Albumaid a gift f'rom Scientific HospitalSupplies Ltd.

1 I)owling RH. Small bowel adaptation and its regttlation. St-and.' (astroenterol 1982; 17 (suppl 74): 53-74.

2 Morin CL, Grey V'I, Garofalo C. Influence of lipids onintestinal adaptation after resection. In: Robinson JWL,Dowling RI!, Riecken Eo, eds. Mlechanismns of intesthnaladaptation Lancaster: MTP' Press, 1982: 175-84.

3 Weinberg ILM, Pusateri JI', Levine GM. Comparison ofdifferent caloric substrates on intestinal adaptation in therat. Gastroenterologv 1989; 96: 15 14-20.

4 Maxton DG, Cvnk E'U, Jenkiins AP, 'T'hompson RPII. 'heeffect oftdietarv tfat on the small intestinal mucosa. Gut 1989;30: 1252--5.

5 Jenkins Al', Thompson RPl'. The cff'cct of dictary fat on thedistribuLtion of muLcosal mass and celi proliferaitioni aloing thesmall intestine. (Gt 1992; 33: 224-9.

6 Jenkins AI', (ihatcei M1A, Bloom SR, 'I'hompsoni RPtI. I'hceffects of' bolus doses of tfat on small intestinal struLctuLre andon rclcase of gastrin, cholcvcstokinin, pcpti.d tyrosine,tyrosine, and ceiteroglucagon. (uti 1992; 33; 218-23.

7 Jenkins AlP, Thompson RPII. 'he trophic eftfect of'Efanmol ointhe rat small intcstinal mLicosa. C/nSin 1989; 77: 555-9.

8 Wescer E, Vandeventer A, Tawsil T. Non-hormional regulationof intestinal adatptatioin. Scand j Gaiistro'nitc'ro/ 1982; 17(suppl 74): 105-13.

9 Jacobs LR. l'ffect ot'short-tcrmi dictary fMt on cell groswth ini ratgastrointestinal muLcosa and pancreas. Atn7 (,lin Nutr 1983;37: 361-7.

10 'Thomson ABR, Kcclan M, Clandinin NMIl', Walker K. I)ietarvtfat selectively alters transport properlties of-rtjojunuLm.IL1 'C/in Invest 1986; 77: 279-88.

11 'I'homsoni ABR, Keelan M, (iarg M1, Clandinin MI. Spectrumot' effects of dictary long-chain fatty acids oni r-at intcstinalgIuCOSCe and lipid ilntiake. (an' I'hYsiol 1'hartnaco/ 1987; 65:2459 65.

12 Sagher FA, t)odge JA, Johnston CF, Shasv C, Bltchaniani Kl),Carr KE'. Rat simall intiestinal morphology aind tissuecegultatorF' peptidcs: tff'ects ot' higl dietary fat. BIr 7 N\'utr1991;65:21 8.

13 \'anderhoof JA, (Granidjcan (J ,KaUf'IMian SS, Burklev K'l,Aintonisoi D)l.. E'ffect ot' high per-cenitage mediumi-Chainltriglyceride diet on ImuLICOsal adaptation folloswing taissisehbsvel rcscctioin in Irits Parent I/itF \'iNwr 1984: 8: 685 9.

14 Scnapati A, Johnisoni (, Brosn IMI , 'IThonpson RI'l Amethod to rcdtice spillage in pailr-fd rats. Alnni 'Iclnol1984; 35: 123 4

15 I osrv ()11, Rosebi-ough NJ , FSarr A I- Randall1 RJ Proteinmeasusrmcnl1CIt si itt1 thcv flioll phenslol reaigent. '7 1ol! (.'ls'm19t5 1; 193 26i 7S

16) 1'rasad1 AS, 1)u .\Iouihcslle 1, KonaiTtih I), ()berIla>s 1). Asimplel Iluoromlti. Iii.mchodl for die dtetrminattiltil of KNAand1t 1)NA in1 tissuIcs 71 a/i? ( unlIeu\t(l 97 '; 80: 59)8 (102.

1 7 (joodilad RA EX Iighi NAX Q}tiantitiats studbies onl cpu heliallrepla.vlcmcn in1 thi gut In: 'lit.}chn IL ed. '1'cetniqucsllZ' inl ti/il/c sciences'.Ice(/iniql/cs il dii.ctstii'X' p/n\'sio/0s\'. \'i '21)(Shannlon: lilsesier BAiomedbical P'ress, 1982: 23.

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from

364 Jenkins, Thompson

18 Armitage P, Berry G. Statistical methods in medical research.2nd ed. Oxford: Blackwell Scientific Publications, 1987.

19 Hart MH, Grandjean CJ, Park JHY, Erdman SH, VanderhoofJA. Essential fatty acid deficiency and postresection mucosaladaptation in the rat. Gastroenterology 1988; 94: 682-7.

20 Jacobs LR, Taylor BR, Dowling RH. Effect of luminalnutrition on the intestinal adaptation following Thiry-Vellaby-pass in the dog. Clin Sci 1975; 49: 26P.

21 Clarke RM. 'Luminal nutrition' versus 'functional work-load'as controllers of mucosal morphology and epithelialreplacement in the rat small intestine. Digestion 1977; 15:411-24.

22 Richter GC, Levine GM, Shiau Y-F. Effects of luminal glucoseversus non nutritive infusates on jejunal mass and absorptionin the rat. Gastroenterology 1983; 85: 1105-12.

23 Greenberger NJ, Rodgers JB, Isselbacher KJ. Absorption ofmedium and long chain triglycerides: factors influencingtheir hydrolysis and transport. J Clin Invest 1966; 45: 217-27.

24 Spiller RC, Trotman IF, Adrian TE, Bloom SR, Misiewicz JJ,Silk DBA. Further characterisation of the 'ileal brake' reflexin man - effect of ileal infusion of partial digests of fat,protein and starch on jejunal motility and release of neuro-tensin, enteroglucagon and peptide YY. Gut 1988; 29: 1042-51.

25 Bloom SR, Polak JM. Enteroglucagon and the gut hormone

profile of intestinal adaptation. In: Robinson JWL,Dowling RH, Riecken EO, eds. Mechanisms of intestinaladaptation. Lancaster: MTP Press, 1982: 189-98.

26 Douglas BR, Jansen JBMJ, de Jong AJL, Lamers CBHW.Effect of various triglycerides on plasma cholecystokininlevels in rats. ] Nutr 1990; 120: 686-90.

27 Kellow JE, Miller LJ, Phillips SF, Haddad AC, ZinsmeisterAR, Charboneau JW. Sensitivities of human jejunum,ileum, proximal colon and gallbladder to cholecystokininoctapeptide. Am] Phvsiol 1987; 252: G345-56.

28 Walsh JH. Gastrointestinal Hormones. In: LR Johnson, ed.Physiology of the gastrointestinal tract. 2nd ed. New York:Raven Press, 1987: 181-253.

29 Hughes CA, Breuer RS, Ducker DA, Hatoff DE, DowlingRH. The effect of cholecystokinin and secretin on intestinaland pancreatic structure and function. In: Robinson JWL,Dowling RH, Riecken EO, eds. Mechanisms of intestinaladaptation. Lancaster: MTP Press, 1982: 435-50.

30 Kripke SA, De Paula JA, Berman JM, Fox AD, Rombeau JL,Settle RG. Experimental short-bowel syndrome: effect of anelemental diet supplemented with short-chain triglycerides.Amj Clin Nutr 1991; 53: 954-62.

31 Vanderhoof JA, Adrian TE, Blackwood DJ, MohammadpourH, Herrington MJ, Park JHY. Effect of high fat diets onmucosal adaptation following massive bowel resection.Gastroentrology 1991; 100: A552.

on Novem

ber 15, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.3.358 on 1 March 1993. D

ownloaded from