Oral functional characteristics and gastrointestinal pathology: an epidemiological approach

Journal of Oral Rehabilitation 2001 28; 668±672

Oral functional characteristics and gastrointestinal pathology: an epidemiological approach A . T O S E L L O * , B . F O T I * , C . S E D A R A T ², J . M . B R O D E U R ³, J . M . F E R R I G N O * , P . T A V I T I A N * , G . S U S I N I * & J . J . B O N F I L * *Service d'Odontologie, HoÃpital Nord (Unite Fonctionnelle de

Recherche Clinique et d'Evaluation), Marseille, France, ²Faculte de Chirurgie Dentaire, Bordeaux, France and ³Universite de MontreÂal, Faculte de MeÂdecine, DeÂpartement de MeÂdecine Sociale et PreÂventive, Canada

One of the main problems for aged people is that of having a balanced diet, improved by correct conditioning of the alimentary bolus. Does the masticatory status in¯uence the appearance of gastrointestinal disorders? Such a question justi®es the present study. The present work concerned all the patients (211) of eight geriatric institutions, and was completed by a retrospective approach (case±control study). The main result brought by the case±control study, showed that

SUMMARY

Introduction Although elderly people may have excellent physical and mental capacities, they often suffer from a gradual loss of autonomy. One of the main problems for the aged is that of having a balanced diet, improved by correct conditioning of the alimentary bolus through the masticatory function. Studies concerning the in¯uence of dentures or of oral functional characteristics on gastrointestinal pathologies, appear to be few. Hunt et al. (1985) did not ®nd a signi®cant link between the fact of being edentulous and the risk of developing a gastric pathology. Similarly, Mumma and Quinton (1970) stated that masticatory ef®cacity did not in¯uence the appearance of certain digestive symptoms such as anorexia, nausea and sickness in partially dentate, hospitalized subjects. A recent study (Gilbert et al., 2 1998) has con®rmed the link between poor masticatory ef®cacity and the decrease in the number of functional, opposable dental (pairs). The correct conditioning of the alimentary bolus may avoid gastrointestinal pathologies as shown by certain ã 2001 Blackwell Science Ltd

poor oral functional characteristics represent a risk for gastrointestinal pathology (age being neutralized; odds ratio (OR): 2.62). This result indicates that the loss of teeth must be compensated by functionally effective dentures. Not only will the patients' physical and psychic health improve but also medical and public health problems will be solved. KEYWORDS : epidemiology, oral health, gastrointest1 inal disorders

studies, notably those conducted by Brodeur et al. (1993). They mainly highlighted the harmful role of non-functional prostheses; moreover these studies were performed on a non-institutionalized sample of the population of Quebec. The aim of the present work was to study the in¯uence of oral functional performance in general on the gastrointestinal pathology, in a sample of the French population living in geriatric institutions.

Materials and methods General experimental design The present investigation was carried out transversally, and concerned all the patients of eight geriatric institutions. It was completed by a retrospective approach (case±control study). The investigation consisted of an anamnesis followed by an oral examination of each subject; the medical record of each individual established by the physicians of the institution, was consulted and, at the same time, the auxiliary nursing staff were questioned.

668

ORAL FUNCTION AND GASTROINTESTINAL PATHOLOGY An assessment relating to the different variables stated below was then made.

Table 1. Oral functional characteristics. The absolute and relative frequencies are indicated. Many subjects are completely edentulous or have full dentures. A signi®cant number has a natural set of teeth

Variables studied 1. Age; 2. Sex; 3. Gastrointestinal pathology. More precisely, in the present work, the following gastrointestinal abnormalities or pathological events were observed: heart burns; precocious satiety; gastroesophagal re¯ux; transit disorders; regurgitation; epigastric pain; tumoural formation; pancreatitis and vomiting. The presence or absence of this pathology was veri®ed. In these institutions such events were detected or diagnosed by attending physicians, with the aid of medical staff. As many elderly patients are reluctant to complain, special attention was required from the medical staff, nurses, during mealtimes. An abnormal behaviour or presenting symptoms triggered off further investigations through, clinical, paraclinical and biological examination. Two groups were established in the present study. Subjects who did not present any of the above manifestations were considered to have no gastrointestinal pathology. Conversely, subjects presenting at least one of these manifestations were recorded as with pathology. 4. The performance of oral functional characteristics; the subjects were classi®ed in the following categories of the present variable: (a) Subjects with complete dentures. (b) Completely edentulous subjects without dentures. (c) Natural set of teeth. These were subjects presenting a complete dental set of teeth or with no more than four teeth missing other than third molars. Also included in this category were subjects presenting a limited edentulous area compensated by a ®xed prosthesis. (d) Partially edentulous subjects (called `teeth + dentures'). These subjects generally had partial prostheses. The categories of variable no. 4 have been grouped so as to form variable no. 4B, that is to say: 4B. Normal set of teeth or not Yes: subjects classi®ed as having a natural set of teeth (cf. above); No: other categories, i.e. subjects with complete dentures, completely bimaxillary edentulous subjects without dentures, partially edentulous subjects, etc. ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 668±672

Complete edentulous subjects (without dentures) Subjects with complete (bimaxillary) prostheses Natural set of teeth Partially edentulous subjects (teeth + dentures)

n

%

44 64 60 43

20á85 30á33 28á44 20á38

Table 2. Diet consistency. The absolute and relative frequencies (%) of the different types of consistency are indicated. Many 8 subjects have a soft or semisolid diet

Soft Semisolid Solid

n

%

25 27 66

21á19 22á88 55á93

5. Diet Diet was classi®ed as soft (when reduced to the state of pureÂe), semisolid (small pieces inferior to 1á5 cm2) or solid (current food). Statistical study Statistical data were studied by using PCSM (Delta Soft, Bordeaux) and Epi.Info* software. For the various tests used in the analysis of the results, the level of statistical signi®cance was ®xed at P ˆ 0á005.

Results and discussion Main features of the population studied The population studied was composed of 211 individuals with an age range between 44 and 100 years. The mean age was 83á29 years. Women were in a large majority (84á46%). The oral functional performance is presented in Table 1, and the type of diet in Table 2. It is important to notice that the presence of a gastrointestinal pathology was not linked to the sex of the subject, which meant the study could be continued within the envisaged framework (v2 ˆ 1á2717; P ˆ 0á26). *WHO, Geneva.

669

670

A . T O S E L L O et al. Table 3. Natural set of teeth and gastrointestinal pathology. The line and column percentages are indicated in brackets for each subject

Gastrointestinal pathology No

Yes

Table 4. Oral functional performance and gastrointestinal pathology. There is a signi®cant link between the variables 9 (v2 ˆ 9á7502; P ˆ 0á002)

Complete bimaxillary prostheses

Edentulous subjects without dentures

Normal set of teeth

Partially edentulous subjects (with dentures)

No

49 (30á4) (76á6)

33 (20á5) (75á0)

53 (32á9) (88á3)

26 (16á1) (61á9)

Yes

15 (30á6) (23á4)

11 (22á4) (25á0)

7 (14á3) (11á7)

16 (32á7) (38á1)

Natural set of teeth No

Yes

108 (67á1) (72á0)

53 (32á9) (88á3)

42 (85á7) (28á0)

7 (14á3) (11á7)

There is a statistically signi®cant link between the `normal set of teeth' (Yes±No) and `gastrointestinal pathology'(Yes±No) variables (v2 ˆ 6á391; P ˆ 0á011).

Analytical study The statistical analysis shows a correlation between the age and the presence of a gastrointestinal pathology (r ˆ 0á1832 with a con®dence interval at 95% [0á0490±0á3109]; statistical signi®cance of the slope is established at P ˆ 0á007). Subjects presenting a gastrointestinal pathology were of a signi®cantly different age from those without this pathology (mean age: 86á06 years for those with the pathology vs. 82á55 for those without the pathology; with P ˆ 0á007 for the comparison test of the mean values). It also appears that subjects with a natural set of teeth suffer from fewer gastrointestinal pathologies than others (study with the chi-squared test: P ˆ 0á011; cf. Table 3). More precisely, a study of the `oral functional characteristics' variable (no. 4), in accordance with the gastrointestinal pathology, indicated that it was subjects with a natural set of teeth vs. partially edentulous subjects who were at the origin of the statistically signi®cant difference concerning the gastrointestinal pathology (comparison test of the columns of Table 4 with corrected P-value equal to 0á00854, also see Fig. 1). Finally, it appears that the types of diet were related to oral functional characteristics (cf. Table 5; result of test v2 ˆ 11á49; P ˆ 0á009). The gastrointestinal pathology was strongly linked to the age of the subjects on one hand, and to the oral functional characteristics on the other; the latter also being linked to age. Age may therefore appear as a confounding variable in a study which aims to establish that poor oral functional characteristics are a risk factor for gastrointestinal pathologies.

Gastric problem

The line and column percentages are indicated in brackets, for each case. There is a statistically signi®cant link between the two variables (v2 ˆ 9á75; ddl ˆ 3; P ˆ 0á02). The multiple comparison test of columns shows a signi®cant difference between subjects with a normal set of teeth v. edentulous subjects with partial dentures.

Fig. 1. Oral functional performance and gastrointestinal pathology (Yes±No). Distribution in percentage. Note the distinctive difference in gastrointestinal pathologies between subjects with normal sets of teeth and subjects with dentures.

The risk which poor oral functional characteristics represent for gastrointestinal pathology was therefore studied after age adjustment in order to neutralize this confounding factor. A covariance analysis was then ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 668±672

ORAL FUNCTION AND GASTROINTESTINAL PATHOLOGY Table 5. Oral functional characteristics and type of diet. Line and column percentages are indicated in brackets for each case. There is a statistically signi®cant link between the two variables (v2 = 11á49; P = 0á0097) Oral functional characteristics Type of diet

Complete prosthesis

Edentulous Normal without set of Partially denture teeth edentulous

Soft or semisolid

11 (21á1) (28á9)

17 (32á7) (34á2)

13 (25) (34á2)

11 (21á1) (55)

Solid

27 (40á9) (71á1)

5 (7á6) (22á7)

25 (37á9) (65á8)

9 (13á6) (45)

performed, with age as the covariant, the gastrointestinal pathology transformed in Boolean variable as determined quantitative variable (0á1 depending on whether it was absent or present). The determining factor was the `oral functional characteristics'. The results of this test with adjustment on age show that gastrointestinal pathologies continue to be signi®cantly linked to `oral functional characteristics' (P ˆ 0á039; cf. Table 6). Consequently, the study could continue with the de®nition, within the framework of a `case±control' Table 6. Analysis of covariance. The lines do not diverge signi®cantly from parallelism: the analysis of covariance is licit with regard to this condition of validity (P > 0á0500)

study, of an odds ratio (OR). The odds ratio determined the risk of a gastrointestinal pathology occurring for subjects presenting poor oral functional characteristics. The `case±control' study is con®gured on the contingency table (Table 7). Two strata of age were created in order to neutralize the confounding factor (®rst stratum ³ 85 years; second stratum < 85 years; 85 years being the median value of the age distribution). The subjects were considered at risk if they did not have a natural set of teeth (either subjects with complete bimaxillary prostheses or completely edentulous subjects with no dentures or partially edentulous subjects), and not at risk if they had a natural set of teeth. The subjects were considered as `cases' if they presented a gastrointestinal pathology, and as `control' if they did not. It emerges that Mantel Haenszel's v2 is equal to 4á04 (P ˆ 0á044) and that the limits of the con®dence interval at 95% for the odds ratio according to Corn®eld are [1á02 < OR < 6á95] (OR MH: 2á62). It is therefore appropriate to say that the fact of being edentulous or of having a more or less well compensated set of teeth, constitutes a risk factor of gastrointestinal pathology with reference to subjects with normal sets of teeth. The results attached to this study enable certain considerations to be made.

Variation source

Sum of square

d.f.

Variance

Factor A (natural set of teeth or not) Deviation from parallelism Rest

0á7261 0á0160 35á5641

1 1 206

0á7261 0á0160 0á1726

Total

36á3062

208

F

P

4á2058 0á0390 0á0928 0á7590

The adjusted mean values differ signi®cantly (P £ 0á0500). The in¯uence of factor A (natural set of teeth or not) is statistically signi®cant regarding the presence of a gastrointestinal pathology. Table 7. Case±control study. Analysis according to a contingency table strati®ed in two age strata (age ³ 85 years and < 85 years). The subjects without a natural set of teeth are considered as exposed to the supposed risk factor; the subjects presenting gastrointestinal pathologies are cases; those not presenting such pathologies are controls. There is a statistically signi®cant link between exposure and pathology, P = 0á044. It is therefore licit to de®ne an odds ratio

Without gastrointestinal pathology (controls) With gastrointestinal pathology (cases)

Age ³ 85 years

Age < 85 years

Natural set of teeth

Natural set of teeth

No (Exposed +)

Yes (Exposed ±)

No (Exposed +)

Yes (Exposed ±)

57 29

19 4

51 13

34 3

v2 MH = 4á04; P = 0á044; OR MH: 2á62. ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 668±672

671

672

A . T O S E L L O et al.

It would appear that mainly subjects having only a few teeth and more or less suitable dentures are preferentially exposed to gastrointestinal pathologies. Therefore, within the framework of this study, one can af®rm that it is advisable to maintain as many pairs as possible of antagonistic teeth in good condition; but one can suppose it is not recommended to keep nonfunctional teeth, in poor condition, connected to poorly effective dentures. In these last cases, it would probably be advisable to envisage a complete prosthesis (cf. data in Table 4 and Fig. 1). Moreover, this study suggests that the poor functional characteristics leads to a semisolid or soft consistency of the alimentary bolus. There has been a suggestion that the origin of some gastric disorders may be connected with such a form of bolus which is a result of a low ®bre diet. Some Canadian studies have stressed the need of a high ®bre intake in order to prevent certain gastro-intestinal disorders requiring medical treatment (Health and Welfare, 1982). In a 4 Scandinavian study Sandstrom and Lindquist (1987) reported on the dif®culty of completely edentulous subjects aged 61 and over, and wearing a prosthesis, when trying to eat raw carrots or whole apples. These are examples of foods with high ®bre content 5,6 (Schneeman, 1987). Brodeur and Simard (1990) also supported the relationship between a low ®bre diet and digestive disorders.

Conclusion The considerations which link functionally de®cient oral characteristics to gastrointestinal disorders indicate that the loss of teeth during the course of life must be as far as possible avoided; in any case, tooth loss must be compensated by functionally effective dentures. Not only will the patients' physical and psychic comfort be improved, but medical and public health problems will also be reduced. One can consider that upon retirement, when the subjects have a lower purchasing power, teeth will not be replaced and gastrointestinal and nutrition problems

may result. It therefore appears completely justi®ed to reduce the costs and improve the payment of dental expenses for such elderly people. The agreed cost would be compensated by the effectiveness (avoided gastrointestinal disorders and health expenses, etc.). If it is of particular relevance for residents of geriatric institutions to maintain subjects with functional teeth, then their oral health and hygiene must be regularly monitored. Moreover, geriatricians concerned by the problem of malnutrition should be better informed about the signi®cance of oral health and an adequate dentition.

References BRODEUR , J.M. & SIMARD , P.L. (1990) ProbleÁmes gastro-intestinaux et nutrition inadeÂquate chez les personnes aÃgeÂes compleÁtement eÂdenteÂes sans protheÁse fonctionnelle Rapport de Recherche. (Document Interne) Groupe de Recherche en Sante 7 Dentaire Communautaire, Universite de Laval.  E , R. & LACHAPELLE , D. (1993) BRODEUR , J.M., LAURIN , D., VALLE ALLEE Nutrient intake and gastro-intestinal disorders related to masticatory performance in the edentulous elderly. Journal of Prosthetic Dentistry, 70, 468. GILBERT , G.H., FOESTER , U. & DUNGAN , R.P. (1998) Satisfaction with chewing ability in a diverse sample of dentate adults. Journal of Oral Rehabilitation, 15, 15. HEALTH AND WELFARE (1982) Canada's Food Guide Handbook (Revised). Supply and Services Canada, Ottawa, Canada. HUNT , R.J., BECK , J.D., LEMKE , J.H., KOHOUT , F.J. & WALLACE , R.B. (1985) Edentulism and oral health problems among elderly rural iowans: the Iowa 65 rural health study. American Journal of Public Health, 75, 1177. MUMMA , R.D. & QUINTON , K. (1970) Effect of masticatory ef®ciency on the occurrence of geriatric distress. Journal of Dental Research, 49, 69. SANDSTROM , B. & LINDQUIST , L.W. (1987) The effect of different prosthetic restorations on the dietary selection in edentulous patients. A longitudinal study of patients initially treated with optimal complete dentures and ®nally with tissue integrated prostheses. Acta Odontologica Scandinavica, 45, 423. SCHNEEMAN , B.O. (1987) Dietary ®ber and gastro-intestinal function. Nutrition Review, 45, 129. Correspondence: Prof. Jean-Jacques Bon®l, Service d'Odontologie, HoÃpital Nord, Chemin des Bourrellys, 13915 Marseille Cedex 20, France. E-mail: jbon®[email protected]

ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 668±672