Oral health indicators for risk of malnutrition in elders

Abstract

Objectives

Using both clinical parameters and subjective measures of oral health, this study aimed to identify useful oral health indicators for the risk of malnutrition in elders.

Design

Cross-sectional study.

Setting

Five community centers run by non-government organizations (NGOs).

Participants

195 community dwelling elders (65 or above).

Measurements

An interviewer-administered questionnaire was completed to collect information on elders' socio-demographic background and oral health perception and practice. Their number of teeth, number of occluding tooth pairs, dental caries, and periodontal condition were examined. General Oral Health Assessment Index (GOHAI), an instrument for assessing oral health related quality of life (OHQoL), was used as a subjective measure of oral health. The elders' nutritional status was evaluated by using the Mini-Nutritional Assessment (MNA).

Results

The mean (SD) DFT was 3.3 (3.1). Over 60% of elders had periodontal pockets; 33% had fewer than 20 teeth and 6% were edentulous. The mean (SD) of occluding tooth pairs was 7.1 (4.8). The mean (SD) total GOHAI score was 56.4 (8.0); 60% reported negative impact of oral health on their quality of life. The mean (SD) MNA score was 25.0 (2.9); 30% had malnutrition or were at risk. After controlling for socio-demographic factors, none of the clinical indicators (dental caries, periodontal status, number of teeth, and number of occluding tooth pairs) were associated with risk of malnutrition (all p>0.05). Poorer OHQoL indicated a higher chance for malnutrition in both adjusted models (OR of 0.914; 95% CI of 0.850-0.982; p=0.014 and OR of 0.915; 95% CI of 0.852-0.984; p=0.017). Tooth loss and untreated decayed teeth (DT) were significant/marginally significant determinants of poor OHQoL.

Conclusion

Elders' tooth loss and unmet treatment need for dental caries were associated with compromised quality of life, which indicated increased likelihood for malnutrition..

Introduction

Adequate nutrition is considered as an essential aspect to preserve the general state of health and functional independence of the elders (1). Malnutrition decreases the immunological defenses, reduces physical functions, and can be a direct or indirect cause of morbidity and mortality among the elderly (1, 2). Risk of malnutrition increases with advancing age (3). Studies have shown that malnutrition is present in 2% to 51% of community-dwelling elders, depending on the definition used and the population studied (4, 5, 6).

The relationship between oral health and nutritional health of elders is attracting scholarly attention (7). Poor oral health renders elders vulnerable to food restrictions and in turn affects their nutritional status (8). It has been reported that aged people with compromised oral condition are more likely to avoid nutrient-rich food including whole grains, vegetables, fruits and meat, because most of them are hard, fibrous in texture and difficult to chew (9, 10). As a result, these elders risk reducing their intake of essential nutrients, such as vitamin C, calcium, non-starch polysaccharides and proteins (1, 11). In addition, oral conditions such as mucosal disorders, oral dryness and pain caused by periodontal disease (gum disease) or dental caries (tooth decay) may also induce difficulty in chewing, leading to reduction of food consumption and malnutrition (8).

Although there are strong reasons to believe oral health conditions are related to nutritional health, there is no consensus on which oral health parameters are indicative for malnutrition (12). Elder may suffer from a wide range of oral health problems, including but not limiting to dental caries, periodontal diseases, tooth loss, and xerostomia (13). Many clinical indicators, such as numbers of teeth, use of removable dentures, saliva flow, carious status, periodontal status, and mastication measurements, were used for measuring oral health in previous studies in regards to nutritional health (12, 14, 15). However, for almost every clinical indicator, studies supporting or refuting its association with elders' nutritional status co-exist. Such controversy applies to number of teeth, which is considered as a main indicator for masticatory efficiency (12). Although a cause and effect relationship between the number of teeth and nutritional state was postulated (16, 17) and some studies supported the association between number of teeth and malnutrition (10, 17, 18), such association was absent in some other studies (8, 14, 19).

Besides objective clinical measures, subjective oral health indicators such as self-perceived oral health (10, 20) and oral health related quality of life (OHQoL) were incorporated to assess the degree to which oral conditions affect physical and psychosocial functioning. This is incentivized by the updated understanding of health enshrined in the World Health Organization (WHO) charter (21). Since health is defined as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity, subject measures provide a very important perspective of the oral health impact on the individuals and the population. Research studies have revealed that subjective and objective measures do not necessarily point to the same conclusion and therefore serve as an important complement to clinical parameters (22, 23). The findings between nutritional status and the subjective oral health measures are however controversial; the association was supported in some studies (12, 14, 15) but was not found in some other studies (20, 24).

Using both clinical parameters and subjective measures of oral health, this study aimed to identify useful oral health indicators for the risk of malnutrition in elders.

Materials and Methods

Sample size calculation

The sample size was calculated for the main objective (i.e. identifying oral health factors associated with malnutrition) by using G* Power version 3.1.9.2. Targeting a statistical power of 0.8 and a significance level of 0.05 and estimating 20% subjects have malnutrition or are at risk (8), 140 subjects are needed to detect an effect size of 0.6.

Participants recruitment

The protocol of this study was reviewed by the Institutional Review Board, The University of Hong Kong and Hospital Authority Hong Kong West Cluster. An ethical approval was obtained (UW15-041). Participants were recruited through 5 community centers run by non-government organizations (NGOs) in Hong Kong. Posters were put up in the neighborhood communities to invite elders to participate. Staff of the centers also contacted their members through phone calls. The formal registration was done on the day of the field work. To be eligible to join, one must be (i) 65 years old or above; (ii) non-institutionalized; and (iii) able to communicate in spoken Cantonese, Mandarin or English. All participants gave their informed written consents for their participation. Patients with cognitive, neurological or psychiatric disorders were excluded.

Questionnaires

Each participant was asked to complete three questionnaires via a face-to-face interview by one of the four calibrated interviewers. The questionnaires were pretested to ensure relevance and clarity. Completing the questionnaires took approximately 15 minutes.

The first questionnaire was structured to collect information on participants' demographic background (age and gender), socioeconomic status (education level, previous occupation, and living condition), family support (living alone or with family members) and their oral health perception and practice.

The second questionnaire was the General Oral Health Assessment Index (GOHAI) (25), which is a subjective measure reflecting oral health related quality of life (OHQoL). The Chinese version of GOHAI validated in a previous study (26) was used. GOHAI is a 14-item instrument intended to report oral function problems and psychosocial impacts associated with oral diseases. It assesses three dimensions including physical functioning, pain and discomfort, and psychosocial functioning. Three items inquire on the participants' overall satisfaction of oral health (1-5 representing very unsatisfactory to very satisfactory). Nine items describe the negative impacts of oral health, whereas two are positively worded to discourage respondent acquiescence. The response to the impact question is collected by using the five point liker scale (1=always, 2=often, 3=sometimes, 4=seldom, and 5=never). Scores for the two questions initially worded positively were re-coded reversely during data processing so that all responses were in the same direction. The total GOHAI score (i.e. a sum of scores for 14 items) ranges from 14 to 70, with a higher total score indicating a better oral health-related quality of life.

The third questionnaire was the Mini Nutrition Assessment (MNA®), which is a nutritional assessment/screening tool increasingly employed in various settings with good sensitivity, specificity and reliability (27). The MNA consists of 18 items covering four dimensions (anthropometry, dietary assessment, global assessment and self-evaluation). To collect the anthropometry data, participants' weight and height were measured by a trained operator using a calibrated electronic scale and a wall-mounted height ruler. The body mass index (BMI) was then calculated as BMI = weight (kg) / [height (m)]2. The other three dimensions (dietary assessment, global assessment and self-evaluation) were assessed by asking specific questions. The MNA is separated by two steps: the MNA-SF (short form) intended to pre-screen those who are at the risk of malnutrition and the full MNA (long form) to further assess those “at risk of malnutrition. A final score is obtained by adding up the 18 items. The nutritional state is classified as ‘normal nutritional status' (scores from 24 to 30), ‘at risk of malnutrition' (scores from 17 to 23.5) and ‘malnourished' (scores lower than 17) (28, 29).

The whole set of questionnaires were pretested among 6 elders with different ages, genders, and backgrounds, to ensure the clarity and relevance of questions.

Oral examination

The oral health status of the participants was evaluated by three trained and calibrated examiners. Participants were examined on a supine position in portable dental chairs. The assessment was based on visual and tactile examination. CPITN probes and illuminated dental mirrors with disposable heads were used. No radiographs were taken. Around 10% participants were randomly selected for duplicate examinations by different examiners for assessing the inter-examiner reliability.

Participants' tooth status was recorded by following the method and criteria recommended by the World Health Organization (30). Dental caries was diagnosed at the cavitation level. The number of decayed and filled teeth (DFT) was calculated.

The periodontal status was assessed using the Community Periodontal Index (CPI) (30). The mouth was divided into six sextants. Ten index teeth were examined, including all first and second molars, upper right central and lower left central incisor. The two molars in each posterior sextant were paired for recording. In case where no index tooth was present in a sextant qualifying for examination, all the remaining teeth in that sextant were examined and the highest score was recorded.

The number of remaining teeth was calculated by excluding teeth missing due to caries and other reasons. The participants were categorized into: (1) < 20 teeth or (2) ≥ 20 natural teeth remaining; the latter was considered sufficient for maintaining a “functional dentition (30, 31). The number of occluding tooth pairs was also examined visually. Articulating paper was not used to examine the occlusal contact since the elders could not endure overly lengthy oral examination.

The examiner also asked the medical history of the elders and recorded if they had the following systemic diseases: heart disease, hypertension, blood disease, abnormal bleeding, kidney disease, diabetes, thyroid disease, tuberculosis, epilepsy and stroke.

Data analysis

Data were analyzed by using the Statistical Package for the Social Sciences (SPSS) software version 20.0. Participants' socio-demographic profile, systematic disease, oral health status, nutritional status, and oral health related quality of life were described. Logistic regression and analysis of covariance (ANCOVA) were performed to identify oral health indicators for malnutrition after controlling for socio-demographic variables and systemic disease. The Enter method, instead of the stepwise method, was used. The alpha level was set at 0.05. The collinearity between independent variables has been tested. Except for between “number of teeth and “number of occluding pairs, all the values of tolerance were well above 0.2, thus collinearity was ruled out (32). As for “number of teeth and “number of occluding pairs, they were entered separately in the multivariate models.

Results

Socio-demographic profiles

A total of 195 eligible elders participated in this study, including one who completed the questionnaire but did not receive oral examination. Their age ranged from 65 to 94 years, with a mean (SD) of 75.3 (6.7) years. There were 132 (67.7%) females and 63 (32.3%) males (Table 1).

Table 1.

Sociodemographic profiles of participating elders

n	%
Age
65-74 years	103	52.8
75-84 years	72	36.9
85 years or above	20	10.3
Age (Mean; SD)	75.3	6.7
Gender
Female	132	67.7
Male	63	32.3
Education level
Primary school or below	129	66.8
Secondary school	54	28.0
College / university	10	5.2
Previous Occupation Managerial or professional
Managers and administrators	8	4.2
Professionals	11	5.8
Self-employed	14	7.4
Clerical or skilled workers
Plant and machine operators and assemblers	32	16.9
Associate professionals	5	2.6
Clerical support workers	13	6.9
Service or labours
Sales	9	4.8
Unskilled laborers	40	21.2
Service industry	14	7.4
Others (mainly housewives)	43	22.8
Living condition
Public permanent housing	54	28.1
Tenement building	7	3.6
Village house	4	2.1
Home ownership scheme	26	13.5
Private housing estates	98	51.0
Others	3	1.6
Family support
Lives with family	156	80.8
Lives alone	37	19.2
Total	195	100.0

Two-thirds (66.8%) of the participants received primary education or had not attended any formal education. As for their previous occupation, 17.4% were “managers and administrators, “professionals or “self-employed; 26.4% were “plant and machine operators and assemblers, “associate professionals or “clerical workers; 33.4% were “sales workers, “service industry or “unskilled laborers; 22.8% were others (mainly housewives). About half (51.0%) lived in private housing estates. The majority (80.8%) of the participants lived with their families.

Systemic disease, nutritional status, oral health status and OHQoL

There were 138 (70.8%) participants with systemic diseases (Table 2). The nutritional status of 129 (70.1%) participants was normal (MNA score 24-30); 53 (28.8%) were at risk of malnutrition (MNA score 17-23.5); and 2 (1.1%) were malnourished (MNA score lower than 17). The mean (SD) MNA score was 25.0 (2.9).

Table 2.

Systemic disease, nutritional and dental status in elders

	n	%
Systemic disease*
Yes	138	70.8
No	57	29.2
Nutritional status
Normal (MNA score>24)	129	70.1
At risk of malnutrition (MNA score: 17-23.5)	53	28.8
Malnourished (MNA score<17)	2	1.1
Number of teeth
Edentulous	11	5.7
1-20 teeth	63	32.5
≥20 teeth	120	61.9
Dental caries
DFT=0	39	20.1
DFT>0	155	79.9
Periodontal status
Highest CPI score=0	0	0
Highest CPI score=1	2	1.2
Highest CPI score=2	73	37.8
Highest CPI score=3	81	41.9
Highest CPI score=4	37	19.2
	Mean	SD
MNA	25.0	2.9
DFT	3.3	3.1
Number of remaining teeth	19.1	8.8
Number of occluding tooth pairs	7.1	4.8

^*Systemic disease refers to any of the following diseases: heart disease, hypertension, blood disease, abnormal bleeding, kidney disease, diabetes, thyroid disease, tuberculosis, epilepsy and stroke.

The inter-examiner reliability for oral examination was high (kappa = 0.944 for DFT; weighted kappa = 0.817 for CPI; intraclass correlation coefficient (ICC) = 0.99 for number of teeth). Around 80% had DFT>0; 46.4% had untreated caries (DT>0). The mean (SD) of DFT and DT were 3.3 (3.1) and 1.0 (1.5), respectively. None of the participants had healthy gingiva (i.e. highest CPI as 0). Only 1.2% had a highest CPI score of 1 (i.e. gingival bleeding only). The highest CPI score of 2 (i.e. calculus), 3 (i.e. shallow pocket), and 4 (i.e. deep pocket) were recorded in 37.8%, 41.9%, and 19.2% of participants, respectively. Periodontal pockets were found in 61% participants. The mean (SD) number of remaining teeth was 19.1 (8.8); 120 (61.9%) participants had 20 or more teeth, while 11 (5.7%) were edentulous. The mean (SD) number of occluding tooth pairs was 7.1 (4.8).

Table 3 shows the distribution of participants' responses to the 14 GOHAI questions. Negative impacts of oral health were more reported on “limiting food choices, “difficulty in chewing hard food, “problem in eating, and “tooth sensitivity. When they were asked to rate their satisfactions and perception on their oral health, the majority chose “3 or “4 from a range of “1 (most unsatisfied) to “5 (most satisfied). The mean (SD) of total GOHAI score was 56.4 (8.0).

Table 3.

Oral health related quality of life (response to GOHAI questions)

	Always	Often	n (%) Sometimes	Seldom	Never
Limit food choices	26 (13.3)	9(4.6)	39(20.0)	13 (6.7)	108(55.4)
Able to bite hard food*	81(41.5)	21(10.8)	40(20.5)	24(12.3)	29(14.9)
Uncomfortable while swallowing	1(0.5)	4(2.1)	15(7.7)	9(4.6)	166(85.1)
Uncomfortable while eating	15(7.7)	12(6.2)	45(23.1)	6(3.1)	117(60.0)
Limit contact with people	2(1.0)	1(0.5)	5(2.6)	6(3.1)	181(92.8)
Use medication to relieve pain	2(1.0)	0(0.0)	13(6.7)	12 (6.2)	168(86.2)
Self-conscious about teeth/gums/denture	9(4.6)	6(3.1)	20(10.3)	14(7.2)	146(74.9)
Emotionally affected	9(4.6)	7(3.6)	18(9.2)	25(12.8)	136(69.7)
Able to speak clearly*	133(68.2)	18(9.2)	14(7.2)	5(2.6)	25(12.8)
Uncomfortable eating in front of others	5(2.6)	5(2.6)	10(5.1)	5(2.6)	170(87.2)
Sensitive to hot, cold, or sweet food	25(12.8)	16(8.2)	48(24.6)	17(8.7)	89(45.6)
	Very unsatisfactory	Unsatisfactory	Neutral	Satisfactory	Very satisfactory
Overall aesthetic satisfaction	17(8.8)	30(15.5)	72(37.1)	52(26.8)	23(11.9)
Perception of overall oral health	22(11.3)	33(16.9)	71(36.4)	54(27.7)	15(7.7)
Overall satisfaction of oral health Mean (SD) of total GOHAI score	19(9.8) 56.4 (8.0)	34(17.5)	69(35.6))	58(29.9)	14(7.2)

^*Items positively worded (i.e. positive impact of oral health)

Asssociation between elders' oral health, nutritional status and OHQoL

The MNA was recoded into two categories (normal and at risk/malnutrition) as dependent variable in logistic regression. The multivariate analysis identified several significant (p<0.05) or marginally significant (0.1>p≥0.05) socio-demographic risk factors for malnutrition, including older age, female gender, higher education level, and less skilled occupation previously engaged (Table 4). The dental indicators, including DFT and number of teeth/occluding tooth pairs were associated with malnutrition (p<0.05) in the unadjusted model, but this association was non-significant after controlling for confounding factor in the adjusted models. Poorer OHQoL was associated with a higher chance for malnutrition in the adjusted models; the OR (95% CI) was 0.914 (0.850-0.982) and 0.915 (0.852-0.984) when “number of teeth and “number of occluding pairs was entered respectively. This indicated that the odds for malnutrition decreased by 8.6% and 8.5% respectively with one unit increase of GOHAI score. Elders of older age were more likely to have malnutrition.

Table 4.

Risk factor for malnutrition of elders

	Unadjusted ModelOR (95% CI) for malnourished/ at risk of malnutrition	P	Adjusted Model 1OR (95% CI) for malnourished/ at risk of malnutrition	P	Adjusted Model 2OR (95% CI) for malnourished/ at risk of malnutrition	P
Age	1.093 (1.039-1.149)	0.001*	1.162 (1.054-1.282)	0.003*	1.161 (1.052-1.280)	0.003*
Systemic disease
No	1		1		1
Yes	1.452 (0.703-2.999)	0.314	1.205 (0.382-3.808)	0.750	1.224 (0.386-3.877)	0.731
Gender
Female	1		1		1
Male	0.596 (0.294-1.205)	0.150	0.371 (0.134-1.026)	0.056	0.365 (0.131-1.016)	0.054
Education
Primary and below		1		1		1
Secondary and above	1.289 (0.662-2.510)	0.455	3.280 (1.085-9.920)	0.035*	3.213 (1.061-9.729)	0.039*
Housing condition##
Basic	1		1	1
Moderate	0.708 (0.258-1.942)	0.503	0.414 (0.095-1.806)	0.241	0.415 (0.095-1.812)	0.242
Good	1.308 (0.641-2.671)	0.461	0.919 (0.328-2.572)	0.872	0.918 (0.327-2.574)	0.871
Past occupation
Service/labour		1		1		1
Clerical/skilled workers	0.858 (0.391-1.883)	0.703	0.754 (0.240-2.370)	0.629	0.762 (0.243-2.387)	0.640
Managerial/professional	0.284 (0.095-0.849)	0.024*	0.287 (0.078-1.060)	0.061	0.287 (0.078-1.061)	0.061
DFT (continuous)	0.850 (0.750-0.963)	0.011*	0.894 (0.744-1.074)	0.231	0.895 (0.745-1.075)	0.235
Periodontal status
Bleeding or calculus only	1		1	1
Shallow pocket (4-5mm)	0.810 (0.381-1.721)	0.584	1.171 (0.387-3.543)	0.780	1.188 (0.392-3.597)	0.761
Deep pocket (≥6mm)	0.818 (0.312-2.149)	0.684	0.885 (0.204-3.842)	0.870	0.901 (0.207-3.919)	0.889
Number of remaining teeth	0.950 (0.916-0.985)	0.006*	0.985 (0.914-1.061)	0.689	-	-
Number of occluding tooth pairs pairs	0.903 (0.844-0.966)	0.003*	-	-	0.969 (0.865-1.086)	0.590
Total GOHAI score	0.946 (0.908-0.985)	0.007*	0.914 (0.850-0.982)	0.014*	0.915 (0.852-0.984)	0.017*

The results were obtained through multiple logistic regression using nutritional status (well-nourished vs. malnourished/at risk) as a dependent variable. The independent variables are shown above. “Family support was not included as an independent variable since it is an item in the MNA questionnaire. “Number of teeth and “number of occluding tooth pairs were entered separately in two models; ## Housing condition was classified into basic (public permanent housing, tenement building), moderate (home ownership scheme, village house), and good (private housing estates); * Factors significantly associated with risk of malnutrition

In order to further identify what factors determine OHQoL, analysis of covariance (ANCOVA) was performed (Table 5). When “number of teeth was entered, an association was found with OHQoL (p<0.001), whereas DFT and CPI were not associated with OHQoL (p>0.05) (Model 1). When replacing DFT with DT in the same analysis, it was shown that both DT and number of teeth were significant determinants of OHQoL (p=0.043 and p<0.001, respectively) (Model 2). When “number of occluding tooth pairs was entered (Model 3 and Model 4), the results revealed that higher number of occluding tooth pairs were associated with better OHQoL (all p<0.001) and DT appeared to be a marginally significant factor affecting OHQoL (p=0.052).

Table 5.

Determinants of oral health related quality of life (total GOHAI)

		Model 1B (95% CI)	p	Model 2B (95% CI)	p	Model 3B (95% CI)	p	Model 4B (95% CI)	p
Age (years)	Continuous	0.280 (0.014, 0.546)	0.039*	0.259 (0.006, 0.512)	0.045*	0.284 (0.016, 0.551)	0.038*	0.261 (0.006, 0.517)	0.045*
Systemic disease	No	0	0	0	0
	Yes	-2.196 (-5.608, 1.216)	0.205	-1.998 (-5.334, 1.338)	0.238 -	2.473 (-5.906, 0.960)	0.156	-2.257 (-5.619, 1.106)	0.186
Gender	Female;	0	0	0	0
	Male	-0.096 (-3.196, 3.003)	0.951	0.313 (-2.756, 3.381)	0.840	0.135 (-2.993, 3.263)	0.932	0.500 (-2.598, 3.597)	0.750
Education	Primary or below	0	0	0	0
	Secondary or above	0.916 (-2.337, 4.170)	0.578	1.220 (-1.893, 4.332)	0.439	1.284 (-1.998, 4.566)	0.440	1.543 (-1.596, 4.683)	0.332
Housing condition	Basic	0	0	0	0
	Moderate	-1.469 (-5.660, 2.722)	0.489	-1.939 (-6.068, 2.190)	0.354	-1.516 (-5.731, 2.699)	0.477	-1.969 (-6.128, 2.190)	0.350
	Good	-2.136 (-5.354, 1.083)	0.191	-2.365 (-5.529, 0.799)	0.141	-2.154 (-5.391, 1.083)	0.190	-2.380 (-5.568, 0.807)	0.142
Past occupation	Service/labour	0	0	0	0
	Clerical/skilled workers	0.782 (-2.642, 4.206)	0.652	0.357 (-3.022, 3.735)	0.835	0.372 (-3.072, 3.815)	0.831	-0.004 (-3.401, 3.393)	0.998
	Managerial/professional	1.501 (-2.328, 5.330)	0.439	1.012 (-2.743, 4.767)	0.594	1.267 (-2.582, 5.117)	0.516	0.810 (-2.968, 4.588)	0.672
Family support	Live alone	0	0	0	0
	Live with family	0.973 (-2.571, 4.517)	0.587	0.659 (-2.832, 4.150)	0.709	0.994 (-2.571, 4.559)	0.582	0.688 (-2.829, 4.205)	0.699
	DFT Continuous	0.034 (-0.486, 0.554)	0.897	-	-	0.036 (-0.487, 0.559)	0.892	-	-
DT	Continuous	-	-	-1.047 (-2.058, -0.036)	0.043*	-	-	-1.016 (-2.040, 0.008)	0.052
Periodontal status	Bleeding or calculus only	0	0	0	0
	Shallow pocket (4-5mm)	2.007 (-1.362, 5.376)	0.240	2.140 (-1.123, 5.403)	0.196	2.115 (-1.268, 5.498)	0.218	2.256 (-1.027, 5.539)	0.176
	Deep pocket (≥6mm)	2.576 (-1.640, 6.791)	0.229	2.389 (-1.577, 6.356)	0.235	2.493 (-1.760, 6.745)	0.248	2.336 (-1.669, 6.342)	0.250
Number of teeth	Continuous	0.468 (0.255, 0.680)	<0.001*	0.425 (0.213, 0.638)	<0.001*	-	-	-	-
Number of occluding tooth pairs	Continuous	-	-	-	-	0.695 (0.366, 1.024)	<0.001*	0.623 (0.292, 0.954)	<0.001*
Intercept		24.819 (3.328, 46.310)	0.024	28.679 (8.403, 48.956)	0.006	29.147 (8.007, 50.288)	0.007	32.722(12.865,52.578)	0.001

Four models were constructed by using analysis of covariance (ANCOVA) to determine factors related to OHQoL, with total GOHAI as the dependent variable. The independent variables are shown above. “Number of teeth and “number of occluding tooth pairs were entered separately in different models;

^*Factors significantly associated with oral health related quality of life (GOHAI score)

Discussion

As similarly shown in other studies among elders, participants in this study were suffering from various dental problems. Dental caries, periodontal pockets, tooth loss, and negative impact on their quality of life were common. Nevertheless, none of the clinical parameters (DFT, CPI, and number of teeth/occluding pairs) was significantly associated with elders' risk for malnutrition. This was in accordance with a previous study which showed that the relationship between specific oral health indicators (e.g. number of natural teeth, masticatory ability, or wearing denture) and nutritional health diminished in multivariate analysis (33). Dental caries or periodontal diseases, if not serious, may not definitely interrupt oral functions. In addition, masticatory function may be compromised only when the number of teeth was very low (34). In contrast to the clinical parameters, the subjective oral health measure OHQoL appeared to be a significant indicator for malnutrition in elders. Poor OHQoL (i.e. low total GOHAI score) significantly increased the chance of malnutrition. As a global measure of oral health, OHQoL reflects the overall functional loss caused by multiple dental problems and a full spectrum of impacts imposed by various oral conditions. Therefore, OHQoL serves as an important complement to clinical parameters (35) and may better explain the impact of oral health on nutritional status and general well-beings. Participants in this study reported a moderate level of GOHAI score (56.4), as benchmarked against recommend cut-off points (25). Their OHQoL appeared to be better than previously found in institutionalized elders in Hong Kong (36), which nearly reached the lowest acceptable level (i.e. GOHAI score of 50).

Among the clinical indicators (DFT, CPI, number of teeth, and number of occluding tooth pairs), only “number of teeth and “number of occluding tooth pairs were identified as a significant factors determining OHQoL. This may be because that caries or periodontal disease may not necessarily compromise oral functions unless there are severe symptoms or a resulting tooth loss. In this study, although DFT did not appear to be a determinant of OHQoL, a significant or marginally significant association between dental caries and OHQoL was found when DFT was replaced by DT. This finding is in line with a previous study (37), but contradictory to the result of another study, which indicated an association between OHQoL and DMFT, but not DT (38). Calculated as number of untreated decayed teeth, DT is a reflection of unmet treatment need. Dental caries is ubiquitous among elders and very few people are free from dental caries when they enter a senior age. Nevertheless, our finding implies that, as far as their treatment need for caries is met, the negative impact on their quality of life can be controlled. There is abundant evidence of a discrepancy between perceived oral treatment need and dental service use by older people (39). Improving elders' access to and affordability for dental care is among the important enabling factors to address this issue (40).

Although one may expect the oral health conditions of the elders to deteriorate as they age, in this study an increase in age was found to be associated with a higher GOHAI score (i.e. a better OHQoL). This finding echoes the result of a study conducted in elders in China (41) and the analysis of national survey data in UK and Australia (42). When assessing quality of life, people may compare their status with what they see as acceptable for a given age (43). The current finding might be a reflection that older people become more tolerant towards oral health problems, trivialize oral health in the general context of their impaired health, and lower their expectation towards oral health (44).

The findings of this study can be better interpreted when the strength and limitation are taken into considerations. Wellestablished and validated instruments (GOHAI and MNA) were adopted to measure OHQoL and nutritional status of the elders. GOHAI has been translated into different languages and validated in various populations (26, 45, 46). It's considered as a suitable tool for measuring functional and psychosocial impacts of oral health in elders (47). MNA is an important tool for grading nutritional status and evaluating malnutrition risk (28). As a scale measuring several dimensions, it outperforms many single parameters and has good sensitivity, specificity and reliability (27). Although haematological data is a possible option for evaluating nutritional status, it did not turn out to be practical since the participants were unwilling to accept invasive procedures. This is a cross-sessional study and the limitation of such design applies. Our findings could only suggest associations; no causal relationships between elders' oral health, quality of life and nutritional status could be determined.

In conclusion, elders' tooth loss and unmet treatment need for dental caries were associated with compromised quality of life, which indicated increased likelihood for malnutrition. This study also showed that the usual dental variables gathered in daily clinical practice, such as dental caries, periodontal status and number of teeth, and number of occluding pairs, do not reliably reflect the possible impact on the nutritional status of the elders. Instead, the subject oral health measure GOHAI is a useful indicator for risk of malnutrition. While subjective measures such as GOHAI are increasingly incorporated into dental assessment, it helps to identify elders who are vulnerable to malnutrition, so that effective interventions and nutritional care can be delivered. Malnutrition is a complex and multifactorial problem requiring a holistic approach for prevention and treatment. A greater integration between dental and nutritional care will be beneficial for the health promotion of the elderly population.

Conflict of Interest Disclosures

The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.

Ethical standards

The protocol of this study was reviewed by the Institutional Review Board (IRB), The University of Hong Kong and Hospital Authority Hong Kong West Cluster. An ethical approval was obtained (UW 15-041). Informed written consent was obtained from each participant.