In total, 111 nursing home residents participated in this study, of which 84 had dementia and 27 were without dementia. The recruitment flowchart is reported in Figure 1. Descriptive characteristics of all participants and of residents with and without dementia separately are given in Table 1. Of the total sample, 62.2% were female and the average age was 83.9 (SD 7.95) years old.

The EQ-5D for proxy was used in 63 participants with dementia. There were no significant differences between people with and without dementia regarding demographics, comorbidity (CCI), quality of life (EQ-5D), and oral-health related quality of life (OHIP-14). Functional ability (Barthel Index) was significantly poorer in the dementia group (median (IQR) dementia group: 10 (10-30) vs non-dementia group: 30 (15-40); P = .003).

3.1 Oral function

Factors regarding oral function of all participants and of residents with and without dementia separately are given in Table 2. Of all participants, 69.4% were dentate and the median number of OU was 0. Regarding functional categories, 48.6% had 0-2 OU without dentures, and only 23.4% had 3 OU or more. Residents with dementia had significantly fewer natural teeth (in dentate participants) (P = .021), less OU (P = .009), and poorer outcome regarding functional categories (P = .001). Subjective swallowing quality was indicated as good by 73% of the participants, while 52.3% indicated the subjective chewing quality as good. Orofacial pain was significantly more prevalent in residents with dementia (48.8%), than those without dementia (14.8%) (P = .002).

TABLE 2. Oral function and nutritional status of all participants and of residents with and without dementia separately Total (n = 111) Dementia (n = 84) No dementia (n = 27) Test value P-value Dental status, dentate [n (%)] 77 (69.4%) 59 (70.2%) 18 (66.7%) X2(1) = 0.123 .726 Present teeth (in dentate P) median (IQR) 16 (7-22) 14 (6-21) 22 (12.75-24.25) Z = −2.302 .021* OU median (IQR) 0 (0-4) 0 (0-3) 4 (0-7) Z = −2.625 .009** Functional categories [n (%)] 0-2 OU no dentures 54 (48.6%) 48 (57.1%) 6 (22.2%) Z = −3.283 .001** 0-2 OU + dentures 31 (27.9%) 21 (25.0%) 10 (37.0%) 3-5 OU 13 (11.7%) 8 (9.5%) 5 (18.5%) ≥ 6 OU 13 (11.7%) 7 (8.3%) 6 (22.2%) Subjective swallowing quality [n (%)] Good 81 (73.0%) 59 (70.2%) 22 (81.5%) X2(2) = 3.084 .214 Moderate 25 (22.5%) 22 (26.2%) 3 (11.1%) Bad 5 (4.5%) 3 (3.6%) 2 (7.4%) Subjective chewing quality [n (%)] Good 58 (52.3%) 40 (47.6%) 18 (66.7%) X2(2) = 4.180 .124 Moderate 38 (34.2%) 30 (35.7%) 8 (29.6%) Bad 15 (13.5%) 14 (16.7%) 1 (3.7%) Orofacial pain [n (%)] 45 (40.5%) 41 (48.8%) 4 (14.8%) X2(1) = 9.796 .002** SXI median (IQR) 6 (5-11.75) 7 (5.5-13) 6 (5-10) Z = −1.220 .222 Soft diet [n (%)] 39 (35.8%) 35 (42.7%) 4 (14.8%) X2(1) = 6.865 .009** Weight in kg M, SD (range) 61.5, 16.9 (31.4-114) 59.9, 16.5 (31.4-114) 66.5, 17.3 (38.9-99.1) t (109) = −1.805 .074 BMI M, SD (range) 22.5, 5.2 (12-37) 22.0, 5.0 (12-37) 24.0, 5.6 (14-37) t (109) = −1.787 .077 Decline in food intake [n (%)] Severe decrease 2 (1.8%) 2 (2.4%) 0 (0.0%) Z = −1.651 .099 Moderate decrease 37 (33.3%) 31 (36.9%) 6 (22.2%) No decrease 72 (64.9%) 51 (60.7%) 21 (77.8%) Weight loss last 3 months [n (%)] >3 kg 8 (7.2%) 5 (6.0%) 3 (11.1%) Z = −0.303 .762 1-3 kg 28 (25.2%) 22 (26.2%) 6 (22.2%) No weight loss 63 (56.8%) 48 (57.1%) 15 (55.6%) Unknown 12 (10.8%) 9 (10.7%) 3 (11.1%) MNA-SF total score M, SD (range) 8.7, 2.7 (2-14) 8.3, 2.7 (2-13) 10.4, 2.0 (6-14) t (109) = −3.128 .002** MNA-SF [n (%)] Normal nutritional status 21 (18.9%) 12 (14.3%) 9 (33.3%) X2(1) = 8.382 .015* At risk of malnutrition 51 (45.9%) 37 (44.0%) 14 (51.9%) Malnourished 39 (35.1%) 35 (41.7%) 4 (14.8%) Abbreviations: P, participants; IQR, interquartile range; M, mean; SD, standard deviation; OU, occlusal units; SXI, summated xerostomia inventory; BMI, body mass index in kg/m2; MNA-SF, mini nutritional assessment short form; X2, chi square test; t, independent sample t test, Z, Mann-Whitney U test; N/A, not applicable. * P < .05, ** P < .01. 3.2 Nutritional status

Information on nutritional status of all participants and of residents with and without dementia separately are shown in Table 2. Significantly more residents with dementia (42.7%) were on a soft diet, compared with residents without dementia (14.8%) (P = .009). The mean outcome of the total score of the MNA-SF for residents with dementia was 8.3 (SD 2.7) and was significantly lower than the mean of 10.4 (SD 2.0) of residents without dementia (P = .002). Residents with dementia scored significantly poorer on the screening score of the MNA-SF (P = .015). According to the MNA-SF, of the residents with dementia, 41.7% were malnourished and 44.0% were at risk of malnutrition. In residents without dementia, 51.9% were at risk of malnutrition and 14.8% were malnourished.

In Tables 3 and 4, the associations between oral function factors, nutritional status and quality of life in nursing home residents are shown. Being dentate (P = .015), lower number of teeth (P = .020), lower number of OU (P = .007), poorer functional category (P = .012), poorer swallowing quality (P < .001), and poorer chewing quality (P < .001) were significantly associated with the presence of orofacial pain.

TABLE 3. Spearman correlation between oral function, nutritional status and quality of life in nursing home residents 1 2 3 4 5 6 7 8 9 1 Present teeth - 2 OU r = .78** - 3 Functional categories r = .60** r = .80** - 4 Swallowing quality r = −.13 r = −.14 r = −.11 - 5 Chewing quality r = −.18 r = −.18 r = −.20* r = .41** - 6 SXI r = −.05 r = −.15 r = −.22 r = .41** r = .37** - 7 MNA-SF r = .42** r = .44** r = .35** r = −.41** r = −.54** r = −.35** - 8 EQ-5D r = .00 r = .02 r = .13 r = −.43** r = −.43** r = −.44** r = .45** - 9 OHIP-14 r = −.22 r = −.23 r = −.26 r = .17 r = .42** r = .37* r = −.32* r = −.31* - Abbreviations: n, 111; OU, occlusal units; SXI, summated xerostomia inventory; MNA-SF, mini Nutritional assessment short form; EQ-5D, euroqol 5 dimension; OHIP, oral health impact profile; r, spearman correlation coefficient. * P < .05, ** P < .01. TABLE 4. Association between dental status, orofacial pain and being on a soft diet and oral function, nutritional status and quality of life in nursing home residents Dental status Orofacial pain Soft diet P-value P-value P-value Dental status [X2] - - - Orofacial pain [X2] .015* - - Soft diet [X2] .068 <.001** - Present teeth [Z] - .020* .215 OU [Z] - .007** .747 Functional categories [Z] .159 .012* .051 Swallowing quality [Z] .591 <.001** <.001** Chewing quality [Z] .122 <.001** <.001** SXI [Z] .697 .008** .002** MNA-SF [t] .687 .001** <.001** EQ-5D [t] .802 .003** <.001** OHIP-14 [t] .100 .001** .481 Abbreviations: n, 111; OU, occlusal units; SXI, summated xerostomia inventory; MNA-SF, mini nutritional assessment short form; EQ-5D, euroqol 5 dimension; OHIP, oral health impact profile; t, independent sample t test; X2, chi square test; Z, Mann-Whitney U test. * P < .05, ** P < .01.

Lower number of teeth (P < .001), lower number of OU (P < .001), poorer functional category (P < .001), poorer swallowing quality (P < .001), poorer chewing quality (P < .001), presence of orofacial pain (P = .001), xerostomia (P = .005) and being on a soft diet (P < .001) were significantly associated with a poorer nutritional status. A poorer quality of life, according to the EQ-5D, was significantly associated with poorer swallowing quality (P < .001), poorer chewing quality (P < .001), presence of orofacial pain (P = .003), xerostomia (P < .001), being on a soft diet (P < .001), and poorer nutritional status (P < .001). A poorer oral health-related quality of life, according to the OHIP-14, was significantly associated with poorer chewing quality (P = .005), presence of orofacial pain (P = .001), xerostomia (P = .014), and poorer nutritional status (P = .039).

In Table 5, the univariate and multiple linear regression between oral function factors and nutritional status (MNA-SF) are reported. Functional categories, quality of swallowing, quality of chewing, orofacial pain, SXI and being on a soft diet were eligible for the multivariable regression model (P < .10). After adjusting for the other oral function factors, only quality of chewing (P = .010) remained significantly associated with nutritional status. This model explained 28% (R2 = 0.28) of the variance in nutritional status.

TABLE 5. Univariate and multiple linear regression of the relationship between oral function factors and nutritional status (MNA-SF) in nursing home residents R R 2 Coefficient 95% CI T P-value Univariate regression Dental status 0.04 0 -0.22 -1.32; 0.88 -0.4 .687 Functional categories 0.35 0.12 0.9 0.44; 1.36 3.85 <.001a Quality of swallowing 0.41 0.17 -1.97 -2.81; −1.14 -4.67 <.001a Quality of chewing 0.54 0.29 -2.02 -2.62; −1.42 -6.66 <.001a Orofacial pain 0.32 0.1 -1.75 -2.73; −0.77 -3.55 .001a SXI 0.32 0.1 -0.13 -0.23; −0.03 -2.63 .011a Soft diet 0.45 0.2 -2.46 -3.40; −1.52 -5.17 <.001a Multiple regression Functional categories 0.53 0.28 0.45 -0.02; 0.92 1.93 .059 Quality of swallowing -0.83 -1.99; 0.33 -1.43 .159 Quality of chewing -1.27 -2.22; −0.31 -2.66 .010* Orofacial pain 0.68 -0.59; 1.93 1.08 .285 SXI -0.04 -0.16; 0.08 -0.67 .508 Soft diet 0.39 -1.17; 1.95 0.5 .619 Abbreviations: n, 111; SXI, summated xerostomia inventory; MNA-SF, mini nutritional assessment short form; CI, confidence interval. a p < 0.10 and included in multivariable regression model. * P < .05.