Shortened dental arch and body mass index in adults 45–65 years of age: results from National Health and Nutrition Examination Survey 2005–2008

Abstract

Objective: Many people have dental arches with unrestored edentulous areas posterior to natural teeth. One dental pattern is the shortened dental arch (SDA). As a result of the lack of teeth, individuals with an SDA may eat a restricted diet, including soft, highly processed foods. Such diets may increase the risk of being overweight or of obesity. We examined whether there was an association between SDA and body mass index (BMI) in adults 45–65 years of age. Methods: The data for this study were US National Health and Nutrition Examination Survey (NHANES) 2005–2008 merged files. There were 5,773 eligible participants. The data were examined for frequencies, and the results were analysed using the chi-square test and logistic regression. Results: There were 69.3% participants with a shortened mandibular arch and a BMI of ≥ 25 compared with 71.8% of participants who had a complete mandibular dental arch and a BMI of ≥ 25 (P = 0.7246). There were 70.6% of participants with a shortened maxillary arch and a BMI of ≥ 25 compared with 71.9% of participants who had a complete maxillary dental arch and a BMI of ≥ 25 (P = 0.8859). The adjusted odds ratio for shortened mandibular dental arch was 0.70 (95% CI: 0.46–1.08) for a BMI of ≥ 25 as compared with individuals with a BMI < 25. The adjusted odds ratio for shortened maxillary dental arch was 1.06 (95% CI: 0.63–1.78) as compared with individuals with a BMI < 25. Conclusions: The research hypothesis that an SDA was related to higher BMI, and the corollary that restored or complete dentition had better odds of a lower BMI, were not supported.

INTRODUCTION

The World Health Organization’s researchers estimate that 30% of the world’s adult population, 65–74 years of age, is edentulous¹. There is a downward trend in edentulism, according to birth cohort, in the USA². In 1957–1958, 18.9% of the population, ≥15 years of age, was edentulous²; by 2009–2012, this had decreased to 4.9%². This decline is primarily a result of the death of cohorts born before 1940². The projected prevalence of edentulism in 2050 is 2.6% of the population, largely because of the death of cohorts born after 1940².

As a result of the downward trend of edentulism, a variety of dental arch tooth patterns are becoming more common. Many people have dental arches with unrestored edentulous areas posterior to natural teeth. This pattern is referred to as a shortened dental arch (SDA) under the following conditions:

• •There are occluding anterior teeth (six occluding pairs)
• •There are unilateral or bilateral edentulous area(s) posterior to the most distal tooth (unilaterally) or teeth (bilaterally)
• •The edentulous area(s) has (have) not been restored³.

There is no agreed upon number of missing posterior teeth in the definition of SDA. In the 1980s, Kayser recognised four occluding pairs as the point between adequate oral function and inadequate oral function4., 5.. Some researchers have defined a unilateral SDA as the presence of an ipsilateral canine, premolar or first molar as the most distal tooth anterior to the edentulous space and a bilateral SDA as having canines, premolars or first molars, or various combinations of these teeth, as edentulous spaces³. Other researchers chose to use 10 occluding pairs of naturally occurring and replacement teeth to define the SDA⁶. In a recent study, investigators chose to develop four scenarios (missing second molars; missing molars; missing molars and second premolars; and missing molars and premolars)⁷.

A systematic review found SDA was defined variously with different combinations of missing premolars and molars in the selected studies⁸. In another review of the SDA literature found studies in which the SDA definition was zero to eight occluding pairs, 10 occluding pairs or four occluding pairs⁹. They reported that there was adaptive capacity for oral function in SDA if at least four symmetrically placed occluding pairs were present⁹. The suggestion was that, in the USA and other industrialised countries, food industries have processed food to the extent that chewing is less important³.

The possibility exists that individuals with an SDA may eat a restricted diet that includes many soft, highly processed foods. Often, such diets consist of refined, calorie-laden items which increase the risk of being overweight or obese. There are 34.9% (78.6 million) adults in the USA – 32.2% of men and 35.5% of women – who are obese10., 11.. Heart disease, stroke, diabetes and some types of cancers are associated with being overweight¹⁰. The medical costs of obesity in the U.S. in 2008 were $147 billion¹⁰.

As being obese or overweight is related to many health problems, and it is unknown if an SDA increases the likelihood of obesity and being overweight, the researchers in this study examined SDA and body mass index (BMI) to determine if there is an association of SDA and being overweight or obese in a group of individuals 45–65 years of age. Individuals 45–65 years of age were more likely to have been exposed to community water fluoridation and to toothpastes containing fluoride during their youth than were individuals > 65 years of age and are more likely to be part of the trend of adults who maintain their teeth as they age.

METHODS

This study received acknowledgement from the West Virginia University Institutional Review Board (protocol number 1505675830) as non-human subject, secondary data-analysis research not requiring Institutional Review Board approval. The research was conducted in full accordance with the World Medical Association Declaration of Helsinki. The data used in this study were de-identified and anonymised before analysis and are publicly available US National Health and Nutrition Examination Survey (NHANES) merged files containing data from 2005 to 2008. The NHANES is a complex, national survey of non-institutionalised US residents. The NHANES researchers utilised questionnaires, examinations and laboratory studies. They over-sampled for adequate sample sizes to represent population subgroups in data analyses. Dental examiners in 2005–2008 evaluated tooth count and the presence or absence of removable dental prostheses (dental examinations in previous and subsequent years did not include data on dental prostheses).

The sample population was limited to individuals with complete data on tooth count (to determine SDA status, the presence of a complete arch, the presence of edentulism or the presence of other arch patterns), age 45–65 years, complete data on prosthetic replacement (yes or no) and complete data on BMI (<25 or ≥25). Participants were limited to the birth cohort who were 45–65 years of age at the time of the study and therefore considered to be more likely to be part of the trend of adults who maintain their teeth as they age. Older adults were excluded as having the potential to over-represent edentulism, and younger adults were excluded as having the potential to over-represent individuals with all teeth present. There were 5,773 eligible participants.

The key dependent variable was BMI. Overweight and obesity were defined as a BMI of ≥25. Normal weight was defined as a BMI of < 25. The key independent variable was dental arch. Dental arches were defined as shortened dental arches, edentulous dental arches, complete dental arches and other dental arches. Eight forms of SDA have been reported:

• •‘Slightly shortened’ dental arch has at least one missing first molar with a distal unrestored edentulous area
• •‘Shortened dental arch I’ has a second premolar with a distal unrestored edentulous area on one side of the arch and the contralateral side has a first, second or third molar present
• •‘Shortened dental arch II’ has a second premolar with a distal unrestored edentulous area on one side of the arch and the contralateral side has a first or second premolar with a distal unrestored edentulous area
• •‘Extreme shortened dental arch I’ has first premolars bilaterally with distal unrestored edentulous areas
• •‘Extreme shortened dental arch II’ has a canine with a distal unrestored edentulous area on one side of the arch and the contralateral side has a canine or first premolar with a distal unrestored edentulous area
• •‘Asymmetric shortened dental arch I’ has a first premolar with a distal unrestored edentulous area on one side of the arch and the contralateral side has a first, second or third molar with a distal unrestored edentulous area
• •‘Asymmetric shortened dental arch II’ has a canine with a distal unrestored edentulous area on one side of the arch and the contralateral side has a first, second or third molar with a distal unrestored edentulous area
• •‘Asymmetric shortened dental arch III’ has a canine with a distal unrestored edentulous area on one side of the arch and the contralateral side has a second premolar with a distal unrestored edentulous area³.

In this study, the eight forms of SDA, in combination and individually, and second premolar occlusion with occluding anterior teeth, were studied in association with BMI. Other variables were selected for adjusted analyses as they had previously been shown to be associated with BMI. These included sex (female or male), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American or Other), age (45–54 years or 55–65 years), education (less than high school, high school graduate, some college/technical school or college/technical school graduate), family income to poverty ratio (0 to <1.25, 1.25 to <2.00, 2.00 to <4.00 or ≥4.00), alcohol consumption (non-use, moderate use or heavy use), smoking status (current, former or never) and moderate exercise (yes or no).

The data were analysed using SAS (version 9.3; SAS, Cary, NC, USA) for frequency, and using the Rao–Scott chi square test and logistic regression on BMI for complex surveys. Weights were used in the analyses.

RESULTS

The sample size was 5,773 participants (Table 1). There were 1,311 participants with a complete maxillary dentition. There were 304 participants with a maxillary SDA using the definitions in which a unilateral SDA has the most distal tooth anterior to the edentulous space being an ipsilateral canine, premolar or first molar and a bilateral SDA has the edentulous spaces being canines, premolars, first molars or various combinations of these teeth. The sample of 304 participants with a maxillary SDA included the entries in Table 1 in which n < 10. The sample also included 1,279 participants with an edentulous maxillary arch and 2,879 participants with other types of maxillary arches. (The limited sample size of the subtypes of SDA, and of SDA defined as ‘anterior and premolar occlusion’, prohibited further analysis beyond description for these types.)

Table 1.

Sample characteristics of dental arches, National Health and Nutrition Examination Survey 2005–2008

Characteristic	Number	Weighted percentage
Maxilla
Scenario 1
Compete dental arch	1,311	30.9
Slightly shortened	192	4.6
Shortened dental arch I	36	0.6
Shortened dental arch II	34	0.5
Extreme shortened dental arch I	*
Extreme shortened dental arch II	20	0.2
Asymmetric shortened dental arch I	*
Asymmetric shortened dental arch II	*
Asymmetric shortened dental arch III	*
Edentulous	1,279	17.1
Other	2,879	45.7
Scenario 2
Compete dental arch	1,311	30.9
Second premolars with anterior occlusion	*
Edentulous	1,279	17.1
Other	3,173	51.2
Mandible
Scenario 1
Complete dental arch	1,343	32.3
Slightly shortened	153	3.4
Shortened dental arch I	47	0.7
Shortened dental arch II	73	1.2
Extreme shortened dental arch I	22	0.1
Extreme shortened dental arch II	320	4.8
Asymmetric shortened dental arch I	*
Asymmetric shortened dental arch II	*
Asymmetric shortened dental arch III	*
Edentulous	837	10.9
Other	2,967	46.2
Scenario 2
Complete dental arch	1,343	32.3
Second premolars with anterior occlusion	31	1.3
Edentulous	837	10.9
Other	2,967	55.5

^*n <10.

The sample included 1,343 participants with a complete mandibular dentition. There were 626 participants with a mandibular SDA using the definition given in a previous paragraph. The sample of 626 participants with a mandibular SDA included the entries in Table 1 in which n < 10. The sample also included 837 participants with an edentulous mandibular arch and 2,967 participants with other types of mandibular arches. (The limited sample size of the subtypes of SDA, and of SDA defined as ‘anterior and premolar occlusion’, prohibited further analysis beyond description of these types.)

There were 1,543 participants who had a BMI of <25 and 4,230 participants who had a BMI of ≥25. The complete sample description is presented in Table 2.

Table 2.

Sample characteristics, National Health and Nutrition Examination Survey 2005–2008

Variable	Number	Weighted percentage	Standard error
Body mass index
0–24	1,543	28.5	0.8
≥25	4,230	71.5	0.8
Mandibular arches
Complete dental arch	1,343	32.3	1.4
SDA*	626	10.6	0.4
Edentulous	837	10.9	0.7
Other	2,967	46.2	0.9
Maxillary arches
Compete dental arch	1,311	30.9	1.5
SDA*	304	6.3	0.5
Edentulous	1,279	17.1	1.1
Other	2,879	45.7	0.9
Sex
Female	2,893	52.8	0.8
Male	2,880	47.2	0.8
Race/ethnicity
Non-Hispanic white	3,100	77.1	2.1
Non-Hispanic black	1,235	10.2	1.3
Mexican American	831	4.8	0.7
Other	607	7.9	1.0
Age
45–54 years	795	52.3	2.4
55–65 years	1,100	47.7	2.4
Education
Less than high school	1,845	19.9	1.2
High school graduate	1,428	26.7	0.9
Some college/technical school	1,395	27.8	1.0
College/technical school graduate	1,096	25.6	1.5
Family income to poverty ratio
0 to <1.25	780	15.4	1.7
1.25 to <2.00	599	15.1	1.0
2.00 to <4.00	794	27.1	1.2
≥4.00	832	42.3	2.5
Alcohol consumption
Non-use	1,768	33.8	1.9
Moderate use	1,899	48.0	1.8
Heavy use	812	18.3	1.0
Smoking status
Current	1,084	19.3	1.0
Former	1,943	32.4	0.8
Never	2,740	48.4	1.2
Moderate exercise
Yes	1,602	35.1	1.8
No	4,171	64.9	1.8

^*Shortened dental arch (SDA) values defined as occluding anterior teeth and unrestored edentulous areas posterior to the most distal ipsilateral canine, premolar or first molar in a unilateral situation or unrestored edentulous areas posterior to the most distal canines, premolars or first molars (or combination) in a bilateral situation.

The weighted percentages from Rao–Scott chi square bivariate analysis showed that 70.6% of participants with a maxillary SDA, 72.4% of participants with an edentulous maxillary arch, 71.9% of participants who had all of their maxillary teeth and 71.0% of participants with other types of maxillary teeth patterns, were overweight or obese (P = 0.8859).

Moreover, 69.3% of participants with a mandibular SDA, 72.5% of participants with an edentulous mandibular arch, 71.8% of participants who had all of their mandibular teeth and 71.6% of participants with other types of mandibular teeth patterns, were overweight or obese (P = 0.7246). The frequencies and the results of the other variables in the bivariate analyses are presented, in detail, in Table 3.

Table 3.

Body mass index (BMI) status versus shortened dental arch (SDA) and other variables of interest: Rao–Scott chi square analysis, National Health and Nutrition Examination Survey, 2005–2008

Variable	Frequency BMI 0–24	Weighted row (%)	Frequency BMI ≥25	Weighted row (%)	(P-value) standard error
Mandibular arches					(0.7246)
Complete dental arch	354	28.2	989	71.8	1.3
SDA*	179	30.7	447	69.3	2.5
Edentulous	246	27.5	591	72.5	1.6
Other	764	28.4	2,203	71.6	1.1
Maxillary arches					(0.8859)
Complete dental arch	343	28.1	969	71.9	1.6
SDA*	77	29.4	227	70.6	2.8
Edentulous	374	27.6	905	72.4	1.9
Other	749	29.0	2,130	71.0	0.9
Sex					(<0.0001)
Male	709	22.8	2,171	77.2	1.1
Female	834	33.6	2,059	66.4	0.9
Race/ethnicity					(<0.0001)
Non-Hispanic white	905	28.8	2,195	71.2	1.0
Non-Hispanic black	295	23.0	940	77.0	1.3
Mexican American	169	20.7	662	79.3	1.3
Other	174	37.0	433	63.0	3.1
Age					(0.5898)
45–54 years	211	29.8	584	70.2	2.4
55–65 years	268	28.4	832	71.6	1.7
Education					(0.0155)
Less than high school	473	27.0	1,372	73.0	1.6
High school graduate	380	28.2	1,048	71.2	1.2
Some college/technical school	354	26.4	1,041	73.6	1.3
College/technical school graduate	330	32.1	766	67.9	1.7
Family income to poverty ratio					(0.7160)
0 to <1.25	205	28.3	575	71.2	2.0
1.25 to <2.00	151	27.4	448	72.6	2.0
2.00 to <4.00	220	30.8	574	69.2	0.9
≥4.00	203	27.5	629	72.5	2.3
Alcohol consumption					(0.0651)
Non-use	420	26.5	1,348	73.5	1.3
Moderate use	552	30.2	1,347	69.8	1.0
Heavy use	215	27.2	597	72.8	2.2
Smoking status					(<0.0001)
Current	428	40.8	656	59.2	1.2
Former	448	23.7	1,495	76.3	1.0
Never	665	26.8	2,075	73.2	1.1
Moderate exercise					(0.2051)
Yes	441	29.6	1,161	70.4	1.2
No	1,102	27.9	3,069	72.1	0.9

^*SDA) values defined as occluding anterior teeth and unrestored edentulous areas posterior to the most distal ipsilateral canine, premolar or first molar in a unilateral situation or unrestored edentulous areas posterior to the most distal canines, premolars or first molars (or combination) in a bilateral situation.

The unadjusted odds ratio for the logistic regression of the maxillary SDA on BMI was 0.94 (95% CI: 0.69–1.28). An adjusted model was developed in which the following variables were added: sex (male vs. female); race/ethnicity (non-Hispanic black, Mexican American or Other vs. non-Hispanic white); education (high school or less than high school vs. more than high school); age (55–65 years vs. 45–54 years); family income to poverty ratio (<2 vs. ≥2); smoking (current or former smoker vs. never smoker); alcohol consumption (moderate or heavy vs. no consumption); and moderate exercise (yes vs. no). This model had an adjusted odds ratio of 1.06 (95% CI: 0.63–1.78). A second model was created which also included the status of the opposing arch (SDA, edentulous or other vs. complete dental arch). The second model had an adjusted odds ratio of 1.14 (95% CI: 0.65–2.02).

The unadjusted odds ratio for the logistic regression of the mandibular SDA on BMI was 0.8994 (95% CI: 0.70–1.12). An adjusted model was developed in which the following variables were added: sex (male vs. female); race/ethnicity (non-Hispanic black, Mexican American or Other vs. non-Hispanic white); education (high school or less than high school vs. more than high school); age (55–65 years vs. 45–54 years); family income to poverty ratio (<2 vs. ≥2); smoking (current or former smoker vs. never smoker); alcohol consumption (moderate or heavy vs. no consumption); and moderate exercise (yes vs. no). This model had an adjusted odds ratio of 0.70 (95% CI: 0.46–1.08). A second model was created that also included the status of the opposing arch (SDA, edentulous or other vs. complete dental arch). The second model had an adjusted odds ratio of 0.65 (95% CI: 0.37–1.11). Logistic regressions are presented in tabular form in Table 4.

Table 4.

Logistic regression of shortened dental arch on body mass index status: National Health and Nutrition Examination Survey 2005–2008

Variable	Unadjusted odds ratio (95% CI)	Adjusted odds ratio model 1 (95% CI)	Adjusted odds ratio model 2 (95% CI)
Mandibular SDA	0.89 (0.70–1.12)	0.70 (0.46–1.08)	0.65 (0.37–1.11)
Complete dental arch	Reference	Reference	Reference
Maxillary SDA	0.94 (0.69–1.28)	1.06 (0.63–1.78)	1.14 (0.65–2.02)
Complete dental arch	Reference	Reference	Reference

The adjusted model 1 additionally includes: sex (male vs. female); race/ethnicity (non-Hispanic black, Mexican American or Other vs. non-Hispanic white); education (high school or less than high school, vs. more than high school); age (55–65 years vs. 45–54 years); family income to poverty ratio (<2 vs. ≥2); smoking (current or former smoker vs. never smoker); alcohol consumption (moderate or heavy vs. no consumption); and moderate exercise (yes vs. no). The adjusted model 2 also includes the status of the opposing arch [shortened dental arch (SDA), edentulous or other vs. complete dental arch]. SDA values were defined as occluding anterior teeth, and unrestored edentulous areas posterior to the most distal ipsilateral canine, premolar or first molar in a unilateral situation or unrestored edentulous areas posterior to the most distal canines, premolars or first molars (or combination) in a bilateral situation.

95% CI, 95% confidence interval.

DISCUSSION

The results of this study did not reject the null hypothesis that there was no difference in the odds ratio for SDA on overweight or obese BMI status compared with the odds ratio for complete dentition on overweight or obese BMI status. To the authors’ knowledge, this study is the first in which BMI and SDA were evaluated for an association in working-age adults, 45–65 years of age.

Bernardo et al.¹² conducted a study of 1,720 Brazilian working-age adults, 20–59 years of age, which involved tooth loss (rather than SDA) as the key independent variable and BMI as the outcome; the association was not statistically significant with age in the adjusted model. The results of Bernardo et al. are similar to the results of the present study.

However, in a study of 2,816 Swedish men and women, 30–74 years of age, there was an association between tooth loss and BMI¹³. The Swedish study differed from the current study in that tooth loss, rather than SDA, was the key independent variable.

Other researchers studied BMI as the independent variable with tooth loss as the dependent variable. In one such study, conducted among working-age Brazilian adults, there were no statistically significant associations in adjusted analyses of tooth loss and overweight/obesity¹⁴. In an older adult population in Brazil, there was an association between more teeth present and a lesser likelihood of obesity.

The present study is relevant because identifying evidence-based justification for treatment is important. Other reasons for treating the SDA include aesthetics, oral health-related quality of life⁵, chewing difficulty¹⁵ and nutritional status¹⁶. However, in evaluating a potential association between BMI and SDA, the researchers did not detect a difference between individuals with an SDA and individuals with complete dentition in terms of BMI as a determinate for treating SDA.

The study strengths include the use of several years of data from a large, nationally representative survey, with examination data as well as questionnaire data. The study weaknesses include the limitations imposed by cross-sectional data collection, in which temporality and causality cannot be determined. Although multiple years were included in the analyses, the subgroup sample sizes of various types of SDA were too small to determine if specific SDA patterns were related to BMI, and it was not possible to increase the sample size by including more years of NHANES data because recent NHANES data does not include information on the presence of maxillary/mandibular dental prostheses. Although the NHANES data for 2005–2008 contained information about prostheses, it did not contain information about periodontal status, which could possibly influence the type of foods selected and ultimately the BMI.

We failed to reject the null hypothesis that the odds ratio for SDA on overweight or obese BMI status would be the same as the odds ratio for complete dentition. The research hypothesis that SDA would be related to higher BMI, and the corollary that restored or complete dentition had better odds of a lower BMI, were not supported by this study.

Acknowledgements

R.C.W. received research support from the National Institute of General Medical Sciences of the National Institutes of Health under Award Number U54GM104942. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Department of Justice (FBOP).

Conflict of interest

The authors report no conflict of interest.