Oral Health and Risk of Malnutrition in Older Adults: A Systematic Review and Meta-analysis

ABSTRACT

Objective:

Older adults are generally at risk of malnutrition due to a decline in physical function, with or without systemic diseases. The objective of this systematic review was to investigate the association between dentition/occlusal support measures and nutritional conditions in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

Methods:

A comprehensive search of PubMed, Web of Science, EMBASE, Google Scholar, and ICHUSHI databases was conducted from July to September 2024. Eight independent researchers performed data extraction and quality assessment using the modified Newcastle–Ottawa Scale. Nineteen studies (12 cross-sectional and 7 cohort studies) were included.

Results:

The review revealed that individuals with 20 or more teeth consistently had significantly higher intake of various nutrients compared to those with fewer teeth. Similarly, compromised occlusal support was associated with reduced intake of vegetables, dietary fiber, and vitamins. Several studies demonstrated a significant association between edentulism or absence of occlusal support and an increased risk of malnutrition. Notably, a meta-analysis of two studies indicated that individuals with fewer than 20 teeth had a 52% increased risk of unintentional weight loss (odds ratio [OR]: 1.52; 95% confidence interval [CI]: 1.08–2.14). Another meta-analysis showed edentulous individuals had a 43% higher likelihood of malnutrition risk compared to dentate individuals (OR: 1.43; 95% CI: 1.13–1.80).

Conclusion:

These findings highlight that fewer remaining teeth, edentulism, and absence of occlusal support are significantly associated with adverse nutritional outcomes. This highlights the critical importance of maintaining functional dentition and integrating oral health assessments into routine geriatric care.

PROSPERO, registration number:

CRD42024531807.

INTRODUCTION

Globally, the population aged 65 years and older is increasing rapidly, with one in six individuals projected to be in this age group by 2030, creating a substantial public health challenge.[1] As life expectancy continues to increase, supporting healthy aging and preserving functional capacity have become urgent priorities for healthcare providers and policymakers.[2] Aging is often accompanied by a progressive decline in physical function, together with physiological, psychological, and socioeconomic changes that are further compounded by chronic diseases and medication use.[3] These factors can compromise dietary intake, restrict food choice, and increase the risk of malnutrition in older adults.[3,4] Adequate nutrition is therefore essential in later life, as it contributes to the maintenance of independence, the prevention of frailty, and a reduction in disease burden.[5] Malnutrition remains a prevalent concern among older adults and is significantly associated with sarcopenia, impaired immune function, poor quality of life, higher healthcare costs, and increased mortality.[5,6]

Oral health is a key determinant of nutritional status. Tooth loss, which remains prevalent in older populations, significantly reduces masticatory efficiency and impairs the ability to chew fibrous foods such as fresh fruits, vegetables, and meats.[7] In response, many older adults substitute these foods with softer, processed alternatives that are easier to chew but often lower in nutrients and higher in calories.[8,9] Over time, this dietary adaptation can contribute to systemic health problems, including cardiovascular disease, osteoporosis, and cognitive decline.[4,5,7]

The bidirectional relationship between oral health and nutritional status further complicates this issue.[2] Poor nutritional status can negatively impact oral health by weakening the immune system, thereby increasing susceptibility to oral diseases such as periodontitis.[10] This creates a detrimental cycle: oral dysfunction impairs nutrition, which in turn worsens oral health, leading to a further decline in oral function.[2,7,10] Despite increasing recognition of the importance of oral health for nutrition and overall well-being, evidence from epidemiological studies remains inconsistent due to methodological differences, and significant knowledge gaps remain. Specifically, the role of factors such as the number of remaining teeth, functional occlusion, and prosthetic use in influencing nutritional outcomes is not yet fully understood.

Given the growing older adult population and the critical importance of maintaining functional capacity and quality of life, there is an urgent need to synthesize the existing evidence regarding the relationship between dentition status and nutritional outcomes. Understanding these relationships is essential for developing evidence-based prevention and intervention strategies that can help older adults maintain both oral health and nutritional well-being. Therefore, this systematic review aimed to investigate the impact of oral conditions on nutritional status by examining how the number of present teeth or functional tooth units (FTUs) relates to nutritional outcomes. We systematically searched existing evidence and evaluated the quality of available studies to provide stronger evidence for clinical practice and public health policy. The findings are expected to highlight the importance of maintaining oral function and integrating oral health assessments into routine geriatric care, helping to prevent malnutrition and support healthy aging.

MATERIALS AND METHODS

STUDY DESIGN AND REGISTRATION

This systematic review focused on a research question: “Does tooth loss affect nutritional status among older adults?” This review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.[11] The protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO, registration number: CRD42024531807).

Database searches and data extraction were independently conducted by eight researchers (FC, HO, KMT, NT, AR, ST, TZT, and OV) with dual verification. In case of disagreement, consensus was achieved through discussion among the research team.

RESEARCH QUESTION AND ELIGIBILITY CRITERIA

The research question was defined using the PECO framework: Population (P; adults), Exposure (E; tooth loss), Comparison (C; edentulous individuals or with prosthetic replacement), and Outcome (O; nutritional status or malnutrition).

The inclusion criteria were as follows:

• Studies involving adult populations.

• Research investigating the association between dentition status (self-reported or clinically evaluated) and nutritional conditions.

• Observational studies (cross-sectional, cohort, case–control) and interventional studies.

The exclusion criteria were as follows:

• Conference abstracts, letters to editors, review articles, laboratory studies, and qualitative studies.

• Studies conducted exclusively among institutionalized populations (hospitalized patients or nursing home residents).

• Articles without access to full text.

SEARCH STRATEGY AND STUDY SELECTION

We searched for studies in adults without applying an upper age limit. To maximize sensitivity, we did not use age-specific terms in the search strategy. The definition of population was applied during screening based on the eligibility criteria. A comprehensive literature search was performed across multiple databases including PubMed, Web of Science, EMBASE, Google Scholar, and ICHUSHI (Japan Medical Abstracts Society Database) from July to September 2024. The detailed search query strategies for each database were provided in Supplementary Material 1. No language restrictions were applied during the search process.

All identified references were imported into an Excel file. After removing duplicates, the remaining articles were screened for eligibility using preestablished inclusion and exclusion criteria. Study selection was performed through a two-stage screening process: initial screening based on titles and abstracts, followed by full-text evaluation of potentially eligible studies. Data searches were undertaken in pairs by eight coinvestigators (FC, HO, KMT, NT, AR, ST, TZT, and OV). For each reviewer pair, the screening completed by the primary reviewer was subsequently cross-checked by the paired reviewer to ensure appropriateness in accordance with PRISMA guidelines. Any discrepancies between reviewers were resolved through discussion until consensus was achieved.

DATA EXTRACTION AND DATA ENTRY

Data extraction was performed independently by eight investigators (FC, HO, KMT, NT, AR, ST, TZT, and OV) using a standardized data extraction form in a Microsoft Excel spreadsheet. All extracted data underwent dual verification, and discrepancies were resolved through team discussion to achieve consensus. The following variables were systematically extracted from each included study: first author, year of publication, country where studies were conducted, demographic characteristics (sample size, number of males, and age range/mean), exposure measurement methods, exposure definition, outcome indicators, assessment methods, comparison groups, language of publication, and key findings.

ASSESSMENT OF RISK OF BIAS

Quality assessment was conducted independently by two researchers (ST and KMT) with disagreements resolved through discussion (Supplementary Material 2). We applied the Newcastle–Ottawa Scale (NOS) to perform a standardized risk-of-bias assessment.[12] This instrument evaluates three domains: (1) participant selection, (2) comparability of groups, and (3) the ascertainment of either the exposure (for cross-sectional studies) or the outcome (for cohort studies) of interest. Overall, each item in the NOS is given a star if it meets criteria for low risk of bias. Since the NOS was originally designed for cohort and case–control studies, a modified version was utilized to accommodate cross-sectional designs.[12,13]

Regarding evaluation of exposure assessment, studies utilizing medical/dental records or clinical oral examination were considered to have valid exposure measurement. Studies that did not specify the method for ascertaining tooth loss were classified as having “unclear” risk of bias, while those relying solely on self-reported number of teeth or tooth loss were classified as high risk of bias. Total scores were calculated as percentages to account for different total scores (≤50%: low quality, >50%–75%: moderate quality, and >75%: high quality).

META-ANALYSIS

Where sufficient data were available, meta-analyses were conducted using a random-effects model to account for between-study heterogeneity. Studies were excluded from quantitative synthesis if they did not report sufficient data for effect size calculation or lacked adequate information to compute odds ratios (ORs).

Due to variations in tooth loss classification methods across studies (e.g., edentate vs. partial or no tooth loss, <20 present teeth vs. ≥20 teeth), comparison was standardized by contrasting the group with the fewest teeth against the group with the most teeth. Pooled ORs were calculated from eligible studies using random-effects meta-analysis. Effect sizes were reported as pooled ORs with 95% confidence intervals (CIs) for categorical outcomes. Between-study heterogeneity was quantified using I² statistic. The risk of publication bias was tested by visually inspecting whether the funnel plot was symmetrical. Egger’s tests were not performed because the number of included studies was <10 and test power was too low to distinguish chance from real asymmetry.[14]

All statistical analyses were performed with EZR (Jichi Medical University, Tochigi, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). More precisely, it is a modified version of R Commander designed to add statistical functions frequently used in biostatistics.[15] A P value of <0.05 was considered statistically significant.

CERTAINTY OF EVIDENCE (GRADE ANALYSIS)

We assessed the certainty of evidence for each critical outcome using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach.[16] The assessment was conducted manually, without dedicated software. For each outcome, we considered the body of evidence contributing to that outcome (studies included in the meta-analysis where pooling was performed). Because the included studies were predominantly observational, we started at low certainty and then downgraded the certainty as appropriate for the following domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Risk-of-bias judgments were informed by our NOS assessment. The final certainty of evidence (high, moderate, low, or very low) and the reasons for any downgrading were summarized for each outcome in a GRADE evidence table. Finally, a summary-of-findings table was produced.

RESULTS

The electronic database search identified 6207 articles after the initial search, of which 814 were duplicates. In the initial screening stage, 5152 articles were excluded as either editorials, letters, case reports, reviews, abstracts, author debates, interviews, or conference proceedings. After secondary screening, 243 articles were reviewed, and an additional 234 articles were excluded as irrelevant for this review (Supplementary Material 3). Handsearching was also conducted to identify additional relevant studies not captured by electronic searches. Nineteen studies were finally selected for inclusion in the systematic review. The study selection process is shown in Figure 1.

Figure 1.

Flow diagram of the literature search

STUDY CHARACTERISTICS

The characteristics of the 19 included studies are presented in Table 1. Twelve studies employed cross-sectional designs, while the remaining seven studies used cohort designs. The studies were conducted in 10 countries. All included studies were conducted in older adult populations; no eligible studies involving younger adults were identified.

Table 1.

Characteristics of studies included in this study

Author (year)	Study design, country	Population	Total Male, n (%)	Exposure	Measurement tool/type of outcome	Results
Outcome: Nutrition (11 studies)
Yoshihara et al. (2005)[17]	Cross-sectional Japan	Community-dwelling adults aged 74 years	57 31 (54.3%)	Number of teeth present Category: Number of teeth 0–19/20+	Food record method/protein intake, mineral intake, vitamin intake, dietary fiber	Total protein, animal protein, sodium, vitamin D, vitamin B, vegetables, and fish were larger among people with 20+ teeth.
Liedberg et al. (2007)[20]	Cross-sectional Sweden	Community-dwelling adults aged 67–68 years	481 481 (100%)	Occlusal tooth contact (contact between mandibular and maxillary teeth)	Dietary history interview by dietitians/nutrition levels were categorized as: “Adequate/inadequate.”	No significant difference in the number of occlusal contacts between adequate and inadequate nutritional intake.
Gil-Montoya et al. (2008)[23]	Cross-sectional Spain	Community-dwelling older adults aged ≥65	2860 1192 (41.7%)	Edentate	MNA score/score < 17: malnutrition	Edentate people had a significant lower mean MNA (P < .001), yet this association was not observed after controlling confounding factors.
Yoshida et al. (2011)[19]	Cross-sectional Japan	Community-dwelling older adults aged 65–85	182 60 (32.9%)	Occlusal tooth contact (evaluated by Eichner classification)	BDHQ/intake of macronutrients	People with no occlusal contact with natural dentition in the molar region had significantly lower intake of vegetables and dietary fiber compared with those with retained contact. There was a significant difference in the intake of various vitamins between the two groups. No difference in BMI or intake of macronutrients.
Kikutani et al. (2013)[21]	Cross-sectional Japan	Community-dwelling older adults aged ≥65	716 240 (33.5%)	Functional occlusion	MNA-SF/cutoff value for malnutrition was not mentioned	Significant correlation was observed between occlusal status and risk of malnutrition (P < 0.05). Those with functionally inadequate occlusion had 3.189 greater malnutrition risk (P < 0.05) than those with natural dentition and adequate function.
Lopez-Jornet et al. (2013)[26]	Cross-sectional Spain	Institutionalized and noninstitutionalized adults aged ≥65	465 213 (45.8%)	Edentate	MNA score/score < 23.5: a risk of malnutrition or malnutrition	Edentulousness did not show statistical association with malnutrition risk or malnutrition. Similarly, having a denture did not show any association with malnutrition risk nor malnutrition.
Iwasaki et al. (2014)[24]	Cross-sectional Japan	Community-dwelling adults aged ≥65	195 63 (32.2%)	Number of teeth present Number of occluding tooth pairs	BDHQ/dietary intake of 13 nutrients and 7 food groups (the number of nutrients below the recommended intake levels was counted)	Occluding tooth pairs were associated with malnutrition in the crude model, yet this association was nonsignificant after controlling for confounders.
Iwasaki et al. (2016)[44]	Cohort Japan (5 years)	Community-dwelling adults aged 75 years	286 144 (50.4%)	FTU Category: FTU ≤5 or >5	BDHQ/changes in dietary intake	Baseline impaired dentition (FTU ≤ 5) was significantly associated with a greater decline in intake of multiple nutrients (protein, sodium, potassium, calcium, vitamin E, and dietary fiber) and food groups (vegetables and meat) than those without impaired dentition.
Krzyminska-Siemaszko et al. (2016)[22]	Cross-sectional Poland	Community-dwelling older adults aged ≥65	3751 1981 (52.8%)	Edentate	MNA-SF score/score < 12: malnutrition	Edentulism was associated with higher odds of malnutrition (OR = 1.26, P = 0.009)..
Wu et al. (2018)[25]	Cross-sectional China	Community-dwelling older adults aged ≥65	195 63 (32.2%)	Number of occluding tooth pairs	MNA score/score < 23.5: a risk of malnutrition or malnutrition	Occluding tooth pairs were associated with malnutrition in the crude model, yet this association was nonsignificant after controlling for confounders.
Logan et al. (2020)[18]	Cohort (13 years) Australia	Community-dwelling men aged 60–69 years	1096 1096 (100%)	Number of teeth present with dentures: Category: 21–28 teeth without denture/21–28 teeth denture/1–20 teeth with denture/1–20 teeth without denture/edentate	FFQ/dietary intake	Having 21 or more natural teeth was positively associated with intakes of fruits, vegetables, and nuts, and higher diet quality scores compared to those with 1–20 teeth or no natural teeth.
Outcome: BMI or body weight (8 studies)
Sheiham et al. (2002)[28]	Cross-sectional UK	Community-dwelling and institutionalized adults aged ≥65	629 280 (44.5%)	Number of POPs	BMI: <20: underweight 20–25: normal >25: overweight	The proportion of underweight was significantly higher in the edentate compared to those with 11 or more teeth (OR: 3.1, P < 0.05). The proportion of underweight was also higher in those with no POP compared to those with POP ≥1 (P = 0.021).
de Andrade et al. (2014)[30]	Cohort (4 years) Brazil	Community-dwelling adults aged ≥60	2143 791 (36.9%)	Edentulism Category: Number of teeth 0/1+	Changes in WC, change in body weight (5% or more)	Edentulism is an important predictor of weight and WC loss among older adults. Risk of experiencing weight loss was greater among edentulous people (relative risk ratio: 2.11, 95% CI: 1.35–3.30).
Gaewkhiew et al. (2019)[27]	Cross-sectional Thailand	Community-dwelling older adults aged ≥60	788 238 (30.2%)	FTU > 0 FTUs defined as pairs of opposing natural teeth and/or artificial teeth on fixed prostheses. The presence of opposing premolars is defined as 1 FTU and molars as 2 FTUs.	BMI/BMI < 18.5: underweight BMI > 25: overweight/obese	Those with FD (FTU > 0) were less likely to be underweight (PR: 0.39, 95% CI: 0.16–0.95) than those with neither FD nor dentures.
Kiesswetter et al. (2019)[33]	Cohort (9 years) Netherlands	Community-dwelling adults aged 55–85 years	893 419 (46.9%)	Number of teeth present, pain while chewing, xerostomia Category: number of teeth 0/1–7/8+	Involuntary weight loss/≥5% in the previous 6 months and/or low BMI (<20: 70 years or younger, ≥22: 70 years and older)	Toothache while chewing was identified as a determinant of incident malnutrition.
Gaewkhiew et al. (2020)[34]	Cohort (1 year) Thailand	Community-dwelling adults aged ≥60	651 178 (27.3%)	FTU: Category: No FTU or FTU > 0	Changes in BMI, WC, or triceps skinfold thickness	FD was defined as numbers of premolars and molars with opposing teeth. Having FD was not associated with changes in BMI or WC or triceps skinfold thickness.
Ritchie et al. (2000)[29]	Cohort (1 year) UK	Community-dwelling adults aged ≥70	563 237 (42%)	Edentate Category: Edentate or dentate	Weight loss (10% or more)	Edentulousness was an independent risk factor for weight loss (OR: 1.64 for 4% weight loss, OR: 2.03 for 10% weight loss).
Takehara et al. (2021)[31]	Cross-sectional Australia	Community-dwelling adults aged ≥70	542 542 (100%)	Number of teeth present: ≥20 or <20	Weight loss (5% or more)	The number of teeth present was significantly associated with weight loss (PR: 1.77, 95% CI: 1.05–2.97).
Shiota et al. (2023)[32]	Cohort (3 year) Japan	Community-dwelling adults aged ≥65	63,602 30,427 (47.8%)	Number of teeth present, pain while chewing, xerostomia Category: number of teeth ≥20/10–19/0–9	Weight gain/loss (5% or more)	Fewer remaining teeth, chewing difficulty, and xerostomia were associated with an increased risk of weight loss.

BDHQ = Brief-type self-administered Diet History Questionnaire, BMI = body mass index, CI = confidence interval, FD = functional dentition, FFQ = Food Frequency Questionnaire, FTU = functional tooth unit, MNA = Mini-Nutritional Assessment, MNA-SF: Mini-Nutritional Assessment Short Form, OR = odds ratio, POP = posterior occluding pair, WC = waist circumference, PR = prevalence ratio

ASSESSMENT OF OUTCOME

The included studies employed two distinctive approaches for nutritional assessment. Eleven studies evaluated nutritional status through dietary intake assessment, while eight studies used anthropometric measures such as body mass index (BMI) or body weight changes.

Among the 11 studies that evaluated nutritional status using dietary intake assessment, 5 studies used the Mini-Nutritional Assessment (MNA) tool. Three studies utilized a validated instrument, specifically the Brief-type self-administered Diet History Questionnaire. The remaining three assessed dietary intake using nonvalidated approaches.

On the other hand, eight studies utilized anthropometric measures such as BMI, weight change, or waist circumference to evaluate nutritional status.

ASSESSMENT OF TOOTH LOSS AS EXPOSURE

Across the included studies, dentition status was assessed using heterogeneous indicators, including the number of remaining teeth, edentulous status, occluding tooth pairs, and posterior occlusal support measures, such as FTUs.

RISK OF BIAS

The risk of bias of the included studies was evaluated using the NOS.[12] The results are presented in Table 2. Ten studies were classified as “high quality,” six as “moderate quality,” and the remaining three as “low quality.” Among the 12 cross-sectional studies, all studies had concerns about nonrespondents, and none reported comparability between respondents and nonrespondents’ characteristics. Only 4 of the 12 cross-sectional studies explicitly reported the use of predetermined sample size.

Table 2.

Newcastle–Ottawa Scale score of the included studies

Cohort studies (maximum 9 stars)
Author (year)	Selection				Comparability		Outcome			Score	Evaluationa
	1	2	3	4	1	2	1	2	3
Ritchie et al. (2000)[29]	*	*	*			*	*		77%	5	Moderate
de Andrade et al. (2014)[30]	*	*	*		*	*	*	*	37%	7	High
Iwasaki et al. (2016)[44]	*	*	*		*	*		*	77%	6	Moderate
Kiesswetter et al. (2019)[33]	*	*		*	*	*	*	*	86% *	8	High
Gaewkhiew et al. (2020)[34]	*	*	*		*	*	*		82.6% *	7	High
Logan et al. (2020)[18]	*	*	*		*	*		*		6	Moderate
Shiota et al. (2023)[32]	*	*		*	*	*		*	56.9%	6	Moderate
Cross-sectional studies (maximum 10 stars)
Author (year)	Selection				Comparability	Outcome			Score	Evaluation a
	1	2	3	4		1	2
Sheiham et al. (2002)[28]	*			**	**	**	*		8	High
Yoshihara et al. (2005)[17]	*			*		**			4	Low
Liedberg et al. (2007)[20]	*			**	*	*			5	Low
Gil-Montoya et al. (2008)[23]	*			**	**	**	*		8	High
Yoshida et al. (2011)[19]				**	*	**			5	Low
Kikutani et al. (2013)[21]	*			**	*	**			6	Moderate
Lopez-Jornet et al. (2013)[26]	*	*			**	**	*		7	Moderate
Iwasaki et al. (2014)[24]	*			**	**	**	*		8	High
Krzyminska-Siemaszko et al. (2016)[22]	*	*		**	**	**	*		9	High
Wu et al. (2018)[25]	*	*		**	**	**	*		9	High
Gaewkhiew et al. (2019)[27]	*	*		**	**	**	*		9	High
Takehara et al. (2021)[31]	*			**	**	**	*		8	High

^aTotal scores were calculated as percentages to account for total scores (≤50%: low quality, >50%–75%: moderate quality, >75%: high quality). Asterisks indicate that the study met the corresponding predefined NOS criterion. The total score represents the number of stars awarded across the Selection, Comparability, and Outcome domains.

ASSOCIATION BETWEEN ORAL HEALTH AND NUTRITIONAL INTAKE

Studies consistently demonstrated that older individuals with 20 or more teeth had significantly higher intake of protein, vitamins B and D, vegetables, and fish, indicating overall better diet quality.[17,18] Indicators of posterior occlusal support, such as molar contact and functional occlusion, were also associated with greater consumption of vegetables, dietary fiber, and macronutrients.[19] However, one study reported that there was no significant association between nutritional levels and occlusal contact.[20]

Two studies demonstrated a significant association between poor oral health and increased malnutrition risk.[21,22] However, several other studies did not observe a significant association after adjustment for confounding factors, indicating some heterogeneity among the findings.[23,24,25,26]

ASSOCIATION BETWEEN ORAL HEALTH AND BMI/BODY WEIGHT

Eight studies examined the association between oral health and BMI or body weight changes. Cross-sectional studies generally showed a higher prevalence of underweight among individuals with poor dentition, including edentulism or lack of occlusal pairs.[27,28]

Findings from cohort studies were consistent: several longitudinal studies demonstrated that fewer remaining teeth or edentulous status was associated with subsequent weight loss among community-dwelling older adults.[29,30,31] Larger population-based cohort similarly reported that impaired oral function was associated with an increased risk of unintentional weight loss.[32] Two studies, however, reported no significant associations between oral health and changes in BMI.[33,34]

META-ANALYSIS

Weight loss risk

A meta-analysis examining the association between number of teeth and weight loss risk included two studies comparing individuals with “fewer than 20 teeth” to those with “20 or more teeth.”[31,32] The pooled analysis demonstrated that individuals with fewer than 20 teeth had a significantly increased likelihood of being at risk of unintentional weight loss compared to those with 20 or more teeth [Figure 2]. The fixed-effect model yielded an OR of 1.36 (95% CI: 1.30–1.42), while the random-effects model, which accounts for between-study heterogeneity, produced the reported estimate of 1.52 (95% CI: 1.08–2.14). Moderate heterogeneity was observed between studies (I² = 60%, P = 0.11), though this did not reach statistical significance. Both included studies employed a 3-year follow-up period and defined weight loss as a 5% or more reduction in body weight from baseline.

Figure 2.

Forest plot showing association between 20+ teeth present/fewer than 20 teeth and unintentional weight loss

Malnutrition risk

Another separate meta-analysis evaluated the relationship between edentulism and malnutrition risk, incorporating two studies that compared edentate versus dentate older adults. The pooled analysis using random-effects model revealed that edentate individuals had a 43% higher likelihood of being at risk of malnutrition or being malnourished compared to dentate individuals (OR: 1.43; 95% CI: 1.13–1.80; I² = 79%) [Figure 3].[22,23] Substantial heterogeneity was present between studies (I² = 79%, P = 0.03). The observed heterogeneity might be attributed to methodological differences in nutritional assessment between the included studies.

Figure 3.

Forest plot showing association between edentate/dentate and risk of malnutrition

Publication bias

To identify the publication bias, funnel plots were generated. However, because only two studies were included in each meta-analysis [Supplementary Figures S1 and S2], visual assessment of funnel plot asymmetry is not informative and is misleading. Therefore, publication bias could not be reliably assessed.

Certainty of evidence (GRADE)

The certainty of evidence for each outcome is summarized in Table 3. For unintentional weight loss, the pooled estimate from two observational studies indicated that having <20 teeth was associated with higher odds of weight loss (OR: 1.52, 95% CI: 1.08–2.14), corresponding to 55 more cases per 1000 (95% CI: 9 more to 112 more), with low certainty of evidence. For risk of malnutrition (MNA score), two observational studies showed higher odds of malnutrition among edentulous participants (OR: 1.43, 95% CI: 1.13–1.80), corresponding to 89 more cases per 1000 (95% CI: 30 more to 146 more); however, the certainty of evidence was very low, primarily due to substantial heterogeneity. Publication bias was not assessed because of the small number of included studies, and we did not downgrade the evidence for this domain.

Table 3.

Certainty of evidence (GRADE)

Outcomes	Number of participants (studies)	Relative effect (95% CI)	Anticipated absolute effects (95% CI)	Certainty of the evidence (GRADE)	Footnote
Unintentional weight loss (5% weight loss for 3 years) Assessed with OR	25,693 cases 38,451 controls (2 observational studies)	OR: 1.52 (1.08–2.14)	55 more per 1,000 (9 more to 112 more)	Low	a
Risk of malnutrition (MNA score) Assessed with OR	2,594 cases 3,894 controls (2 observational studies)	OR: 1.43 (1.13–1.80)	89 more per 1,000 (30 more to 146 more)	Very low	a , b

CI = confidence interval, GRADE = Grading of Recommendations Assessment, Development and Evaluation, MNA = Mini-Nutritional Assessment, OR = odds ratio

^aPublication bias was not assessed due to the small number of studies, and we did not downgrade for publication bias

^bDowngraded for inconsistency due to substantial unexplained heterogeneity across studies

DISCUSSION

This systematic review provides comprehensive evidence regarding the impact of tooth loss on nutritional status among community-dwelling older adults. Our findings demonstrated that poor oral health, characterized by fewer remaining teeth, edentulism or compromised occlusal support, is significantly associated with adverse nutritional outcomes. The meta-analysis revealed that individuals with fewer than 20 teeth had a 52% increased risk of unintentional weight loss compared to those with 20 or more teeth. Additionally, edentulous individuals showed a 43% higher likelihood of malnutrition risk compared to dentate individuals.

The observed associations between oral health and nutritional outcomes can be explained through several interconnected mechanisms. First, mechanical factors play a crucial role, as tooth loss directly impairs masticatory function, leading to difficulties in chewing fibrous foods such as fruits, vegetables, and proteins.[30,35] This mechanical limitation often results in dietary modifications, with individuals shifting toward softer, processed foods that may be energy-dense but nutrient-poor. However, the problem of nutritional status among elderly individuals is not only a question of preferences or dietary habits.[36] Other factors such as reduced physical function, psychological aspects, and the living environment can also affect nutritional intake.[37]

An association between nutritional condition and oral health has been reported by many studies. As people age, oral diseases become more prevalent, often resulting in a reduced number of teeth present.[38] Such impairments compromise chewing and swallowing capabilities, limit dietary variety, and ultimately reduce overall nutritional intake, thereby increasing the risk of malnutrition. Conversely, malnutrition negatively impacts skeletal muscle mass and strength, including the orofacial and swallowing muscles, which can further aggravate declines in oral function.[39] Malnutrition may also affect oral function through another mechanism.[40] Poor nutritional status weakens the immune system, increasing susceptibility to oral diseases. Furthermore, certain nutrients like carotenoids and vitamins C and E possess anti-inflammatory and antioxidant properties.[41] When oral dysfunction limits the intake of these protective nutrients, it may elevate the risk of developing periodontitis.[42,43] These interconnected relationships create a detrimental cycle: compromised oral function contributes to malnutrition, which increases vulnerability to oral diseases, leading to further deterioration of oral function. Therefore, preserving good oral function in older adults is crucial for preventing individuals from becoming trapped in this harmful cycle.

The consistent findings from two included studies showed that individuals with 20 or more teeth were less likely to experience weight loss.[31,32] Similarly, individuals with 20 or more teeth demonstrated superior nutritional intake patterns in two studies.[17,18] These results support the concept of a functional dentition threshold, which appears to represent a critical point affecting the ability to maintain adequate nutritional status or body weight. Beyond simply counting teeth, four studies further reported the importance of occlusal support on nutritional condition.[19,27,28,44] These studies examined posterior occlusal contact and evaluated whether participants had some occlusal support.[19,27,28] The importance of posterior occlusal contact, particularly in molar regions, was emphasized across these studies, highlighting the role of functional units rather than merely the number of teeth present. Collectively, these findings suggest that maintaining functional occlusion—whether through natural teeth or removable dentures—is crucial for preserving nutritional status in the older population.

This emphasis on functional dentition and its link to nutritional decline contrasts with the earlier systematic review. Gaewkhiew et al. (2017)[45] reported an inconsistent association between tooth loss and weight change, identifying risks for both weight gain and loss. In contrast, our systematic review and meta-analysis of recent studies suggest a more consistent and specific trend toward unintentional weight loss. Gaewkhiew et al.[45] suggested that tooth loss might lead to weight gain in some cases, potentially due to dietary shifts following extraction of loose or diseased teeth. However, the studies identified in our review, particularly recent large-scale cohorts, predominantly highlighted the association with weight loss.[29,30,31,33] This discrepancy may be attributed to the characteristics of the study populations in recent large-scale cohort studies included in our review, wherein the impact of tooth loss emerges more prominently as unintentional weight loss due to frailty and sarcopenia rather than obesity. While tooth loss can theoretically lead to weight gain due to soft-diet choices, our findings suggest that in the later stages of life, the pathway to undernutrition and weight loss is dominant.

Our findings align with a systematic review by Zelig et al.,[46] demonstrating a clear link between tooth loss and compromised nutritional status. However, our review extends beyond their analytical approach. While Zelig et al.[46] analyzed nutritional status solely in terms of nutritional intake using established measurement tools such as the MNA, we recognized that nutritional status assessment is not limited to such standardized tools. It can also be performed using BMI or unintentional weight loss as standardized indicators. Given that some cohort studies were not initially designed for dental or nutritional assessment purposes, we broadened our inclusion criteria for nutritional evaluation to encompass not only standardized tools such as the MNA but also individual indicators such as BMI and weight loss.

STRENGTH AND LIMITATIONS

Several limitations should be considered when interpreting our findings. First, the substantial heterogeneity between studies limits the precision of pooled estimates and suggests that the true effect may vary across different populations and settings. Second, the predominance of cross-sectional studies limited our ability to establish definitive causal relationships. Third, the limited number of studies eligible for meta-analysis restricts the power to detect small effect sizes and limits subgroup analyses. In addition, the diverse methodological approaches across studies made it difficult to make direct comparisons or to conduct pooled analyses. While we made efforts to minimize publication bias by searching several databases, citation tracking, and handsearching, we cannot rule out publication bias, given the limited number of included studies and the possibility of nonindexed work. Fourth, most included studies were conducted in developed countries with established healthcare systems, potentially limiting the generalizability of findings to resource-limited settings. Lastly, all eligible studies were conducted in older adults, so the findings should not be generalized to younger adult populations. This also indicates an evidence gap regarding the association between dentition/occlusal support and malnutrition-related outcomes in younger adults.

Despite these limitations, our review has several notable strengths. The comprehensive search strategy across multiple databases and languages minimizes selection bias. The rigorous study selection process and dual verification of data extraction enhance the reliability of our findings. The pooled estimates indicated that having <20 teeth was associated with unintentional weight loss, and that edentulism was associated with risk of malnutrition. However, the overall certainty of evidence was low to very low, because the included studies were exclusively observational—and therefore started at low certainty within the GRADE framework—and were further downgraded due to heterogeneity. Therefore, these findings should be interpreted cautiously and warrant confirmation in well-designed prospective studies.

CLINICAL IMPLICATIONS

Our findings have substantial implications for clinical practice and public health policy. The identification of oral health as a significant predictor of nutritional risk suggests that oral health assessments should be integrated into routine geriatric care and nutritional screening protocols. Healthcare providers should consider oral health status when evaluating older adults for malnutrition risk, particularly in community-dwelling populations where early intervention may prevent progression to severe malnutrition. In particular, a functional dentition threshold of ≥20 teeth may serve as a pragmatic screening marker to identify older adults who may require nutritional assessment. The threshold of 20 teeth was shown to be a clinically relevant marker for nutritional assessment. Preventive strategies should focus on maintaining functional dentition through comprehensive oral healthcare, including regular dental examinations, preventive treatment, and timely restorative interventions.

For individuals who have already experienced significant tooth loss, our findings highlight the critical importance of appropriate prosthetic rehabilitation. The studies examining FTUs suggest that well-fitting dentures or other prosthetic replacements may help reduce the risk of malnutrition associated with tooth loss. However, the effectiveness of prosthetic interventions requires further investigation, as the current evidence provides limited insight.

CONCLUSION

This systematic review provides robust evidence that poor oral health is significantly associated with an increased risk of malnutrition and unintentional weight loss among older adults. The findings support the clinical importance of maintaining functional dentition and highlight the need for integrated approaches to oral health and nutritional care in aging populations. While methodological heterogeneity limits the precision of effect estimates, the consistency of findings across diverse populations and study designs strengthens the observed associations.

CONFLICTS OF INTEREST

There are no conflicts of interest.

ETHICAL POLICY AND INSTITUTIONAL REVIEW BOARD STATEMENT

Not applicable.

AUTHORS’ CONTRIBUTIONS

Concepts, design, and definition of intellectual content: ST, AR, and FC. Literature search and data acquisition: ST, KMT, AR, FC, OV, NT, H. Okubo, and TZT. Data analysis and statistical analysis: ST. Manuscript preparation: ST and KMT. Manuscript editing: KMT and FC. Manuscript review: AR, FC, OV, NT, H. Okubo, TZT, and H. Ogawa.

DECLARATION OF PATIENT CONSENT

Not applicable.

DATA AVAILABILITY STATEMENT

Not applicable.

DECLARATION OF GENERATIVE AI AND AI-ASSISTED TECHNOLOGIES IN THE WRITING PROCESS

During the preparation of this work, the authors used ChatGPT (OpenAI) in order to improve readability and language. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

LIST OF ABBREVIATIONS

• BDHQ Brief-type self-administered Diet History Questionnaire

• BMI Body mass index

• CI Confidence interval

• GRADE Grading of Recommendations Assessment, Development and Evaluation

• FTU Functional tooth unit

• MNA Mini-Nutritional Assessment

• NOS Newcastle–Ottawa scale

• PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses

• OR Odds ratio

SUPPLEMENTARY MATERIALS

SUPPLEMENTARY MATERIAL 1: SEARCH QUERY

• 1. Research question:

Does tooth loss affect nutritional status among the adults?

• 2. Searches
  • 1)Electronic search: A search of the literature will be conducted in the following databases: PubMed, Web of Science, EMBASE, CINAHL, Google Scholar, and ICHUSHI up to May 2024. The search will be limited to human studies.
  • 2)No language restrictions will be applied for searching studies.
• 1.Search Query
  • 1)PubMed

• 2)Web of Science

• 3)EMBASE

• 4)Google scholar

• 5)Ichushi (Japan Medical Abstracts Society Database)

SUPPLEMENTARY MATERIAL 2: NEWCASTLE–OTTAWA QUALITY ASSESSMENT SCALE

CROSS-SECTIONAL STUDIES

Selection: (Maximum 5 stars)

• 1). Representativeness of the sample:
  • a.Truly representative of the average in the target population. * (all subjects or random sampling)
  • b.Somewhat representative of the average in the target group. * (nonrandom sampling)
  • c.Selected group of users/convenience sample.
  • d.No description of the derivation of the included subjects.
• 2)Sample size:
  • a.Predetermined and calculated *
  • b.Not predetermined or calculated.
• 3)Nonrespondents:
  • a.Comparability between respondents and nonrespondents’ characteristics is established, and the response rate is satisfactory*(>90% response rate)
  • b.The response rate is unsatisfactory, or the comparability between respondents and nonrespondents is unsatisfactory (<90% response rate)
  • c.No description of the response rate or the characteristics of the responders and the nonresponders.
• 4)Ascertainment of the exposure (risk factor):
  • a.Validated measurement tool used. **
  • b.Nonvalidated measurement tool, but the tool is available or described*
  • c.No description of the measurement tool

Comparability: (Maximum 2 stars)

• 1)The subjects in different outcome groups are comparable, based on the study design or analysis.

• Confounding factors are controlled.

  • a.The study controls for the sex (select one)*
  • b.The study control for other demographic factors*

Outcome: (Maximum 3 stars)

• 1)Assessment of the outcome
  • a.Independent blind assessment**
  • b.Record linkage**
  • c.Self report*
  • d.No description
• 2)Statistical test:
  • a.Statistical test used to analyze the data clearly described, appropriate, and measures of association presented including confidence intervals and probability level (P value). *
  • b.Statistical test not appropriate, not described, or incomplete.

This scale has been adapted from the Newcastle–Ottawa Quality Assessment Scale for cohort studies to perform a quality assessment of cross-sectional studies for the systematic review, “Are Healthcare Workers’ Intentions to Vaccinate Related to their Knowledge, Beliefs and Attitudes? A Systematic Review”

Overall rating;

Total scores were calculated as percentages to account for total scores.

• ≤ 50%: low quality

• >50% - 75%: Moderate quality

• >75%: High quality

COHORT STUDIES

Note: A study can be awarded a maximum of one star for each numbered item within the Selection and Outcome categories. A maximum of two stars can be given for Comparability

Selection (Maximum 4 stars)

• 1)Representativeness of the exposed cohort
  • a)truly representative of the average adults (describe) in the community ↓
  • b)somewhat representative of the average adults in the community ↓
  • c)selected group of users eg nurses, volunteers
  • d)no description of the derivation of the cohort
• 2)Selection of the non exposed cohort
  • a)drawn from the same community as the exposed cohort ↓
  • b)drawn from a different source
  • c)no description of the derivation of the non exposed cohort
• 3)Ascertainment of exposure
  • a)secure clinical record or clinical examination ↓
  • b)structured interview ↓
  • c)written self report
  • d)no description
• 4)Demonstration that outcome of interest was not present at start of study
  • a)yes ↓
  • b)no

Comparability (Maximum 2 stars)

• 1)Comparability of cohorts on the basis of the design or analysis
  • a)study controls for demographic, denture (select the most important factor) ↓*
  • b)study controls for any additional factor (medical factor) ↓

(This criteria could be modified to indicate specific control for a second important factor.)

Outcome (Maximum 3 stars)

• 1)Assessment of outcome
  • a)independent blind assessment ↓
  • b)record linkage ↓
  • c)self report
  • d)no description
• 2)Was follow-up long enough for outcomes to occur
  • a)≥12 months (select an adequate follow-up period for outcome of interest) ↓
  • b)<12 months
• 3)Adequacy of follow-up of cohorts
  • a)complete follow-up - all subjects accounted for ↓
  • b)subjects lost to follow-up unlikely to introduce bias - small number lost - >20% (select an adequate %) follow-up, or description provided of those lost) ↓
  • c)follow-up rate <80% (select an adequate %) and no description of those lost
  • d)no statement

Overall rating (Maximum 9 stars):

Total scores were calculated as percentages to account for total scores.

• ≤50%: low quality

• >50%–75%: Moderate quality

• >75%: High quality

SUPPLEMENTARY MATERIAL 3: EXCLUDED STUDIES AT SECONDARY SCREENING

Supplementary Figure S1.

Funnel plot of the risk of bias for association between 20+/fewer and unintentional weight loss

Supplementary Figure S2.

Funnel plot of the risk of bias for association between edentate/dentate and malnutrition

Funding Statement

This research was supported by a grant from the Heiwa Nakajima Foundation, Tokyo, Japan.