Combined association of chewing capacity and depression with constipation: a cross-sectional study

Abstract

Background

This study aimed to investigate the independent and combined effects of chewing capacity and depression on the risk of constipation, using data from the National Health and Nutrition Examination Survey (NHANES) 2005–2010.

Methods

A total of 10,814 participants were included in the analysis. Chewing capacity was assessed using functional tooth units (FTUs), defined as the number of functional posterior occlusal units, and categorized into three groups (≤ 3, 3–9, and 10–12). Depression was defined as a Patient Health Questionnaire-9 (PHQ-9) score of 10 or higher. The PHQ-9 is a widely used self-administered instrument that assesses the severity of depressive symptoms over the preceding two weeks. Constipation was determined based on the Bristol Stool Form Scale (BSFS) and bowel movement. Multivariable logistic regression models were used to estimate odds ratios (ORs) for constipation. A restricted cubic spline (RCS) analysis was conducted to evaluate dose–response relationships.

Results

Participants with lower FTUs and higher PHQ-9 scores had a significantly increased risk of constipation. In multivariable-adjusted models, individuals with FTUs 10–12 exhibited a lowest constipation risk (OR = 0.637, 95% CI: 0.504–0.806), while depression was also strongly associated with constipation (OR = 1.942, 95% CI: 1.602–2.356). The joint analysis revealed that participants with both FTUs ≤ 3 and depression had the highest constipation prevalence (OR = 2.363, 95% CI: 1.600–3.489). Mediation analysis indicated that depression partially mediated the relationship between FTUs and constipation.

Conclusion

Both impaired chewing capacity and depression were independently linked to a higher likelihood of constipation, and their co-occurrence appeared to exert a synergistic impact. These findings highlight the importance of integrating oral health rehabilitation and mental health interventions to prevent and manage constipation in at-risk populations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-025-04123-3.

Introduction

Constipation is a common gastrointestinal disorder influenced by multiple factors, including diet, hydration, physical activity, and psychological well-being [1]. Emerging evidence suggests that oral health, particularly chewing capacity, plays a crucial role in digestive function [2, 3]. Functional tooth units (FTUs), which reflect the number of opposing posterior teeth available for mastication, are essential for efficient food breakdown and subsequent gut motility [4]. Impaired chewing capacity has been associated with altered dietary patterns, reduced fiber intake, and impaired digestion, all of which may contribute to constipation [5]. Despite this physiological relevance, the relationship between FTUs and constipation remains underexplored in epidemiological research.

Psychological factors, particularly depression, have also been implicated in constipation risk. Depression is linked to altered gut-brain axis function, autonomic nervous system dysregulation, and changes in neurotransmitter levels, all of which can impair gastrointestinal motility [6, 7]. The Patient Health Questionnaire-9 (PHQ-9) is a widely used screening tool for depression, and higher PHQ-9 scores have been associated with an increased prevalence of functional gastrointestinal disorders [8]. Prior studies have demonstrated that individuals with depression are more likely to experience constipation due to reduced physical activity, medication use, and altered central nervous system signaling [9, 10]. However, research examining the interplay between oral health, mental health, and constipation remains limited.

Most existing studies have independently investigated either chewing capacity or depression in relation to constipation, with little focus on their combined effects. Given the complex interactions between oral function, mental health, and gastrointestinal physiology, it is plausible that impaired chewing capacity and depression may jointly exacerbate constipation risk. Since both factors are prevalent among general populations, understanding their combined influence could provide critical insights for targeted interventions. To address this gap, the present study utilizes data from the National Health and Nutrition Examination Survey (NHANES) to examine the independent and joint associations of FTUs and PHQ-9 scores with constipation risk. By integrating oral health and mental health factors, this study aims to provide a more comprehensive understanding of constipation etiology and inform potential preventive strategies.

Materials and Methods

Study design and participants

NHANES utilizes a complex, stratified sampling design to obtain a nationally representative sample of the U.S. population. Its primary objective is to assess the health and nutritional status of individuals in the United States [11]. The survey is approved by the National Center for Health Statistics Institutional Review Board, and all participants provide written informed consent before participation. NHANES collects extensive data, including demographic characteristics, dietary intake, medical examinations, laboratory measurements, and questionnaire responses [12].

During the NHANES 2005–2010 cycle, a total of 28,237 individuals were initially included. Exclusions were made for participants with missing data on bowel-related variables, hypertension, BMI, education, alcohol intake, smoking history, diabetes, poverty level, uric acid, bilirubin, cardiovascular disease (CVD), hyperlipidemia, PHQ-9 score, functional tooth units (FTUs), or other relevant covariates (Fig. 1).

Fig. 1.

Flowchart of participant screening

Definition of constipation

In the NHANES dataset, constipation was identified based on stool frequency or stool consistency, as assessed through the Bowel Health Questionnaire (2005–2010) [13]. Participants assessed their usual stool consistency using the Bristol Stool Form Scale (BSFS), which categorizes stool into seven types. Constipation was defined as BSFS Type 1 (hard, separate lumps similar to nuts) or Type 2 (lumpy, sausage-like stools). Non-constipation included BSFS Type 3 (sausage-shaped with surface cracks), Type 4 (smooth and soft, resembling a sausage or snake), Type 5 (soft blobs with distinct edges), Type 6 (mushy with ragged edges), and Type 7 (entirely liquid without solid pieces [14, 15]. Bowel movement frequency was determined through self-reported data, where participants indicated how often they had bowel movements. Individuals reporting fewer than three bowel movements per week were also categorized as constipated [16].

Definition of depression and chewing capacity

The PHQ-9 is a widely used self-report tool for assessing depressive symptoms over the past two weeks [17]. It comprises nine items covering key depression symptoms, including low mood, anhedonia, sleep disturbances, fatigue, appetite changes, guilt, concentration difficulties, psychomotor changes, and suicidal thoughts. Each item is rated based on symptom frequency, yielding a total score from 0 to 27 [18]. A score ≥ 10 is generally used to indicate moderate to severe depressive symptoms, and is commonly employed as a cutoff for clinical depression [19, 20].

The primary exposure was chewing capacity, assessed during a dental examination conducted by trained and calibrated dentists. Chewing capacity was measured based on the number of functional posterior occlusal units, including premolars and molars, defined as FTUs [21]. The number of FTUs was defined as pairs of opposing natural or artificial teeth, excluding third molars. Two opposing premolars constituted one FTU, while two opposing molars were counted as two FTUs [22]. Participants were categorized into three groups: ≤ 3, 3–9, 10–12 [23].

Covariates

Potential confounding and modifying variables were selected based on prior research and clinical expertise. These included age, sex, races, education level, poverty-to-income ratio (PIR), body mass index (BMI), and smoking status. BMI was calculated as weight in kilograms divided by height in meters squared. Clinical indicators such as uric acid, creatinine, and total bilirubin were measured in the NHANES laboratory. Cardiovascular disease (CVD) diagnosis was based on self-reported physician diagnoses obtained through a standardized medical condition questionnaire. Alcohol drinking status was classified into four categories based on consumption patterns: never drinkers (lifetime abstainers), former drinkers (abstinent in the past year), moderate drinkers (1 or 2 drinks per day for females and males, respectively), and heavy drinkers (> 1 or > 2 drinks per day for females and males, respectively, and/or frequent binge drinking) [24, 25]. Diabetes Mellitus was defined by the presence of any of the following: (a) physician-diagnosed diabetes, (b) hemoglobin A1c (HbA1c) ≥ 6.5%, (c) fasting plasma glucose (FPG) ≥ 7.0 mmol/L, (d) random blood glucose ≥ 11.1 mmol/L, (e) 2-h post-load glucose during an oral glucose tolerance test (OGTT) ≥ 11.1 mmol/L, or (f) current use of insulin or other antidiabetic medications [26]. Hyperlipidemia was defined by meeting any one of the following criteria: (1) elevated triglycerides (TG) ≥ 150 mg/dL; (2) abnormal cholesterol profile, including total cholesterol (TC) ≥ 200 mg/dL, low-density lipoprotein (LDL) ≥ 130 mg/dL, or reduced high-density lipoprotein (HDL) levels— < 40 mg/dL for males and < 50 mg/dL for females; or (3) current use of lipid-lowering medications [26]. Variables such as dietary fiber intake, total fluid consumption, and physical activity were not included in the regression models due to inconsistent or missing data across the NHANES 2005–2010 cycles, which could have compromised sample size and model stability. Pearson correlation analyses among all covariates were performed and are shown in Supplementary Fig. 1.

Statistical analysis

NHANES employs a complex, multistage probability sampling design involving stratification, clustering, and sampling weights. Statistical estimation and inference in this context are based primarily on the survey design structure rather than assumptions about the underlying data distribution. Unlike conventional analyses, the validity of parametric tests in complex survey data does not depend on normality assumptions. Instead, the key lies in the appropriate incorporation of sampling weights to produce unbiased, population-representative estimates. The weighted Student’s t-test is frequently used in this setting, as it accounts for design features such as sampling weights, stratification variables, and primary sampling units, thereby ensuring that statistical inferences accurately reflect the U.S. civilian non-institutionalized population [27]. The baseline characteristics of participants were summarized and compared between constipation and non-constipation groups. Continuous variables were expressed as mean (95% confidence interval) and compared using either a t-test or Wilcoxon rank-sum test, depending on the results of the Kolmogorov–Smirnov normality test. Categorical variables were presented as frequencies (percentages) and compared using the Chi-square test.

Additionally, the potential nonlinear associations between FTUs, PHQ-9 score, and constipation were explored using restricted cubic spline (RCS) curves, a flexible regression method that models complex dose–response relationships by fitting piecewise polynomial functions at specified percentiles (knots) of the predictor variable distribution [28]. Multivariable-adjusted logistic regression analyses were applied to determine the odds ratio (OR) alongside a 95% confidence interval (CI) for assessing the relationship between FTUs, PHQ-9 score and constipation. To reduce the risk of overadjustment while maximizing data availability, we developed two multivariable models. Model 1 adjusted for age, sex, BMI, education level, poverty index, smoking status, alcohol consumption, and marital status. Model 2 included all covariates from Model 1 and further adjusted for diabetes, races, serum creatinine, total bilirubin, uric acid, and hyperlipidemia. We also made variance inflation factor (VIF) test and Box-Tidwell test to check the linearity assumption in logistic regression models. To control for multiple testing, p-values were adjusted using the Benjamini–Hochberg (BH) procedure to control the false discovery rate (FDR).

To assess the combined effects of PHQ-9 score and chewing capacity on constipation, we created dummy variables representing four joint exposure categories. These were based on cross-classification of depression status (depression vs. non-depression) and FTU categories (≤ 3, 3–9, 10–12). The reference group consisted of individuals without depression and with FTU values between 10 and 12. We then evaluated the association of each joint exposure group with the risk of constipation. To assess the robustness of these associations to potential unmeasured confounding, we calculated the E-value, which represents the minimum strength of association that an unmeasured confounder would need to have with both the exposure and the outcome to fully explain away the observed effect. To further explore the potential mediating role of depressive symptoms in the relationship between chewing capacity and constipation, we conducted a mediation analysis using a weighted regression framework. In this model, chewing capacity (measured by FTUs) was treated as the exposure, constipation as the binary outcome, and the PHQ-9 score as the mediator. We estimated the indirect effect by applying the product-of-coefficients method, in which the effect of FTUs on PHQ-9 score was multiplied by the effect of PHQ-9 score on constipation. Both linear and logistic regression models were weighted to account for the complex NHANES survey design. Statistical significance of the total, direct, and indirect effects was assessed using the bootstrap method with resampling, providing robust confidence intervals for mediation estimates [29]. To perform sensitivity analysis, comparison between included and excluded participants and multiple imputations for missing variables were performed in Supplementary Table 1, and inverse probability of weighting (IPW) regression analysis was used on the unweighted raw data to deal with potential confounders. We also conducted model diagnostics using McFadden's R-squared and the Akaike Information Criterion (AIC), as shown in Supplementary Table 2. Forest plots illustrating effect sizes across different subgroups are provided as well in Supplementary Fig. 2.

Table 1.

Weighted characteristics between constipated and non-constipated participants

	Overall (n = 10814)	Non-constipation (n = 9397)	Constipation (n = 1417)	P value
Age (years)	47.9 (47.2,48.7)	48.2 (47.5,49.0)	45.8(44.7,46.9)	< 0.001
Sex (Male %)	5584(49.9)	5113(53.0)	471(28.2)	< 0.0001
BMI (kg/m2)	28.7(28.5,28.9)	28.9 (28.6,29.1)	27.7(27.2,28.1)	< 0.0001
Race (%)				< 0.0001
Black	2062 (9.842)	1694 (9.037)	368 (15.461)
Other	3127(16.381)	2772(16.636)	355(14.598)
White	5625(73.777)	4931(74.327)	694(69.941)
Poverty income ratio	3.161(3.073,3.248)	3.211(3.123,3.298)	2.812(2.690,2.933)	< 0.0001
Education (%)				< 0.0001
Less Than High School	2945(17.566)	2520(17.204)	425(20.093)
More Than High School	5282(58.337)	4676(59.288)	606(51.698)
High School	2587(24.097)	2201(23.508)	386(28.209)
Total bilirubin (mg/dL)	0.766(0.756,0.775)	0.774(0.764,0.783)	0.709(0.684,0.734)	< 0.0001
Uric acid (mg/dL)	5.476(5.431,5.520)	5.539(5.493,5.584)	5.039(4.952,5.126)	< 0.0001
BUN (mg/dL)	13.070(12.861,13.279)	13.154(12.939,13.369)	12.485(12.207,12.764)	< 0.0001
Creatinine (mg/dL)	0.903(0.895,0.911)	0.907(0.898,0.916)	0.873(0.855,0.891)	0.002
Smoke (%)				< 0.001
Former	2897(25.977)	2599(26.810)	298(20.164)
Never	5514(51.906)	4722(51.294)	792(56.175)
Current	2403(22.117)	2076(21.896)	327(23.661)
Alcohol drink (%)				0.005
Former	2204(16.651)	1889(16.366)	315(18.636)
Heavy	2145(20.551)	1882(20.666)	263(19.748)
Mild	3522(36.000)	3112(36.587)	410(31.907)
Moderate	1580(16.483)	1388(16.566)	192(15.901)
Never	1363(10.316)	1126 (9.815)	237(13.809)
Marital status				< 0.0001
Married	6806(67.037)	6006(67.891)	800(61.072)
Never married	1478(13.854)	1232(13.455)	246(16.637)
Separated	2530 (19.110)	2159 (18.654)	371 (22.291)
FTU	8.299(8.066,8.532)	8.368(8.142,8.594)	7.817(7.411,8.223)	0.002
≤ 3	3076(20.015)	2634 (19.469)	442 (23.826)
3–9	2607 (21.685)	2256 (21.556)	351 (22.587)
10–12	5131(58.300)	4507(58.976)	624 (53.587)
CVD (%)	1188 (8.245)	1019(8.145)	169 (8.939)	0.355
PHQ-9 score	2.846(2.710,2.982)	2.701(2.570,2.833)	3.855(3.534,4.176)	< 0.0001
Depression (%)	884 (6.780)	686 (5.922)	198(12.765)
Hyperlipidemia (%)	8075(73.634)	7073(74.107)	1002(70.331)	0.025
Diabetes Mellitus (%)	1928 (12.878)	1708 (13.120)	220 (11.189)	< 0.0001

Abbreviation: body mass index (BMI), blood urea nitrogen (BUN), cardiovascular disease (CVD), functional tooth units (FTU), Patient Health Questionnaire-9 (PHQ9)

Fig. 2.

Restricted cubic spline (RCS) for the association between PHQ-9 score and chewing capacity with the risks of constipation

A two-sided P < 0.05 was considered statistically significant. All analyses were performed using SPSS version 26.0 (IBM Corp, Armonk, NY) and R (version 4.3.2) [30, 31].

Results

Study participants and baseline characteristics

In the final cohort, 10,814 American adults were included, which is considered representative of a population-based sample size of 149, 990, 493 participants, of whom 1,417 were classified as constipated (Fig. 1 and Table 1). Compared with non-constipated participants, those with constipation were younger (45.8 vs. 48.3 years, P < 0.001), more likely to be female (71.8% vs. 46.9%, P < 0.0001), and had lower BMI (27.7 vs. 28.9 kg/m², P < 0.0001), socioeconomic status (poverty-income ratio 2.81 vs. 3.21, P < 0.0001), and educational attainment (20.1% vs. 17.2% with less than high school, P < 0.0001). They also had lower levels of uric acid, total bilirubin, BUN, and creatinine. The constipation group exhibited a higher proportion of never smokers (56.2% vs. 51.3%) and never drinkers (13.8% vs. 9.8%), and were less likely to be married (61.1% vs. 67.9%, all P < 0.01). Importantly, they had fewer functional tooth units (FTUs: 7.82 vs. 8.37, P = 0.002), and a higher percentage of participants with FTUs ≤ 3 (23.8% vs. 19.5%). The constipation group also demonstrated significantly higher PHQ-9 scores (3.86 vs. 2.70, P < 0.0001) and greater prevalence of depression (12.8% vs. 5.9%). The prevalence of diabetes mellitus was slightly lower among constipated individuals (11.2% vs. 13.1%, P < 0.0001), whereas cardiovascular disease showed no significant group difference (P = 0.355).

Correlation Between PHQ-9 Score, chewing capacity, and constipation

The relationship between PHQ-9 score, chewing capacity, and constipation was assessed using RCS curves (Fig. 2). In Fig. 2A, the RCS analysis demonstrated that chewing capacity, as measured by FTUs, was inversely associated with constipation risk, showing a linear trend (P overall < 0.001, P non-linear = 0.7247), suggesting that a lower number of functional tooth units might contribute to a greater risk of constipation. In contrast, PHQ-9 score exhibited a positive and nearly linear association with constipation risk (P overall < 0.001, P non-linear = 0.9202), indicating that higher depressive symptoms were correlated with an increased probability of constipation (Fig. 2B).

Associations of PHQ-9 Score and chewing capacity with constipation

Multivariable logistic regression analysis was conducted to evaluate the associations between PHQ-9 score, chewing capacity, and constipation (Tables 2 and 3). Among individuals without depression, 12.28% (1219/9930) reported constipation, compared to 22.40% (198/884) among those with depression. In the unadjusted model (Model 0), a higher PHQ-9 score was linked to a greater likelihood of constipation (P < 0.001). This association remained robust after adjusting for age, sex, BMI, education, poverty, smoke, alcohol drink and marital status in Model 1. In the fully adjusted Model 2, which additionally accounted for diabetes mellitus, races, creatinine, total bilirubin, uric acid and hyperlipidemia, the PHQ-9 score continued to show a significant positive association with constipation risk, reinforcing the relationship between depression and constipation. Each one-point increase in PHQ-9 score was linked to a 5.0% increase in constipation risk, while individuals meeting criteria for depression had a 94.2% higher risk compared to those without depression.

Table 2.

Multivariable-adjusted logistic regression analysis of the association between chewing capacity (FTU), depression (PHQ-9), and constipation

	No. of participants	No. of outcomes of constipation	Model 0	Model 1a	Model 2b
FTU	10,814	1417	0.973(0.957,0.989) ***	0.952(0.933,0.971) **	0.957(0.937,0.977) **
≤ 3	3076	442	ref	ref	ref
3–9	2607	351	0.856(0.715,1.026)	0.780(0.632,0.963) *	0.809(0.651,1.005)
10–12	5131	624	0.742(0.620,0.889) **	0.603(0.481,0.756) ***	0.637(0.504,0.806) ***
PHQ-9 score	10,814	1417	1.066(1.051,1.082) ***	1.051(1.034,1.068) **	1.050(1.032,1.067) **
Non-depression	9930	1219	ref	ref	ref
Depression	884	198	2.324(1.936,2.791) ***	1.971(1.624,2.392) **	1.942(1.602,2.356) **
Joint variable
Q1	4837	559	ref	ref	ref
Q2	294	65	2.313(1.618,3.305) ***	2.051(1.420,2.963) ***	2.056(1.416,2.985) ***
Q3	2350	291	1.074(0.887,1.299)	1.236(1.021,1.496) *	1.220(1.008,1.476) *
Q4	257	60	3.007(2.152,4.200) ***	2.927(2.006,4.269) ***	2.749(1.859,4.065) ***
Q5	2743	369	1.322(1.095,1.597) **	1.705(1.343,2.164) ***	1.614(1.258,2.069) ***
Q6	333	73	2.320(1.680,3.203) ***	2.480(1.710,3.597) ***	2.363(1.600,3.489) ***

^a Model 1 adjusted for age, sex, BMI, education, poverty, smoke, alcohol drink and marital status

^b Model 2 adjusted for age, sex, BMI, education, poverty, smoke, alcohol drink, marital status, diabetes, races, creatinine, total bilirubin, uric acid, and hyperlipidemia

Covariates were selected based on baseline characteristics with P < 0.05 in Table 1

^*P < 0.05, **P < 0.01, ***P < 0.001

Q1: Non-depression + 10 ≤ FTU ≤ 12

Q2: Depression + 10 ≤ FTU ≤ 12

Q3: Non-depression + 3 < FTU ≤ 9

Q4: Depression + 3 < FTU ≤ 9

Q5: Non-depression + FTU ≤ 3

Q6: Depression + FTU ≤ 3

Table 3.

Sensitivity analyses by inverse probability of weighting

Variable	OR (95% CI)
FTU	0.956 (0.932, 0.982)
PHQ-9 score	1.051 (1.034, 1.070)

Approximately 14.4% of participants with FTUs ≤ 3 (442/3,076) had constipation, compared to 13.5% among those with FTUs 4–9 and 12.6% among those with FTUs ≥ 10. When evaluating chewing capacity, participants with higher FTUs had a significantly lower risk of constipation in Model 0 (P < 0.001). This protective association persisted after adjusting for sociodemographic and health-related covariates in Model 1. In the fully adjusted Model 2, individuals with FTUs > 10 remained at a significantly lower risk of constipation compared to those with FTUs ≤ 3, indicating a strong negative correlation between chewing capacity and constipation. Each one-unit increase in FTUs was associated with a 4.3% decrease in the likelihood of constipation. Compared to individuals with FTUs ≤ 3, those with 3–9 and 10–12 FTUs had 19.1% and 36.3% lower risk of constipation, respectively. In the fully adjusted Model 2, both chewing capacity and depressive symptoms remained significantly associated with constipation after FDR correction using the BH method. The adjusted p-values were 0.003 for FTUs and 0.0004 for PHQ-9 score, respectively, and details were shown in supplementary Table 5.

In supplementary Table 3 and Table 4, we made VIF and Box-Tidwell test to evaluate the adjusted factors in model 2 and all confounding factors showed no multicollinearity.

Combined association of PHQ-9 Score and chewing capacity with constipation

A joint analysis was performed to investigate the combined effects of PHQ-9 score and chewing capacity on constipation risk. The results (Table 2) demonstrated that individuals with both high PHQ-9 scores (depression) and low chewing capacity (FTUs ≤ 3) had the highest risk of constipation. Compared to individuals with low PHQ-9 scores (non-depression) and high chewing capacity (FTUs 10–12), those with high PHQ-9 scores and low chewing capacity had an odds ratio (OR) of 2.363 (95% CI: 1.600–3.489) for constipation. Furthermore, individuals with high PHQ-9 scores and moderate chewing capacity (FTUs 3–9) were more likely to experience constipation (OR: 2.749, 95% CI: 1.859–4.065). These findings suggest that depressive symptoms and reduced chewing capacity have a synergistic effect in increasing constipation risk, highlighting the importance of addressing both psychological and oral health factors in constipation management.

Sensitivity analyses

For the chewing capacity analysis, FTUs of 10–12 had an E-value of 2.52, while the E-value for FTUs treated as a continuous variable was 1.26. For individuals with depression, the E-value was 3.29. For the joint effect of FTUs and PHQ-9 score, the Q6 group (depression + FTUs ≤ 3) had an E-value of 4.16.

After IPW, the OR of FTU was 0.956, and OR of PHQ-9 score was 1.051.

Mediation analysis

Figure 3 illustrates the mutual mediation effects among FTUs, depression, and constipation. The PHQ-9 score significantly mediated 7.50% (95% CI: 3.67%–11.30%) of the association between FTUs and constipation, indicating that a small but significant portion of the total effect of chewing capacity on constipation operates indirectly through depressive symptoms.

Fig. 3.

Mediation analysis of PHQ-9 score on FTUs and constipation

Discussion

This study offers important insights into the complex causes of constipation. We examined how chewing capacity, measured by functional tooth units (FTUs), and depressive symptoms, measured by PHQ-9 scores, relate to constipation. Both reduced FTUs and higher PHQ-9 scores were individually linked to greater constipation risk. Notably, the risk was highest when both conditions were present. These findings suggest that oral health and mental health should be considered alongside dietary and lifestyle factors in understanding and managing constipation.

In line with our findings, population-based studies from diverse geographic regions—including Asia and New Zealand, older adults and children—have reported similar associations between impaired oral health, depression, and functional gastrointestinal disorders [32–34]. Several mechanisms may explain how chewing capacity affects constipation. First, effective chewing, supported by adequate functional tooth units (FTUs), helps break down food mechanically and improves nutrient absorption [5]. When chewing is impaired, individuals may avoid fibrous foods, leading to lower fiber intake—a key factor in healthy bowel movements [35]. Recent studies also suggest that impaired oral health can contribute to systemic inflammation, which has been implicated in gastrointestinal dysmotility and gut-brain axis dysfunction [36, 37]. Additionally, inadequate mastication may alter the gut microbiota composition by limiting prebiotic substrate availability, further exacerbating constipation risk [38]. In addition to microbiota and neuroendocrine pathways, systemic inflammation may serve as a common mechanistic link among impaired oral health, depression, and constipation. Chronic periodontal disease and tooth loss have been associated with elevated systemic inflammatory markers, such as C-reactive protein and interleukin-6 [39, 40].

Depression has emerged as an important factor in gastrointestinal disorders, including constipation. Epidemiological studies have shown that individuals with depression exhibit a higher prevalence of functional constipation, even after adjusting for lifestyle factors and comorbidities [41, 42]. A large-scale cohort study by Wang et al. found that depressive symptoms predicted a higher likelihood of developing chronic constipation, regardless of dietary fiber and moisture intake [9]. Mechanistically, the gut-brain axis plays a pivotal role in linking depression and constipation, with dysfunctions in neuroendocrine signaling, gut motility, and intestinal microbiota composition implicated in this association [43, 44]. Depression-induced hyperactivation of the hypothalamic–pituitary–adrenal (HPA) axis leads to elevated cortisol levels, which in turn impair colonic motility and disrupt the intestinal barrier [45, 46]. Altered serotonergic signaling, a hallmark of both depression and functional constipation, further exacerbates gut dysmotility by reducing serotonin availability in the enteric nervous system [47]. Recent systematic reviews and meta-analyses further strengthen the evidence linking oral health, depression, and gastrointestinal function. A comprehensive review by Kumar et al. highlighted the critical role of mastication in promoting gut motility and nutrient absorption, suggesting that impaired chewing may directly contribute to constipation through altered dietary intake and mechanical digestive efficiency [5]. Meta-analytic data have also confirmed the association between depressive symptoms and functional gastrointestinal disorders, including constipation [48]. For instance, Wang et al. reported that individuals with depression were significantly more likely to experience constipation, independent of dietary and lifestyle factors [49]. Moreover, recent Mendelian randomization studies support a potential causal relationship between major depressive disorder and constipation, further implicating depression as an etiological factor rather than a mere correlate [41].

Concurrently, recent reviews on the gut-brain axis emphasize the role of gut microbiota dysbiosis as a shared biological substrate for both depression and constipation. Liu et al. and Shen et al. noted that decreased abundance of short-chain fatty acid (SCFA)-producing bacteria in depressed individuals may impair intestinal barrier integrity and colonic motility, leading to functional bowel disturbances [50, 51]. These findings underscore the therapeutic potential of microbiota-targeted interventions such as probiotics or dietary modulation in patients with overlapping depressive and gastrointestinal symptoms.

Importantly, our study’s outcome definitions were aligned with the Rome IV diagnostic criteria for functional constipation, which emphasize both stool consistency (as measured by the Bristol Stool Form Scale) and bowel movement frequency as diagnostic indicators. Published in 2016, the Rome IV consensus statement remains the global standard for classifying functional gastrointestinal disorders and provides a validated framework for population-based research [52]. Our application of Rome IV-compatible definitions enhances the clinical relevance and comparability of our findings across international studies.

Depression may serve as a crucial mediator in the association between reduced FTUs and constipation through multiple physiological and behavioral mechanisms. Prior studies have established a connection between tooth loss and depression. Cross-sectional analyses have consistently identified a significant association between edentulism and depressive symptoms [53, 54]. Furthermore, an instrumental variable analysis provided evidence supporting a causal link, showing that each additional missing tooth increased the risk of clinical depression by 0.81 percentage points among US adults [55]. Impaired chewing capacity limits dietary choices, often leading to lower fiber intake and reduced mastication-induced gut stimulation, both of which contribute to colonic dysmotility. Simultaneously, depression exacerbates these effects by further decreasing appetite, altering autonomic nervous system regulation, and increasing HPA axis activity, which in turn impairs gastrointestinal motility [43, 56, 57]. Moreover, depression is associated with gut microbiota dysbiosis and serotonergic dysfunction, both of which are implicated in constipation pathophysiology [58]. Together, these clinical and mechanistic insights support the notion that oral health impairment and depression may synergistically influence constipation through overlapping pathways involving the gut-brain axis, mastication-mediated nutrient sensing, and systemic inflammation. This integrated perspective provides a strong rationale for incorporating dental care and mental health screening into comprehensive gastrointestinal management strategies. However, in our mediation analysis, depression accounted for only 6.0% of the total association between chewing capacity and constipation, indicating that the majority of the relationship is likely driven by alternative pathways. Masticatory ability may influence digestive function through several physiological mechanisms independent of psychological status. For example, impaired chewing results in larger food particle sizes, which can delay gastric emptying and alter intestinal motility, thereby contributing to constipation [59, 60].

From a clinical perspective, our findings suggest the importance of early screening and supportive interventions for individuals with depression and impaired chewing capacity. Routine assessment of mastication, mental health, and bowel habits could help identify high-risk patients. These strategies align with recent clinical practice guidelines on constipation management, which emphasize individualized treatment plans incorporating lifestyle modification, psychological assessment, and pharmacological interventions when necessary [61, 62]. In parallel, updated recommendations for depression screening and treatment advocate for regular mental health evaluations for all adults, including pregnant and postpartum women. Effective treatment generally involves antidepressants or psychotherapy, or a combination of both, with special considerations for pregnant or breastfeeding women [63]. Moreover, integrated care models, supported by growing evidence from multidisciplinary interventions, underscore the value of coordinated efforts among dental professionals, mental health providers, and primary care clinicians [64]. Implementing such comprehensive strategies may enhance the effectiveness of constipation prevention and management in vulnerable populations. However, given the cross-sectional design of this study, these interpretations should be made with caution. The associations observed do not establish causality, and further prospective or interventional studies are needed to validate these findings and explore potential therapeutic strategies [65]. Interventions that incorporate oral rehabilitation, psychological support, and dietary modification may ultimately reduce constipation burden and improve quality of life in vulnerable populations. Although the cross-sectional nature of our study precludes determination of causality, evidence from longitudinal studies and instrumental variable analyses offers additional support for the hypothesized pathways. For example, a population-wide natural experiment using instrumental variable methods found that tooth loss was causally associated with an increased risk of depression [55]. These findings help contextualize our results by supporting the possibility that impaired oral function may precede the development of depressive symptoms, which in turn may contribute to gastrointestinal dysregulation, including constipation.

Furthermore, previous systematic reviews on constipation risk factors and management highlight the critical role of lifestyle, dietary, and medical influences. These reviews consistently underscore the importance of adequate fiber and fluid intake, physical activity, and medication review. Effective management strategies, therefore, should integrate patient education, behavioral modifications, and appropriate therapeutic interventions. Such evidence-based, multidisciplinary approaches offer a promising framework for addressing the complex interplay between oral health, mental well-being, and gastrointestinal function [66, 67].

Future research should prioritize longitudinal cohort studies to determine the temporal and potentially causal relationships between impaired chewing function, depression, and constipation. Intervention studies are also needed to evaluate whether improving oral health (e.g., through dental rehabilitation) and managing depressive symptoms can alleviate constipation symptoms [68]. Additionally, mechanistic studies exploring the biological pathways—including neuroendocrine, inflammatory, and microbiota-mediated mechanisms—linking these conditions would enhance our understanding and inform targeted prevention strategies [69]. Interventions that incorporate oral rehabilitation, psychological support, and dietary modification may ultimately reduce constipation burden and improve quality of life in vulnerable populations.

Although the E-values for chewing capacity (2.52) and depression (3.29) suggest moderate robustness to unmeasured confounding, they do not rule out the possibility of hidden bias. For comparison, E-values greater than 3.0 are commonly reported in studies of cardiovascular risk factors, reflecting higher levels of confounder robustness [70]. In our study, it would require an unmeasured confounder with a relative risk of at least 2.5 with both the exposure and the outcome to fully account for the observed associations [71]. While such confounding is possible, particularly from factors like early-life adversity, genetic predisposition, or unrecorded dietary and behavioral variables, these would need to be relatively strong and correlated with both oral health and bowel function. Therefore, our findings appear reasonably robust, though future longitudinal studies are needed to confirm these associations.

Several limitations should be acknowledged. First and foremost, the cross-sectional design represents a fundamental limitation of this study. This design prevents the establishment of temporal relationships or causal inferences, which is particularly problematic when evaluating potentially bidirectional associations among depression, chewing capacity, and constipation [72]. For instance, chronic constipation may not only result from depression and reduced FTUs but may also contribute to depressive symptoms and poor oral health via diminished quality of life, appetite changes, and reduced self-care. While mediation analysis was conducted to explore indirect associations, it cannot determine causality in the absence of longitudinal data. Second, depression was assessed using the self-reported PHQ-9, which may be influenced by somatic symptoms such as constipation-related discomfort, potentially introducing misclassification bias. Future studies incorporating clinical diagnostic tools or excluding participants with chronic somatic conditions may help validate these findings. Additionally, several important confounders—such as medication use (e.g., opioids, antidepressants, anticholinergics), dietary fiber and fluid intake, physical activity, and comorbidities like irritable bowel syndrome or thyroid dysfunction—were not included in our models due to inconsistent availability across NHANES cycles, which may contribute to residual confounding. Furthermore, chewing capacity was assessed based on FTUs without accounting for the functional status of dental prostheses, potentially overestimating masticatory ability in edentulous individuals, particularly those with poorly fitting dentures [22]. Although the large, nationally representative NHANES sample enhances the generalizability of our findings, selection bias remains possible, as individuals with severe depression or poor oral health may be less likely to participate in dental examinations or complete questionnaires. Lastly, due to the use of secondary data from a public database, we were unable to conduct a formal power analysis to assess the adequacy of statistical power for detecting clinically meaningful associations.

Conclusion

Our findings demonstrate that both depressive symptoms and reduced FTUs were independently associated with an elevated risk of constipation. Notably, the co-occurrence of depression and diminished chewing capacity (FTUs ≤ 3) was linked to the highest constipation prevalence, indicating a synergistic interaction between psychological and oral health factors. These results underscore the critical need to integrate oral health rehabilitation and mental health interventions into public health strategies aimed at mitigating constipation risk. Future prospective cohort studies with longitudinal designs and objective measurements are warranted to validate these associations, clarify causal pathways, and explore targeted interventions addressing these modifiable risk factors.

Authors’ contributions

JFH, YJX conceived and designed the study, acquired the data and drafted the manuscript; TL and XDM analyzed the data; JFH and TL contributed to the interpretation of the results and critical revision of the manuscript for important intellectual content; XDM developed the software and provided technical support. TL had the primary responsibility for final content. All authors have read and approved the final manuscript. The authors reported no conflicts of interest.

Funding

None.

Data availability

The raw data supporting the conclusions of this article can be found here: https://www.cdc.gov/nchs/nhanes/.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Consent to participate was obtained and the National Center for Health Statistics (NCHS) ethics committee approved NHANES study protocol. Study protocols for NHANES were approved by the NCHS ethnics review board (Protocol #2011–17). All information from the NHANES program is available and free for the public, so the agreement of the medical ethics committee board was not necessary. All participants provided written informed consent to participate in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.