Effect of Physical Activity on the Association Between Diet and Constipation: Evidence From the National Health and Nutrition Examination Survey 2007-2010

Abstract

Background/Aims

Previous studies have shown that diet and physical activity can influence constipation. However, the combined effect of diet and physical activity on constipation remains unclear.

Methods

Constipation was defined based on stool consistency and frequency, while overall diet quality was assessed using Healthy Eating Index (HEI)-2015 scores. Participants were categorized into low (metabolic equivalent [MET]-min/wk < 500) and high physical activity groups (MET-min/wk ≥ 500). The association between diet and constipation across physical activity groups was analyzed using survey logistic regression and restricted cubic splines.

Results

Higher HEI-2015 scores were associated with reduced constipation risk in the high physical activity group when constipation was defined by stool consistency (odds ratio [OR], 0.98; 95% confidence interval [CI], 0.97-0.99). However, in the low physical activity group, increased HEI-2015 scores did not significantly affect constipation risk (OR, 1.01; 95% CI, 0.97-1.05). Similar results were found when constipation was defined based on stool frequency. In the high physical activity group, increased HEI-2015 scores were significantly associated with a reduced constipation risk (OR, 0.96; 95% CI, 0.94-0.98). Conversely, in the low physical activity group, increased HEI-2015 scores did not affect the risk of constipation (OR, 0.96; 95% CI, 0.90-1.03).

Conclusions

Our findings suggest that a higher HEI-2015 score is negatively associated with constipation among individuals with high physical activity levels but not among those with low physical activity levels. This association was consistent when different definitions of constipation were used. These results highlight the importance of combining healthy diet with regular physical activity to alleviate constipation.

Introduction

Constipation is a prevalent gastrointestinal condition that affects a significant proportion of the adult population worldwide.¹ It is characterized by infrequent and difficult bowel movements, leading to discomfort and a decreased quality of life.^2,3 The societal and economic burdens of constipation are substantial, necessitating effective management strategies.⁴

Numerous modifiable behaviors, including exercise, dietary modifications, pharmacotherapy, and physical therapy, have been investigated in relation to the risk and management of constipation.^5,6 Physical activity (PA) improves bowel function by enhancing gastrointestinal motility and reducing transit time.^7,8 Dietary factors, particularly fiber intake and fruit and vegetable consumption, play crucial roles in promoting regular bowel movements and preventing constipation.⁹ The Healthy Eating Index (HEI) is a widely used measure to assess overall diet quality, with the HEI-2015 aligned with the Dietary Guidelines for Americans.^10-14 Higher HEI scores indicate greater adherence to the recommended intake levels and reflect a healthier diet.¹⁵ Recent studies have demonstrated an inverse association between HEI-2015 and constipation risk in the United States (US).¹³

However, previous investigations have predominantly focused on examining the independent effects of diet quality and PA on constipation, with limited attention being given to their combined effects. Consequently, this study aims to address this research gap by using data from the National Health and Nutrition Examination Survey (NHANES) and Food Pattern Equivalence Database (FPED) to explore the potential association between diet and constipation across different levels of PA. By elucidating the combined effects of diet and PA on constipation, this study aims to provide valuable insights into the development of comprehensive strategies for the prevention and management of constipation. Understanding the interplay between these lifestyle factors can guide healthcare professionals and individuals seeking to alleviate constipation symptoms and enhance overall gastrointestinal well-being.

Materials and Methods

Data Source

This cross-sectional study used data from the NHANES database. The NHANES employs a complex, stratified, multistage probabilistic cluster design to collect comprehensive health and nutritional data from a representative sample of the US population.

This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the National Center for Health Statistics (NCHS) Institutional Review Board (continuation of protocol: #2005-06). Written informed consent was obtained from all participants. Because the NHANES data were de-identified, institutional review board approval was not required.

We utilized data from 2 NHANES cycles, specifically the 2007-2008 and 2009-2010 cycles, which included a total of 12 153 participants aged 20 years or older. These cycles include specific questionnaires related to gut health. Participants with incomplete gut health, dietary recall, or PA questionnaire data were excluded from the analysis. Extreme energy intake may be due to diet, obesity, disease, or other factors. When statistical analyses are performed, the dietary behaviors of these individuals may have a significant impact on the overall results and are not reflective of the dietary patterns of the general population. Consequently, individuals with exceptionally high energy intake values (below 800 kcal/day and above 8000 kcal/day for males, and below 500 kcal/day and above 5000 kcal/day for females) were excluded from our study.¹³ Moreover, individuals who self-reported having ulcerative colitis, Crohn’s disease, celiac disease, or colon cancer, or who were pregnant were also excluded. After applying these exclusion criteria, 5309 participants were included in the final analysis (Fig. 1).

Figure 1.

Study flowchart. NHANES, National Health and Nutrition Examination Survey.

Definition of Constipation

Bowel health was assessed using computer-assisted personal interviews conducted in dedicated interview rooms. The participants completed a bowel health questionnaire that included information on stool consistency and frequency over the past 30 days. Stool consistency was evaluated using the Bristol stool form scale (BSFS),¹⁶ which consists of color pictures and detailed descriptions of 7 stool types (type 1 [separate hard lumps, like nuts], type 2 [sausage-like, but lumpy], type 3 [like a sausage but with cracks in the surface], type 4 [like a sausage or snake, smooth and soft], type 5 [soft blobs with clear-cut edges], type 6 [fluffy pieces with ragged edges, a mushy stool], and type 7 [watery, no solid pieces]). Participants were shown a card displaying these stool types and were asked to identify the number corresponding to their usual or most common stool type. Stool frequency was assessed by asking the participants how often they typically had bowel movements.

Owing to the limited association observed between stool consistency and stool frequency in prior NHANES data, previous NHANES studies have consistently used either stool consistency or stool frequency as criteria for defining constipation. Therefore, in this study, we used 2 distinct definitions of constipation.^17,18 The first definition defined constipation as having fewer than 3 stools per week.¹⁹ The second definition classifies constipation as reporting stool type 1 or 2 on the BSFS.²⁰

Healthy Eating Index

Dietary intake data were collected through two 24-hour recall interviews conducted by trained dietary interviewers.²¹ The first interview was conducted face-to-face, while the second interview took place over the telephone 3-10 days later. During these interviews, participants were asked to recall the types and amounts of food and beverages they had consumed in the past 24 hours. The average of the 2 sets of recall data was used to estimate dietary intake.²² The Food and Nutrition Database for Dietary Studies,¹³ NHANES Personal Food Data, and FPED were used to calculate energy and nutrient intake for all food items.

The HEI was used to assess the overall diet quality. The HEI-2015 score was derived from the obtained dietary data and reflected adherence to dietary guidelines. The US Department of Agriculture Food Pattern Equivalency Database was used to convert the dietary data into standardized quantities of food groups to calculate the HEI-2015 score. The HEI-2015 score is not based on the absolute amounts of food components but rather on their energy density per 1000 kcal. It comprises 13 components, including 9 adequacy components: total fruit, whole fruit, total vegetables, vegetables and legumes, total protein foods, seafood, plant protein (0-5 points each), whole grains, dairy products, and fatty acids (0-10 points each).¹⁴ Higher scores indicated a higher intake of these components. In addition, 4 moderating components were considered: sodium, refined grains, added sugars, and saturated fats (0-10 points each). Higher scores for these components indicate lower intake. SAS codes were used to calculate HEI-2015 scores.²³

Physical Activity

PA information was collected using the Global Physical Activity Questionnaire,²⁴ developed by the World Health Organization. Participants were asked to report their PA behavior over the past 30 days. The questionnaire assessed 3 PA levels: vigorous work/recreational activity, moderate work/recreational activity, and walking/cycling activity. The participants provided information on the number of days per week and the amount of time (in minutes) they engaged in each type of PA during a typical week. To determine PA intensity, weekly MET were calculated. The NHANES provides MET values corresponding to each activity category.²⁵ First, MET-min/week was calculated by multiplying the total number of minutes per week for each activity by the respective MET values. The total MET-min/week for all activities was obtained by summing the values across all activity categories. Based on the recommendations of the US Department of Health and Human Services, participants were categorized into 2 groups: low PA (MET-min/wk < 500) and high PA (MET-min/wk ≥ 500).²⁶ This classification allowed the differentiation of individuals based on their PA levels.

Covariates

The covariates considered in this study included sex (male or female), age, education level (≤ high school vs > high school), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, and other race), household income level (under $20 000 vs over $20 000), alcohol intake (never, ever, and current), smoking status (never, ever, and current), comorbidities (stroke, chronic kidney disease [CKD], chronic obstructive pulmonary disease [COPD], cerebrovascular disease [CVD], diabetes, depression, vitamin D deficiency, hyperlipidemia, hypertension, heart failure, coronary artery disease), and body mass index (BMI) measured in kg/m². Serum 25-hydroxyvitamin D levels were considered because previous studies found them to be associated with constipation.²⁷ Participants with serum 25-hydroxyvitamin D levels ≤ 50 nmol/L (20 ng/mL) were classified as having vitamin D deficiency and were included in the analysis as covariates.²⁸

Statistical Methods

The data in this study were weighted using NHANES-provided weights for the dietary interview sample to appropriately account for the complex survey design of the NHANES.²⁹ Statistical analyses were performed using SAS 9.4 (version 9.4, SAS Institute, Cary, NC, USA) and R Studio software (version 4.2.1). Statistical significance was set at P < 0.05 (two-sided).

Weighted continuous data are presented as mean (SE), and group differences were assessed using ANOVA.³⁰ Categorical data were reported as numbers (n) and weighted percentages (%), and group differences were examined using chi-square tests. Logistic regression analysis was conducted to investigate the association between HEI-2015 and constipation in individuals with different PA levels. The results were reported as odds ratios (OR) with corresponding 95% confidence intervals (95% CI). Three models were constructed: Model 1 was unadjusted; Model 2 was adjusted for age, sex, and race; and Model 3 was further adjusted for education, household income, alcohol consumption, smoking, stroke, CKD, COPD, CVD, diabetes, depression, vitamin D deficiency, hyperlipidemia, hypertension, heart failure, coronary artery disease, and BMI based on Model 2. Linear trends were assessed by considering the median of each quartile as a continuous variable. Restricted cubic splines were used to explore the nonlinear relationship between chronic constipation and HEI-2015.³¹ All covariates in Model 3 were adjusted in the spline regression analysis. To minimize potential sample size reduction resulting from missing covariates, we employed a multiple imputation approach using chained equations with five interpolations. Specifically, we utilized SAS programs SURVEYIMPUTE and MIANALYZE, implementing fully conditional specification methods, to interpolate missing values for the covariates.³² This rigorous methodology was adopted to assess the robustness of our results during further analyses.

Results

Study Participants and Baseline Characteristics

We selected 12 153 adult participants from the NHANES database. After excluding individuals with incomplete or invalid data, including those with missing dietary recall, PA questionnaires, and bowel health questionnaire data, 6427 participants remained. Further exclusions were made for extreme total energy intake, specific medical conditions (ulcerative colitis or Crohn’s disease, celiac disease, and colon cancer), pregnancy, and missing covariates, resulting in a final sample size of 5309 participants (Fig. 1).

Among the study population, 53.76% were male. When constipation was defined based on stool frequency, significant associations (P < 0.05) were observed between constipation and various factors, including sex, BMI, household income, stroke, CVD, depression, hyperlipidemia, coronary artery disease, PA levels, and HEI-2015 scores (Table 1). When constipation was defined based on stool consistency, some associations changed. Specifically, significant associations (P < 0.05) were found between constipation and the HEI-2015 scores, age, BMI, sex, diabetes, and depression (Table 1).

Table 1.

General Characteristics of the Participants in the National Health and Nutrition Examination Survey 2007-2010, Weighted

Characteristic	Definition of constipation
	Constipated (< 3 stools/wk)				Constipated (consistency)
	No	Yes	P-value		No	Yes	P-value
	n = 5149	n = 160			n = 4947	n = 362
HEI-2015 total score	54.61 ± 0.47	49.04 ± 1.49	< 0.001		54.63 ± 0.48	52.07 ± 0.95	0.008
Age (yr)	45.80 ± 0.46	42.68 ± 1.82	0.122		45.94 ± 0.43	42.55 ± 1.28	0.008
BMI (kg/m2)	28.40 ± 0.14	27.03 ± 0.66	0.049		28.44 ± 0.13	27.25 ± 0.46	0.007
Sex			< 0.001				< 0.001
Male	2817 (53.91)	37 (18.51)			2727 (54.05)	127 (37.15)
Female	2332 (46.09)	123 (81.49)			2220 (45.95)	235 (62.85)
Races/ethnicity			0.165				0.109
Non-Hispanic white	2760 (73.66)	80 (67.02)			2666 (73.95)	174 (67.04)
Non-Hispanic black	812 (8.88)	44 (15.52)			778 (8.72)	78 (13.71)
Mexican American	849 (7.78)	18 (7.07)			813 (7.74)	54 (8.06)
Other race	728 (9.68)	18 (10.39)			690 (9.59)	56 (11.19)
Education Level			0.416				0.121
Less than high school	477 (4.63)	11 (3.52)			450 (4.44)	38 (6.65)
High school or more	4672 (95.37)	149 (96.48)			4497 (95.56)	324 (93.35)
Annual family income			0.052				0.582
Under $20 000	1130 (16.12)	52 (23.02)			2651 (54.86)	211 (56.24)
Over $20 000	4019 (83.88)	108 (76.98)			1290 (25.30)	76 (22.27)
Smoke			0.634		1006 (19.84)	75 (21.49)
Never	2771 (54.95)	91(55.15)					0.090
Former	1339 (25.19)	27(21.66)			544 (9.06)	60 (13.52)
Now	1039 (19.86)	42(23.19)			861 (14.47)	65 (16.14)
Alcohol drinker			0.691		3542 (76.46)	237 (70.34)
Never	584 (9.44)	20 (7.34)					0.083
Former	895 (14.53)	31 (16.61)			1083 (16.04)	99 (20.13)
Now	3670 (76.03)	109 (76.05)			3864 (83.96)	263 (79.87)
Baseline comorbidities
Stroke	139 (1.88)	8 (6.56)	0.036		138 (2.07)	9 (1.32)	0.352
COPD	327 (5.92)	10 (6.62)	0.733		317 (6.02)	20 (4.95)	0.556
CKD	713 (10.89)	19 (9.03)	0.497		686 (10.74)	46 (12.01)	0.660
CVD	456 (6.37)	22 (13.02)	0.032		451 (6.75)	27 (4.10)	0.074
Diabetes	546 (7.15)	15 (5.93)	0.505		530 (7.33)	31 (4.24)	0.039
Hypertension	1987 (32.13)	51 (24.34)	0.062		1913 (32.09)	125 (29.29)	0.457
Hyperlipidemia	3819 (71.99)	96 (54.55)	< 0.001		3658 (71.28)	257 (74.13)	0.350
Depression	366 (5.72)	29 (15.25)	< 0.001		350 (5.39)	45 (14.14)	< 0.001
Vitamin D deficiency	3453 (73.29)	94 (69.22)	0.444		3310 (73.23)	237 (72.38)	0.847
Coronary heart disease	187 (2.69)	3 (0.85)	0.071		183 (2.74)	7 (1.31)	0.214
Congestive heart failure	103 (1.28)	6 (6.50)	0.005		102 (1.45)	7 (1.18)	0.682
Physical activity level, MET-min/wk			0.052				0.287
Insufficiently active (< 500)	856 (15.12)	42 (25.32)			832 (15.18)	66 (18.67)
Active (≥ 500)	4293 (84.88)	118 (74.68)			4115 (84.82)	296 (81.33)

HEI, Healthy Eating Index; BMI, body mass index (weight in kilograms divided by height in meters squared); COPD, chronic obstructive pulmonary disease; CKD, Chronic kidney disease; CVD, cardiovascular disease; MET, metabolic equivalent.

Percentages and means (95% CI) were estimated using United States population weights.

Race/ethnicity was determined using the preferred terminology from the National Center for Health Statistics: non-Hispanic white, non-Hispanic black, or Mexican American. Mexican American individuals were oversampled rather than broader groups of individuals from Latin America. Others included Asians, other Hispanics, native Alaskans, and multiracial individuals.

P-values were computed separately for each covariate and indicated statistically significant differences between step groups at P < 0.05.

Data are presented as mean ± SE or n (%).

Association Between Healthy Eating Index-2015 Scores and Constipation at Different Physical Activity Levels

To examine the association between HEI-2015 scores and constipation at different PA levels, logistic regression analysis was performed after adjusting for various covariates. Table 2 presents the results. Using the stool consistency definition of constipation, no significant association was found between HEI-2015 scores and constipation in the population with low PA levels (OR, 1.01; 95% CI, 0.97-1.05). However, in the population with high PA levels, there was a significant association between HEI-2015 scores and reduced constipation risk. Each 1-point increase in the HEI-2015 score was associated with a 2% reduction in the odds of constipation (OR, 0.98; 95% CI, 0.97-0.99). Additionally, in Model 3, the highest tertile of HEI-2015 scores showed a significantly lower risk of constipation compared with the lowest tertile (OR, 0.55; 95% CI, 0.35-0.87; P for trend = 0.013). Similar findings were observed when constipation was defined based on stool frequency. The low-PA level population showed no significant association between HEI-2015 scores and constipation (OR, 0.96; 95% CI, 0.90-1.03). However, in the high PA level population, higher HEI-2015 scores were significantly associated with a reduced constipation risk. Each 1-point increase in the HEI-2015 score was associated with a 4% decrease in the odds of constipation (OR, 0.96; 95% CI, 0.94-0.98). In Model 3, the highest tertile of HEI scores had a significantly lower risk of constipation compared to the lowest tertile (OR, 0.30; 95% CI, 0.15-0.63; P for trend < 0.01). Furthermore, spline curves were used to illustrate the dose-response relationship between HEI-2015 and constipation risk at different PA levels. The curves showed a linear relationship between HEI-2015 scores and the risk of constipation in all PA groups. Nonlinearity tests indicated no significant deviations from linearity (P > 0.05) for either stool consistency or stool frequency definitions of constipation (Fig. 2).

Table 2.

Effect of Physical Activity on the Relationship Between Dietary Fiber Intake and Constipation Using Different Definitions of Constipation

Physical activity level, MET-min/wk	Healthy Eating Index
	Low (< 48)	Middle (48-59)	High (≥ 59)	P for trend	HEI score (continued)	P-value
Constipated (consistency)
Insufficiently active (< 500)
Model 1	1 (ref)	1.31 (0.53, 3.23)	0.97 (0.36, 2.61)	0.976	1.00 (0.97, 1.03)	0.934
Model 2	1 (ref)	1.47 (0.61, 3.5)	1.06 (0.4, 2.82)	0.855	1.00 (0.97, 1.03)	0.926
Model 3	1 (ref)	1.52 (0.49, 4.71)	1.44 (0.44, 4.74)	0.454	1.01 (0.97, 1.05)	0.505
Active (≥ 500)
Model 1	1 (ref)	0.82 (0.61, 1.11)	0.59 (0.4, 0.86)	0.006	0.98 (0.97, 0.99)	0.006
Model 2	1 (ref)	0.82 (0.60, 1.13)	0.58 (0.4, 0.84)	0.003	0.98 (0.97, 0.99)	0.002
Model 3	1 (ref)	0.81 (0.56, 1.17)	0.55 (0.35, 0.87)	0.013	0.98 (0.97, 0.99)	0.016
Constipated (< 3 stools/wk)
Insufficiently active (< 500)
Model 1	1 (ref)	0.61 (0.2, 1.85)	0.6 (0.12, 2.91)	0.512	0.97 (0.91, 1.03)	0.298
Model 2	1 (ref)	0.61 (0.2, 1.86)	0.62 (0.12, 3.18)	0.543	0.97 (0.91, 1.03)	0.322
Model 3	1 (ref)	0.72 (0.18, 2.81)	0.54 (0.08, 3.61)	0.463	0.96 (0.9, 1.03)	0.254
Active (≥ 500)
Model 1	1 (ref)	0.71 (0.37, 1.34)	0.35 (0.19, 0.65)	0.001	0.97 (0.95, 0.98)	< 0.001
Model 2	1 (ref)	0.68 (0.34, 1.36)	0.30 (0.16, 0.56)	0.001	0.96 (0.95, 0.98)	< 0.001
Model 3	1 (ref)	0.63 (0.31, 1.28)	0.30 (0.15, 0.63)	0.004	0.96 (0.94, 0.98)	0.001

MET, metabolic equivalent.

Model 1: non-adjusted; Model 2: adjusted for age, sex, race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, and other race); Model 3: adjusted for Model 2 plus education level (≤ high school vs > high school), household income level (< $20 000 vs over $20 000), alcohol intake (never, ever, and current), smoking status (never, ever, and current), comorbidities (stroke, chronic kidney disease, chronic obstructive pulmonary disease, cardiovascular disease, diabetes, depression, vitamin D deficiency, hyperlipidemia, hypertension, heart failure, and coronary artery disease), and body mass index.

Data are presented as OR (95% CI).

Figure 2.

Restricted cubic spline of Healthy Eating Index (HEI) and odds ratio of constipation at different physical activity levels. (A) Association between HEI-2015 scores and constipation at different physical activity levels: constipation defined as < 3 stools/wk. (B) Association between HEI-2015 scores and constipation at different physical activity levels: constipation defined by consistency. Note: adjusted for age, sex, education level (≤ high school vs > high school), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, other race), household income level (< $20 000 vs over $20 000), alcohol intake (never, ever, and current), smoking status (never, ever, and current), and comorbidities (stroke, chronic kidney disease, chronic obstructive pulmonary disease, cardiovascular disease, diabetes, depression, vitamin D deficiency, hyperlipidemia, hypertension, heart failure, and coronary artery disease), and body mass index.

Sensitivity Analyses

These results remained consistent after interpolation (Table 3). In the low PA-level group, there was no significant correlation between HEI-2015 scores and stool consistency (OR, 1.00; 95% CI, 0.97-1.03; P = 0.940) or stool frequency (OR, 0.97; 95% CI, 0.91-1.02; P = 0.230). Conversely, in the high PA-level group, there was a significant correlation between HEI-2015 scores and both stool consistency (OR, 0.98; 95% CI, 0.97-0.99; P = 0.002) and stool frequency (OR, 0.96; 95% CI, 0.95-0.98; P < 0.01).

Table 3.

Sensitivity Analysis: Hazard Ratio (95% CI) for Healthy Eating Index and Constipation (Different Definitions) after Utilizing Survey Impute for Multiple Interpolation (N = 263) Across Various Levels of Physical Activity

	Insufficiently active (< 500)	P-value	Active (≥ 500)	P-value
	OR (95% CI)		OR (95% CI)
Constipated (consistency)
Model 1	0.99 (0.97, 1.02)	0.702	0.98 (0.97, 0.99)	0.002
Model 2	1.00 (0.97, 1.03)	0.943	0.98 (0.97, 0.99)	0.002
Constipated (< 3 stools/wk)
Model 1	0.97 (0.91, 1.03)	0.272	0.97 (0.95, 0.98)	< 0.001
Model 2	0.97 (0.91, 1.02)	0.234	0.96 (0.95, 0.98)	< 0.001

Model 1: non-adjusted; Model 2: adjusted for age, sex, education level (≤ high school vs > high school), race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American, other race), household income level (< $20 000 vs over $20 000), alcohol intake (never, ever, and current), smoking status (never, ever, and current), comorbidities (stroke, chronic kidney disease, chronic obstructive pulmonary disease, cardiovascular disease, diabetes, depression, vitamin D deficiency, hyperlipidemia, hypertension, heart failure, and coronary artery disease), and body mass index.

Discussion

In this population-based study utilizing the NHANES nutritional data, we examined the relationship between PA, overall diet quality (as measured by the HEI-2015 score), and constipation. We used 2 different definitions of constipation: stool consistency and stool frequency. After accounting for potential confounding factors, we observed inconsistent associations between adherence to the current US dietary guidelines (HEI-2015 score) and constipation across different levels of PA. Specifically, we found no significant association between the HEI-2015 score and constipation in the low-PA group. However, in the high-PA group, there was a noteworthy inverse association between HEI-2015 score and constipation. This association was consistent regardless of the definition of constipation. These findings suggest that the relationship between diet quality and constipation is influenced by PA.

PA and diet are lifestyle factors that can be modified to improve health outcomes.^33,34 Engaging in appropriate PA has been demonstrated to have a positive impact on bowel movements, including increasing their frequency and preventing or alleviating constipation.³⁵ Numerous studies have reported a potential association between PA and a reduced risk of constipation.^36,37 However, there are also differing viewpoints based on previous NHANES studies, which have indicated a weak correlation between insufficient PA and stool consistency or having bowel movements less than three times per week.¹⁸ The effects of PA on constipation may vary depending on factors, such as exercise intensity, duration, and mode. Most studies have focused primarily on the benefits of recreational PA, with limited research on non-recreational PA.³⁸ Currently, there is limited evidence to support the notion that different types of PA yield equivalent health benefits, and no definitive conclusions have been reached regarding the optimal PA intensity for each type. Therefore, this study aimed to strengthen the validity of these findings by assessing PA intensity by integrating various types of PA.

Dietary factors are crucial in the prevention and management of constipation.^39,40 Several components of the HEI-2015 have demonstrated associations with constipation. Fruits and vegetables, rich in soluble fiber, possess a strong affinity for water, facilitating defecation by softening stools, increasing stool frequency, and improving consistency.⁴¹ Oligosaccharides in beans stimulate Bifidobacteria growth and act as a natural laxative.⁴² Whole grains have been linked to lower constipation scores.⁴³ Coarse grains, abundant in dietary fiber, increase fecal bulk and expedite intestinal/colonic transit time. Significantly, a negative correlation has been identified between higher fatty acid ratios (polyunsaturated fatty acids [PUFA]/saturated fatty acids) and constipation.⁴³ Saturated fatty acids may disrupt the composition of intestinal flora, potentially leading to intestinal dysbiosis, while PUFA, such as omega-3 or long-chain PUFA, regulate inflammation and the immune system, fostering a healthy symbiotic relationship among intestinal bacteria.^43,44 Additionally, fruits, vegetables, and whole grains contribute to increased water intake and the maintenance of water balance. Protein aids in the absorption and retention of water, and higher water intake is associated with a reduced risk of constipation.¹⁷ Conversely, sodium intake was found to be positively associated with the likelihood of experiencing constipation.¹³ This association may be attributed to increased salt intake reducing fecal water content, making it challenging for feces to move through the digestive tract and leading to difficulties in defecation. It is essential to note that individuals typically consume a combination of these dietary components rather than a single nutrient in isolation. This complexity makes it challenging to isolate the effects of specific nutrients due to the interactions among these dietary elements. By utilizing the HEI-2015 to evaluate overall dietary quality, we overcame this challenge and closely examined the relationship between dietary quality and constipation, aligning more closely with real-world dietary practices.

HEI-2015, based on evidence-based recommendations from the US Department of Agriculture and Department of Health and Human Services, provides a standardized assessment of an individual’s overall dietary quality.⁴⁵ Previous research demonstrated that higher adherence to the HEI-2015 is associated with a reduced risk of constipation. However, it is important to note that increased dietary fiber intake and moderate PA decrease the prevalence of constipation.⁴⁶ The beneficial effects of dietary fiber on constipation may be influenced by population-level PA, which aligns with the objectives of our study and provides a novel perspective for investigating the impact of dietary quality on constipation.^47,46 Although PA and diet have been individually linked to a reduced risk of constipation, few studies have examined their combined effects. Our study revealed that, among US adults, a higher overall diet quality was significantly associated with a decreased risk of constipation in individuals with high PA levels. Furthermore, we observed similar associations when different definitions of constipation were used. Additional analyses confirmed the robustness of our findings.

Previous studies have suggested that a combination of regular PA, increased fiber intake, and adequate fluid intake can prevent and relieve constipation.⁴⁸ In a study conducted within the NHANES framework, Li et al²⁰ also identified a synergistic effect between dietary fiber and PA in improving constipation, particularly when considering stool consistency. However, unlike our study, this association disappeared when constipation was defined based on stool frequency. Several factors may contribute to this observed discrepancy. Firstly, it is possible that dietary fiber, as a subcomponent of the HEI-2015, may not adequately represent the overall diet quality. It may also fail to fully capture the cumulative effects of other components included in the HEI-2015, such as fruits, vegetables, whole grains, and fats, regarding constipation improvement. Secondly, differences in how PA was defined and measured in Li et al’s study,²⁰ including variations in the PA questionnaire used before and after 2006 in the NHANES study, may have contributed to the differences in findings between the 2 studies. PA stimulates colonic contractions and increases the colonic transit time, whereas high dietary quality promotes bowel regularity and improves stool consistency. In conclusion, the relationship between exercise, diet, and constipation is complex, and further investigation is warranted to gain a deeper understanding of the combined effects of these lifestyle factors. It is essential to consider various exercise and dietary patterns in future studies to elucidate their interplay and impact on constipation.

The strengths of our study include its nationally representative design, inclusion of multiple types of PA (both work-related and leisure activities), and comprehensive and detailed dietary data. However, this study had several limitations. First, due to its cross-sectional nature, we could not establish a causal relationship. Future prospective observational and interventional studies are needed to elucidate the effects of different PA and dietary patterns on various forms of constipation. Second, NHANES data were collected through interviews and self-reported questionnaires, which may introduce inaccuracies and recall biases. Third, the 24-hour recall method inherently incorporates inaccuracies and may not accurately represent actual intakes, making it susceptible to recall bias and the potential for both underreporting and overreporting.

In conclusion, our study contributes to the growing body of evidence regarding the complex interplay between PA, diet quality, and constipation. We found that higher overall diet quality was associated with a reduced risk of constipation among individuals with high PA levels. These findings suggest the potential synergistic effect of PA and diet quality on bowel function. However, further research is required to investigate the combined effects of exercise and dietary patterns on constipation in diverse populations. Prospective studies and interventions targeting lifestyle modifications are warranted to ascertain causality and optimal strategies for preventing and managing constipation.