Original Contribution

Functional GI Disorders

Characterizing Normal Bowel Frequency and Consistency in a Representative Sample of Adults in the United States (NHANES)

  • The American Journal of Gastroenterology 113, 115123 (2018)
  • doi:10.1038/ajg.2017.213
  • Download Citation
Received:
Accepted:
Published online:

Abstract

Objectives:

Our current understanding of normal bowel patterns in the United States (US) is limited. Available studies have included individuals with both normal and abnormal bowel patterns, making it difficult to characterize normal bowel patterns in the US. The current study aims to (1) examine frequency and consistency in individuals with self-reported normal bowel habits and (2) determine demographic factors associated with self-reported normalcy.

Methods:

This study used data from adult participants who completed bowel health questions as part of the National Health and Nutrition Examination Survey (NHANES) in 2009–2010 and who reported normal bowel patterns (N=4,775). Data regarding self-perceived bowel health; stool frequency; stool consistency (using the Bristol Stool Form Scale (BSFS)); and demographic factors were analyzed.

Results:

95.9% of the sample reported between 3 and 21 BMs per week. Among men, 90% reported a BSFS between 3 and 5, while for women it was 2–6. After controlling for age, the following demographic variables were associated with normalcy: male sex, higher education, higher income, <2 daily medications, and high daily fiber intake. Hispanic ethnicity was significantly associated with abnormal self-reported bowel habits.

Conclusions:

This is the first study to evaluate normal bowel frequency and consistency in a representative sample of adults in the US. The current findings bolster the common “3 and 3” metric of normal frequency (3 BMs/day to 3 BMs/week) while also suggesting different criteria for normal consistency for men and women. Finally, this study provides novel information about demographic factors associated with normal frequency and consistency.

Introduction

Our current understanding of what constitutes normal bowel patterns in the United States (US) population is limited. The importance of understanding normal bowel habits is relevant not only for patient and physician expectations but also for clinical trial eligibility criteria and outcome measures. Most published studies evaluating bowel patterns are limited by small sample sizes and have not adequately assessed sex or demographic differences in normal bowel patterns (1, 2, 3). In addition, most available studies have included individuals with both normal and abnormal bowel patterns, making it difficult to estimate what constitutes normal bowel patterns in the US. Since estimated prevalence of functional gastrointestinal disorders such as constipation and diarrhea range up to 20 and 17%, respectively (4, 5, 6, 7), it is essential to evaluate a sample of individuals without reported constipation or diarrhea in order to understand normal bowel function in the general population.

Bowel frequency is the most commonly used metric of bowel habits in both clinical and research settings. One of the most widely cited studies to assess bowel frequency in the population was conducted on factory workers and general medical patients in the greater London area in the 1960s (1). That study found 99% of individuals reported bowel frequency within the range of 3 bowel movements (BMs) per week and 3 BMs per day, regardless of their self-perceived bowel function. This range has since been confirmed by several studies but has not been tested in a representative US sample. For example, in a study conducted at the University of North Carolina in the early 1980s on college students and hospital employees, 94.2% of individuals reported stool frequency within this range (though over 17% of these individuals reported having bowel dysfunction) (2); and in one of the few studies conducted among individuals with self-reported normal bowel habits involving 204 individuals in Italy, 100% of study participants reported bowel frequency within the range of 3 BMs per week and 3 BMs/week (3).

Another common measure used to assess bowel patterns is stool consistency, which is commonly measured by the Bristol Stool Form Scale (BSFS) that varies from 1 (hard lumps) to 7 (watery) (8). The BSFS was initially believed to be a guide to assess intestinal transit (9) though subsequent studies have questioned its ability except at the extreme ends of the scale (10, 11). Although a BSFS between 3 and 5 has generally been considered normal (12, 13), one study of 146 women without symptoms of functional bowel disorders in Olmstead County, Minnesota suggested that the normal range on the BSFS should be between 2 and 4 (ref. 14).

Given the need to better understand what defines normal bowel habits in a representative sample of the US population without self-reported diarrhea or constipation, the aims of this study were (i) to examine the range of bowel habits using measures of frequency and consistency in individuals with self-reported normal bowel function and (ii) to determine demographic and environmental factors associated with self-reported normal bowel function.

Methods

Study cohort

The National Health and Nutrition Examination Survey (NHANES) is a survey research program that is designed to analyze a nationally representative sample of non-institutionalized respondents in the United States. NHANES is conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control (Atlanta, GA, USA). All participants provide written informed consent before completing the NHANES and there are no patient identifiers in the publicly available NHANES database. Participants are selected using a stratified multistage probability design with oversampling of certain age and ethnic groups. Sample weights in NHANES allow inferences to the non-institutionalized population of the United States.

Data from adult participants (age ≥20 years) in NHANES 2009–2010, who completed two specific bowel health questions about constipation and diarrhea were used for this study (N=5,164). The questions in the NHANES questionnaire were: “During the past 12 months, how often have you been constipated?” and “During the past 12 months, how often have you had diarrhea?” These bowel health questionnaires were completed in the Mobile Examination Center (MEC) Interview Room using a Computer-Assisted Personal Interview (CAPI) System. Any missing data (N=782) and any participants, who selected “refused”/“do not know” (N=10), were excluded from the study. Those who did not report having diarrhea or constipation ‘always’ or ‘most of the time’ comprised the population with self-reported normal bowel habits.

Bowel health questionnaire

Other bowel health questionnaires from the NHANES 2009–2010 were then used to analyze bowel symptoms in participants with self-reported normal bowel habits. Stool consistency was assessed using the BSFS (BSFS: range from type 1–7). Participants were asked to define their stool by recording the number type that corresponded to their usual/most common stool type. Stool frequency was also assessed with the following questions, “How many times do you have a BM per week?” The range varied from 1 to 63 BMs per week. BMs per week was then recorded into five sub-groups: <3, ≥3–7, ≥8–14, ≥15–21, ≥21 BMs per week. These categories were determined based on logical cutpoints (≥3–7=up to one BM per day; ≥8–14=one to two BMs per day; ≥15–21=two to three BMs per day).

Co-variables

The following co-variables were included to evaluate demographic factors potentially associated with normality: age, sex, race/ethnicity, education, poverty income ratio, body mass index, self-reported medical comorbidities, smoking behavior, number of daily medications, physical activity, and dietary intake. Age was divided into decades (20–29, 30–39, 40–49, 50–59, 60–69, and ≥70 years). Race and ethnicity was recorded into the following classifications: Non-Hispanic White, Non-Hispanic Black, Hispanic (including Mexican American), and other race (including multi-racial). Education was divided into three levels: less than high school, high school, and more than high school level. Poverty income ratio was categorized into two groups: ≤2 times the poverty threshold and >2 times the poverty threshold. Three body mass index groups were assessed: normal weight (<25.0 kg/m2), overweight (25.0 to 29.9 kg/m2), and obese (≥30 kg/m2). Number of daily medications (including over-the-counter drugs) was separated into two variables: (i) <2 and (ii) ≥2 medications. Participants with medical comorbidities, including hypertension, diabetes, coronary heart disease, stroke, cancer, thyroid problem, gout, and celiac disease, were defined by self-reported diagnosis. Smoking behavior was split into three categories: non-smoker, former smoker, and current smoker. Vigorous physical activities were defined as “hard effort that cause[d] large increases in breathing or heart rate for at least 10 min continuously.” Dietary fiber, liquid, and caffeine intake were obtained from the first day of the 24 h dietary recall from the NHANES 2009–2010.

Statistical analyses

We first categorized individuals based on their self-described frequency of constipation or diarrhea, and examined the correlation to reported BM frequency as well as consistency. We then identified individuals with self-perceived normal bowel habits, and summarized their background characteristics in both male and female groups. We also examined self-reported stool frequency and stool consistency characterized by the BSFS type in this cohort, and compared them between males and females. To identify factors associated with normal bowel pattern, we performed logistic regression while adjusting for age to examine the associations with ethnicity, socioeconomic factors, medical history, dietary fiber/liquid intake, smoking, and caffeine intake. P-value has been adjusted for Bonferroni correction (P<0.002 is considered significant). We also examined these factors in association with normal BMs specifically in men and women and tested the interaction with sex.

All estimated standard errors were calculated using sampling weights accounting for the complex survey design of NHANES. Taylor series linearization was used for variance estimate (15). All statistical analyses were performed using STATA statistical software version 8.2 (College Station, TX, USA).

Results

A total of 5,164 adults participating in the NHANES 2009–2010 survey were included in initial analyses to establish a sample of individuals with self-described normal bowel habits. The distribution of stool frequency and stool consistency of these individuals based on self-reported frequency of constipation and diarrhea in the past 12 months is shown in Table 1. Stool frequency and consistency in individuals who reported having diarrhea or constipation ‘sometimes’ within the past 12 months were more similar to those who reported having diarrhea or constipation ‘rarely/never’ compared to those who reported having these symptoms ‘most of time/always’. Therefore, individuals who reported having diarrhea and constipation either ‘rarely/never’ or ‘sometimes’ in the past 12 months were included in the self-reported normal habit group. In total, this group included 4,775 (92%) of the 5,164 individuals participating in the NHANES 2009–2010 survey (Table 2).

Table 1: Distribution of bowel health questionnaire and their association with self-reported constipation and diarrhea in subjects
Table 2: Factors associated with self-reported normal bowel condition among subjects from NHANES 2009–2010Continued

Among individuals with self-reported normal bowel habits 52% of this sample was male, 50% was White, and 50% had achieved at least a high school education. Each age group (in decades) was well-represented from age 20–70+. Demographics for the entire sample as well as demographics divided by sex are shown in Table 2.

Stool frequency and consistency in individuals with self-reported normal bowel function

Among individuals with self-reported normal bowel habits 95.9% reported between 3 and 21 BMs per week (Figure 1). Individuals (60.5%) reported having between 3–7 BMs per week, 29.9% reported between 8 and 14 BMs per week, and 5.7% reported between 15 and 21 BMs per week. In contrast only 2.7% reported <3 BMs per week and 1.4% reported >21 BMs per week. The distribution of stool frequency based on percentile of individuals was as follows: 5th percentile=3, 10th percentile=4, 50th percentile=7, 90th percentile=14, and 95th percentile=21.

Figure 1
Figure 1

Histogram of bowel movements distribution in self-reported normal bowel habits from the 2009 to 2010 National Health and Nutrition Examination Survey (NHANES).

Among individuals with self-reported normal bowel habits, 86.2% reported BSFS type between 3 and 5 (Figures 2 and 3). A BSFS type 1 was reported by 4.5% while a BSFS type 2 was reported by 4.7%. Likewise, a BSFS type 6 was reported by 6.9% while a BSFS type 7 was reported by 0.7%. The distribution of BSFS based on percentile of individuals was as follows: 5th percentile=2, 10th percentile=3, 50th percentile=4, 90th percentile=5, and 95th percentile=6. For both women and men, the percent reporting self-perceived normal bowel habits was negatively related to age (Figure 4). Overall, men more frequently reported normal bowel habits than women. In general, the 5th and 95th percentile for stool frequency in both men and women fell within the 3 BMs per day and 3 BMs per week. However, for stool consistency the 5th and 95th percentile for men fell between 3 and 5, while for women it was between 2 and 6.

Figure 2
Figure 2

Histogram of Bristol stool type distribution in the self-reported normal bowel habits from the 2009 to 2010 National Health and Nutrition Examination Survey (NHANES).

Figure 3
Figure 3

Histogram of Bristol stool type distribution in the self-reported normal bowel habits in men and women from the 2009 to 2010 National Health and Nutrition Examination Survey (NHANES).

Figure 4
Figure 4

Self-reported normal bowel habits (N=4,775) in the adult participants (N=5,164) who completed the questions on diarrhea and constipation in the 2009–2010 National Health and Nutrition Examination Survey (NHANES).

Demographic and environmental factors associated with self-perceived normal bowel function

After controlling for age in decades, prevalence odds ratios (POR) revealed that the following demographic variables were significantly associated with self-reported normal bowel habits (Table 3): male sex, higher education, higher poverty income ratio, and lower number of daily medications. In the entire sample (men and women combined), the only lifestyle-related demographic factor that was significantly associated with self-reported normal bowel habits was high daily fiber intake (>20.1 gm/day). No statistically significant differences were found in prevalence of self-reported normal bowel habits in the US adult population according to physical activity, body mass index, smoking behavior, or high caffeine intake. On the other hand, those with self-reported history of thyroid problems and those who reported Hispanic ethnicity had a lower likelihood for self-reported normal bowel habits. Finally, participants within the standard ranges of normal bowel frequency (≥3–21 per week) and normal common stool types (3–5 BSFS) were 8.5- and 4.9-fold more likely to have self-reported normal bowel habits (reporting having diarrhea and/or constipation “sometimes” or “rarely/never”, respectively).

Table 3: Weighted multivariable models for factors associated with normal bowel habits among US population from NHANES, 2009–2010

Table 3 also outlines PORs for men and women separately. In these analyses, physical activity produced the only significant difference between men and women when evaluating normal BMs. In men, but not in women, vigorous physical activity was associated with normalcy (P=0.02).

Discussion

This is the first study to evaluate and define self-perceived normal bowel frequency and consistency in a large representative sample of adults in the general United States population. In the present study, 90% of the sample reported 3–21 BMs per week. In addition, 90% of individuals reported their most common stool type between 2 and 6, although this varied by sex. Among men, 90% reported a BSFS between 3 and 5, while for women it was between 2 and 6. These findings contribute to existing literature by bolstering the common “3 and 3” metric of normal bowel frequency (i.e., 3 BMs per day to 3 BMs per week), however, they also suggest that men and women should have different criteria for normal bowel consistency.

Although previous studies have found similar results with regard to bowel frequency, none included samples that were representative of the general US population. For example, the first study to describe a typical range of normal BMs used a sample of factory workers and general medical patients in London (1). Subsequent studies evaluated similarly non-generalizable samples in Italy, the United Kingdom, Singapore, Midwestern United States, and North Carolina (2, 3, 14, 16). Furthermore, the majority of the above-mentioned studies did not evaluate self-perceived normalcy of bowel patterns and did not reliably characterize normal bowel consistency.

Defining a population-based range of normal BM frequency and consistency among individuals in the US with self-reported normal bowel patterns is important for both research and clinical practice. In research settings, this data may be used to establish inclusion and exclusion criteria as well as to determine thresholds for improvement in symptoms. In clinical practice, this data may be used in the context of evidence-based patient education and clinical decision making. Based on our results, we recommend maintaining the common metric of 3BMs/week to 3BMs/day to define normal frequency and to use a range of 3–5 for men and 2–6 for women on the BSFS when defining normal consistency. Including a BSFS type 2 as within normal for women would be consistent with the study by Bharucha et al., which included 146 women without symptoms of a functional bowel disorder from the Mayo Clinic. Interestingly in Bharucha’s study, the upper 90th percentile was a BSFS type 3.5 compared with 6 in our study. The reason for this discrepancy is likely due to the more homogeneous population and a smaller sample size.

In addition to confirming metrics for normalcy, this study also provides novel information about demographic factors associated with normalcy when considering frequency and consistency of BMs in the general US population. Among the strongest factors associated with normal BMs in this sample were: living above poverty level; higher education; high fiber intake; and fewer medications. Although not significant in the whole sample, there was also a significant interaction between men and women when evaluating the effects of physical activity on bowel patterns, with more vigorous physical activity associated with more normalcy in men but not women. Prior studies have shown that increased fiber intake may benefit patients with functional bowel disorders (17, 18) and that increased physical activity can improve gas transit and reduce bloating among healthy individuals (19). However, there are several studies have that have reported conflicting findings on the effect of fiber and physical activity in constipated patients (18, 20, 21) and this area likely warrants further research.

Although it is beyond the scope of this paper to explain discrepancies in normal bowel patterns based on education, ethnicity, and/or income, it is possible that the dietary factors mentioned above contribute to the demographic predictors of normalcy found in this study. Previous studies have found differences in dietary fiber and vegetable intake based on income, race, and ethnicity, which may explain our findings regarding these variables (19).

Finally, our study provides interesting data regarding changes in aging when considering normal frequency and consistency of stool. Our data indicate that, over the adult lifespan, men and women are less likely to report normal bowel patterns as they age. Overall, men appear more likely to report normal patterns when compared to women. Interestingly, women’s self-reported normalcy of bowel patterns in this sample appears to decline until menopause and then plateau, while men’s self-reported normalcy of bowel patterns appears to decline with age (Figure 4). These findings are consistent with other literature suggesting normal, age-related changes in motility (20, 21)

An important consideration when interpreting this data lies in the fact that 83.6% of patients who reported diarrhea and 74.7% of those who reported constipation fell within the normal range of bowel frequency (3–21 per week) and 36% of those who reported diarrhea and 67% of those who constipation fell within the normal range for bowel consistency (3–5 on the BSFS), see Table 1. This overlap is notable as it reflects differences in individual interpretation of normalcy. These two groups also reflect differences in likelihood to seek treatment for constipation/diarrhea and may indicate some amount of hypervigilance or hypersensitivity to bowel habits in the group who reported diarrhea or constipation always or most of the time. These findings can be used clinically in order to normalize frequency and/or consistency in patients presenting for treatment of constipation or diarrhea.

There were several limitations to this study. First, this study relied entirely on retrospective and self-report data. Second, this study only evaluated frequency and consistency of BMs and did not assess other gastrointestinal symptoms. Third, we do not have data available on laxative or anti-diarrheal medication use, which may limit our ability to interpret the data. These limitations were due to the nature of the data available through the NHANES database. Future studies should continue to evaluate and characterize normalcy in bowel patterns with the goal to elucidate demographic and lifestyle factors that might contribute to normal bowel patterns in the US population.

Study Highlights

References

  1. 1.

    , , et al. Variation of bowel habit in two population samples. Proc R Soc Med 1966;59:11–12.

  2. 2.

    , , et al. Bowel patterns among subjects not seeking health care. Use of a questionnaire to identify a population with bowel dysfunction. Gastroenterology 1982;83:529–534.

  3. 3.

    , , et al. Self-perceived normality in defecation habits. Dig Liver Dis 2005;38:103–108.

  4. 4.

    , , et al. Bowel Disorders. Gastroenterology 150:1393–1407.e5.

  5. 5.

    , , et al. Cumulative incidence of chronic constipation: a population-based study 1988-2003. Aliment Pharmacol Ther 2007;26:1521–1528.

  6. 6.

    , , et al. Functional bowel disorders in Iranian population using Rome III criteria. Saudi J Gastroenterol 2010;16:154–160.

  7. 7.

    , , et al. Epidemiology of functional diarrhea and comparison with diarrhea-predominant irritable bowel syndrome: a population-based survey in China. PLoS ONE 2012;7:e43749.

  8. 8.

    , , . Detection of pseudodiarrhoea by simple clinical assessment of intestinal transit rate. BMJ 1990;300:439–440.

  9. 9.

    , . Stool form scale as a useful guide to intestinal transit time. Scand J Gastroenterol 1997;32:920–924.

  10. 10.

    , , et al. Stool characteristics and colonic transit in irritable bowel syndrome: evaluation at two time points. Scand J Gastroenterol 2013;48:295–301.

  11. 11.

    , , et al. Do stool form and frequency correlate with whole-gut and colonic transit? results from a multicenter study in constipated individuals and healthy controls. Am J Gastroenterol 2010;105:403–411.

  12. 12.

    , , . Validity and reliability of the Bristol Stool Form Scale in healthy adults and patients with diarrhoea-predominant irritable bowel syndrome. Aliment Pharmacol Ther 2016;44:693–703.

  13. 13.

    , , et al. Association of low dietary intake of fiber and liquids with constipation: evidence from the National Health and Nutrition Examination Survey. Am J Gastroenterol 2013;108:796–803.

  14. 14.

    , , et al. Insights into normal and disordered bowel habits from bowel diaries. Am J Gastroenterol 2008;103:692–698.

  15. 15.

    , . Analysis of Health Surveys. Wiley InterscienceBiometric Research Branch at the National Cancer Institute: Bethesda, MD. 1999.

  16. 16.

    , , et al. Defecation frequency and timing, and stool form in the general population: a prospective study. Gut 1992;33:818–824.

  17. 17.

    , , et al. The effect of fiber supplementation on irritable bowel syndrome: a systematic review and meta-analysis. Am J Gastroenterol 2014;109:1367–1374.

  18. 18.

    , , . Association between physical activity, fiber intake, and other lifestyle variables and constipation in a study of women. Am J Gastroenterol 2003;98:1790–1796.

  19. 19.

    , . Income and race/ethnicity influence dietary fiber intake and vegetable consumption. Nutr Res 2014;34:844–850.

  20. 20.

    . Basic mechanisms of the aging gastrointestinal tract. Dig Dis 2007;25:112–117.

  21. 21.

    . The aging gut: physiology. Clin Geriatr Med 2007;23:757–767 v-vi.

Download references

Author information

Affiliations

  1. Division of Gastroenterology, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, USA

    • Shuji Mitsuhashi
    • , Sarah Ballou
    • , Zhenghui G Jiang
    • , William Hirsch
    • , Judy Nee
    • , Johanna Iturrino
    • , Vivian Cheng
    •  & Anthony Lembo

Authors

  1. Search for Shuji Mitsuhashi in:

  2. Search for Sarah Ballou in:

  3. Search for Zhenghui G Jiang in:

  4. Search for William Hirsch in:

  5. Search for Judy Nee in:

  6. Search for Johanna Iturrino in:

  7. Search for Vivian Cheng in:

  8. Search for Anthony Lembo in:

Competing interests

Guarantor of the article: Anthony Lembo, MD.

Specific author contributions: The idea for the article was conceived by A.L and Z.G.J. The manuscript was drafted by S.B. and S.M. Statistical analysis was performed by S.M. and Z.G.J. The draft manuscript was critically reviewed by W.H., J.N., J.I, and V.C. All authors reviewed and approved the final version of the manuscript.

Financial support: This project was funded in part by T32DK007760.

Potential competing interests: None.

Corresponding author

Correspondence to Anthony Lembo.