This study using CT colonography shows that females and constipated (less frequent defecation) patients have a significantly longer colon, and that these trends also are shown in analyses limited to the length of the proximal colon. However, colon length is not associated with physique, regardless of gender. These observations are confirmed by two different statistical methods. To our knowledge, this is the first study to show that bowel habits (defecation frequency) clearly correlate with colon length in a cohort study. A previous case–control study (n = 51) utilizing CT colonography reported that patients with functional constipation had a significantly longer colon compared with diarrhea type IBS or control patients [22]. These results suggest that defecation frequency is associated with colon length and are consistent with the results of the present study.
The difference in colon length between “daily” and “constipated” groups was an average of 11 cm and the difference between the “daily” and “intermediate” groups was an average of 7 cm. It is important to consider how such seemingly small difference could affect defecation frequency. A plausible assumption is that constipated bowel habits could lengthen the colon as a result of fecal impaction. According to a recent murine physiology study, however, elongation of colon by longitudinally stretching the colonic muscle has the effect of inhibiting the colonic migrating motor complexes which are responsible for fecal pellet propulsion in the murine large bowel, finally leading to colonic accommodation and slow transit [25]. However, a partially obstructed outlet in a murine colon resulted in an elongated impacted colon with slowed transit [26]. A similar mechanism may be active in the human colon.
In the present study, proximal colon elongation was present while distal colon elongation was not present in patients classified as constipated. In a nuclear scintigraphy study in children, although unable to quantify colon length, transverse colon elongation is more common but sigmoid colon elongation is not present in those with slow-transit constipation [27]. Even in adults, slow transit constipation could influence proximal colon elongation. To diagnose slow transit constipation, objective tests, e.g., daily stool weight (< 35 g/d), colonic transit time, and anorectal function are required, and diagnostic evaluations should be performed while the patient is not taking laxatives [28]. These objective tests may be onerous for patients to undergo. In addition to the Bristol Stool Form Scale which can estimate colonic transit time [29], proximal colon length measured by CT colonography might be a new surrogate biomarker to identify slow transit constipation.
Notably, the present study has clearly demonstrated that the female colon is longer than the male colon, consistent with two previous relatively large-scale cohort studies; one study applied CT colonography to asymptomatic American patients (n = 505) [21] and another used barium enema examination for Japanese patients (n = 920) [15]. Sex hormones can affect bowel transit time in women during different stages of their menstrual cycle, although they do not appear to have a major effect on bowel function under normal physiological conditions [30]. During childhood, constipation is common in boys, but is much more common in female adults aged 15–50 years, that is, during their reproductive years [30]. In addition, women with severe constipation have a high incidence of having undergone gynecological surgery [30]. In the present study, indeed, female patients had a higher prevalence of undergoing laparotomy and patients with major laparotomy had a trend toward a longer proximal colon. These observations imply that hormonal and/or gynecological changes are associated with bowel habits and may subsequently alter the colon length.
Measurement methods other than CT colonography have individual drawbacks in nature, as mentioned above in the introduction. At present, CT colonography is the most reliable method to measure actual colon length. However, previous studies using CT colonography showed various values for the length of the entire colon, as summarized in Table 5 [17,18,19, 21, 31]. It is obvious that patients with incomplete colonoscopy have a long colon [17]. The colon in Americans appears longer than in Japanese patients, but we must consider differences in the measurement software used in the USA and Japan. As part of image processing, both software packages (V3D Colon, Viatronix Inc., USA, and AZE Virtual Place, Canon Medical System, Japan) automatically create a centerline that serves as the focal point for three-dimensional fly-through image review. The length of the centerline from the anus to the cecal tip represents the length of the entire colon, but the centerlines automatically generated depend on the software used. For instance, a submerged lumen with fecal fluid and collapsed segments of colon are not identically considered by the different software, and the software is evolving year by year, causing distinct differences. Nagata et al. analyzed the American colon using an open-access database (the Cancer Image Archive) and reported that the total colorectal length in Americans (n = 650) was significantly longer than in Japanese (n = 650) but the difference was very small [31], which suggests that racial difference in colon length likely do not exist. Considering that colon length was not associated with height in a previous study [21] as well as in the present study, the difference between American and Japanese colon lengths is negligible.
Table 5 Reports measuring the entire colon length using CT colonography This study has several acknowledged limitations. First, this is a post hoc analysis using data obtained from a previous multi-center trial, and the sample size was not calculated. Second, all patients were Japanese manifesting positive FIT, and the generalizability of these findings may be limited. Third, we did not apply the Rome IV criteria to assess bowel habits but asked about defecation frequency alone. Accordingly, patients with functional constipation may be classified into a non-constipated group. Fourth, colon length was measured by one author (KU). This could result in a lack of objective analysis because colon length was manually estimated in case of failure of the automatic process.
In conclusion, this study has clearly demonstrated that bowel habits (defecation frequency) and gender both correlate with the length of the entire colon measured by CT colonography, and in particular, the proximal colon. Proximal colon length might be a new surrogate biomarker to objectively diagnose slow transit constipation.