Main

Colorectal cancer is the third most common cancer in men and the second in women worldwide and accounts for eight percent of all cancer deaths (Fredriksson et al, 1989; Organization, 2011). It has been proposed that the risk associated with a sedentary lifestyle could be explained by hyperinsulinism or insulin resistance, which might stimulate the growth of colonic cancer cells (McKeown-Eyssen, 1994; Giovannucci, 1995).

Sedentary behaviour, which is distinctly different from physical inactivity, is defined as activities that are done sitting or in reclining posture that expend <1.5 times the basal metabolic rate. It is characterised by prolonged sitting or lying down and absence of whole body movement, such as watching television, desk-bound work, using computer and game-consoles, sitting at work, and sitting in automobiles. (Pate et al, 2008; Owen et al, 2010; Wilmot et al, 2012). Several studies have demonstrated that sedentary behaviour is an independent risk factor for health problems and diseases. Sedentary behaviour may reduce overall energy expenditure (Brown et al, 2009), and it has been positively associated with obesity (Wijndaele et al, 2010), weight gain (Blanck et al, 2007), diabetes, cardiovascular disease (Katzmarzyk et al, 2009; Dunstan et al, 2010; Wijndaele et al, 2011), and prostate, ovarian, breast, and endometrial cancers (Mathew et al, 2009; George et al, 2010; Lynch, 2010). In 2012, a meta-analysis showed that people who spend the higher amounts of time in sedentary behaviours have greater odds of having metabolic syndrome (Edwardson et al, 2012). Another review focused on the association between prolonged TV viewing in adults and risk of type 2 diabetes and cardiovascular disease(Grontved and Hu, 2011).

Some likely biological mechanisms have been put forth to suggest why sedentary behaviour may contribute to the development and progression of colorectal cancer risk, which included adiposity accumulation and metabolic dysfunction. Sedentary behaviour might result in adiposity, which is likely an independent contributor to colorectal cancer and a mediating variable on the other pathways. Colorectal carcinogenesis may be promoted by increased levels of sex hormones, insulin resistance, chronic inflammation, and altered secretion of adipokines (Neilson et al, 2009; Lynch, 2010), which are also mediating pathways of adiposity facilitating colorectal carcinogenesis.

Nowadays, people spend 7.7 h in sedentary behaviours every day, and this number may continue to rise (Matthews et al, 2008). Recently, there has been an increase in the number of studies assessing the association between sedentary behaviour and colon or rectal cancer. However, the results are inconsistent. Two previous systematic reviews (Wolin et al, 2009; Boyle et al, 2012) focused only on the association between physical activity and colon cancer and did not independently analyse the association between sedentary behaviour and colon and rectal cancers. Another review(Boyle, 2012) summarised the association between sedentary behaviour and colon cancer but did not use quantitative techniques to compute summary risk estimates. Thus, we conducted a meta-analysis to evaluate the relationship between sedentary behaviour and colon and rectal cancer.

Materials and methods

Search strategy

Our meta-analysis was conducted according to the checklist of the Meta-analysis of Observational Studies in Epidemiology(Stroup et al, 2000). PubMed, Embase, and Google Scholar were searched up to May 2013. Text word, title word, abstract and subject headings were searched to cover sedentary behaviours and colon and rectal cancer. We used ‘sedentary’ or ‘sitting’ or ‘occupational sitting time’ or ‘occupational work’ or ‘TV’ ‘television’ or ‘screen time’ or ‘computer and game-console use’ or ‘car driving’ combined with ‘colon’ or ‘colorectal’ or ‘rectum’ or ‘rectal’ or ‘bowel’ and ‘cancer’ or ‘neoplasm’ or ‘carcinoma’ as the search terms. In addition, we reviewed the reference lists of retrieved articles to identify any studies that were not identified from the preliminary literature searches.

Inclusion criteria

Studies were considered eligible if they met the following criteria: (1) the study was a case–control or cohort study design, (2) was published in English, (3) was self-reported or reported a job title-based or objective measure of sedentary behaviour, (4) the study reported the relative risk (RR) or odds ratio (OR) with 95% confidence intervals (CIs) for the association between sedentary behaviour and colon and/or rectal cancer.

Data extraction

We extracted the following information from each retrieved article: name of the first author, date and country of study, study design, characteristics of study participants at baseline, duration of follow-up, number of cases, definition and measurement of sedentary behaviour, outcomes, RR/OR (corresponding 95% CI), and confounding factors that were adjusted in the analysis. Data extraction was conducted independently by two authors (YJC and YG), and any disagreements were resolved by discussion.

Quality assessment

We used a quality assessment tool with reference to MOOSE (Stroup et al, 2000), QATSO (Wong et al, 2008), and STROBE (von Elm et al, 2007). The total score available was six points (one point for a prospective study design; two points if a self-reported measure of time spent in sedentary behaviour or classification of job title-based was used; one point if two or more demographic confounders were controlled for in the analysis; one point if analyses controlled for physical activity; and one point for an objective measure of the health outcome). We assigned scores of 0–2, 3–4, and 5–6 for low, moderate, and high quality of studies, respectively. Each study was rated independently by two authors (YJC and YG), with ratings reported in Table 1.

Table 1 Characteristics of studies included in meta-analysis

Statistical analysis

We used random-effect meta-analyses to estimate the summary RRs for the associations between sedentary behaviour and the risks of colon and rectal cancers. We combined the case–control and cohort studies in the primary meta-analysis because ORs and RRs provide similar estimates of risk when the outcome is rare (Greenland, 1987). If a study reported results for occupational and recreational physical activity separately, these were considered as separate reports. One study (Whittemore et al, 1990) was conducted in North America and China and was considered as two independent studies. Any studies stratified by sex and sedentary behaviour domain were also treated as separate reports.

Statistical heterogeneity among studies was evaluated using the I2 statistic, where values of 25%, 50%, and 75% represent cutoff points for low, moderate, and high degrees of heterogeneity, respectively (Higgins and Thompson, 2002). In sensitivity analyses, we conducted leave-one-out analyses (Wallace et al, 2009) for each study to examine the magnitude of influence of each study on pooled risk estimates. Subgroup analyses for sex, study design, ethnicity, sedentary behaviour domain, study quality, body mass index (BMI), and physical activity were conducted to examine the robustness of the primary results. For publication bias, we used the Begg test (Begg and Mazumdar, 1994),the Egger test (Egger et al, 1997), and visual inspection of a funnel plot to assess it. All statistical analyses were performed using STATA version 11.0 (Stata Corp, College Station, Texas, USA). All tests were two sided with a significance level of 0.05.

Results

Literature search

The search identified 1655 articles from the PUBMED, EMBASE, and Google Scholar databases, of which 108 articles were identified as potentially relevant. After retrieving the full text for detailed evaluation, 23 studies examining the association between sedentary behaviour and colon or rectal cancer were identified. A flow chart showing the study selection is presented in Figure 1.

Figure 1
figure 1

Flow chart of study selection.

Study characteristics

Table 1 showed the main characteristics of the included studies that were published between 1984 and 2013. The sample size of studies ranged from 294 to 1 223 908, with a total of 4 324 756. For colon cancer, the number of cancer cases ranged from 70 to 7900, with 27 231 reported colon cancer outcomes. For rectal cancer, the number of cancer cases ranged from 41 to 5328, with 13 813 reported rectal cancer outcomes. With regard to study location, one study was conducted in Australia (Boyle et al, 2011), one in New Zealand (Fraser and Pearce, 1993), two in Turkey (Vetter et al, 1992; Dosemeci et al, 1993), one in China (Whittemore et al, 1990), one in Taiwan (Tang et al, 1999), five in the United States of America (Garabrant et al, 1984; Peters et al, 1989; Whittemore et al, 1990; Howard et al, 2008; Campbell et al, 2013),one in Canada (Parent et al, 2011), and eleven in Europe (Gerhardsson et al, 1986; Arbman et al, 1993; Thune and Lund, 1996; Levi et al, 1999; Tavani et al, 1999; Colbert et al, 2001, Weiderpass et al, 2003; Friedenreich et al, 2006; Johnsen et al, 2006; Moradi et al, 2008; Simons et al, 2013). Five studies reported results for both men and women, three studies reported results for men and women separately, eight studies reported results for men only, and one study reported results for women only. The major adjustment confounding factors included age, density, social class, education, smoking, alcohol use, BMI, and physical activity. There were ten high quality studies and thirteen moderate quality studies in our meta-analysis. The average score for all included studies was 4.4.

Sedentary behaviour and the risk of colon cancer

The results from the random-effect meta-analysis of sedentary behaviour and the risk of colon cancer are shown in Figure 2. The pooled RR was 1.30 (95% CI: 1.22–1.39). Subgroup analysis by study design shows that the combined RR was 1.27 (95% CI: 1.18–1.36) in cohort studies, and the combined OR was 1.46 (95% CI: 1.27–1.68) in case–control studies. A low heterogeneity was detected with an I2=7.7% across case–control studies, and a moderate heterogeneity was observed in cohort studies (I2=50.4%; P=0.005).

Figure 2
figure 2

Forest plot of sedentary behaviour and risk for colon cancer.

Sedentary behaviour and the risk of rectal cancer

The results from the random-effect meta-analysis of sedentary behaviour and the risk of rectal cancer are shown in Figure 3. The pooled RR was 1.05 (95% CI: 0.98–1.13). Subgroup analysis by study design shows that the combined RR for cohort studies was 1.06 (95% CI: 1.01–1.12), and there was no evidence of a significant heterogeneity (I2=0.0%, P=0.472). For case–control studies, the combined OR was 1.06 (95% CI: 0.85–1.33), and there was a medium heterogeneity (I2=39.9%, P=0.068).

Figure 3
figure 3

Forest plot of sedentary behaviour and risk for rectal cancer.

Subgroup analyses

Table 2 showed the results from subgroup analyses examining the stability of the primary results and exploring the resource of potential heterogeneity. The associations between sedentary behaviour and the risk of colon and rectal cancer were similar in most subgroup analyses. Given the influence of physical activity on the association between sedentary behaviour and colon cancer, we also conducted stratified analysis by physical activity. The results remained materially unchanged after adjusting for physical activity (RR: 1.30, 95% CI: 1.16–1.46, P=0.505).

Table 2 Subgroup analyses of relative risk of colon and rectal cancer

Sensitivity analyses

Exclusion of a study (Garabrant et al, 1984) that analysed sedentary behaviour and colon cancer yielded a pooled RR of 1.26 (95% CI: 1.20–1.33), and the statistical heterogeneity was forcefully attenuated (P=0.213, I2=15.7%). We excluded any single study in turn and pooled the results of remaining included studies, which did not change the overall combined RR, with a range from 1.26 (95% CI: 1.20–1.33; P=0.213) to 1.32 (95% CI: 1.24–1.41; P=0.016).

For rectal cancer, we found that exclusion of any single study recalculated the pooled OR for the remainder of the studies and showed that none of them was identified as a possible source of heterogeneity among all the included studies. A low heterogeneity of available data on sedentary behaviour and rectal cancer was observed (P>0.05).

Publication bias

Visual inspection of the funnel plot did not reveal any asymmetry (see Figure 4). We found no evidence of significant publication bias for rectal cancer (Egger test, P=0.083, Begg test, P=0.722,). Visual inspection of the funnel plot showed a weak indication of publication bias in colon cancer (see Figure 5), and there was evidence of asymmetry for association with colon cancer risk (Egger test, P<0.001), whereas the Begg test did not identify evidence of substantial publication bias.

Figure 4
figure 4

Funnel plot of sedentary behaviour and rectal cancer.

Figure 5
figure 5

Funnel plot of sedentary behaviour and colon cancer.

Discussion

The results of our meta-analysis suggest that the risk of colon cancer is 30% higher among sedentary people. Null reverse association exists between sedentary behaviour and rectal cancer, but, in cohort studies, sedentary behaviour is associated with rectal cancer. Similar results were obtained in most subgroup analyses.

Our research result is consistent with a large cohort study, the NOCCA study (Pukkala et al, 2009), for which the investigators found that the risk of colon and rectal cancer was related to some jobs. However, we excluded the study because it focused on the link between occupation and colon and rectal cancer and did not clearly define the occupations as sedentary work. Our included studies either clearly defined the kind of sedentary work or reported the numbers of sedentary hours per day.

Our subgroups analyses identify three findings. A major finding is that the relationship between sedentary behaviour and colon cancer may be independent of physical activity, which is important because it suggests that sedentary behaviour could be an independent determinant of colon cancer distinct from that of physical inactivity. Another finding is that sedentary behaviour increased the risk of colon cancer both in case–control and cohort studies, and the sedentary behaviour was associated with a significantly increased risk for rectal cancer in cohort studies, but not in case–control studies. Finally, we also found that the pooled RR for participants from the United States of America (1.51; 95% CI: 1.33–1.71) was higher than the pooled RR for European participants (1.22; 95% CI: 1.15–1.29). There was null statistically significant risk of colon cancer in Asian participants (1.79; 95% CI: 0.98–3.25), which may result from the limited number of included studies (two case–control studies comprising 2054 participants). On the basis of the above findings, we should pay close attention to the sedentary behaviour risk in different countries. As we know, the more developed the county, the higher the economic level, which may frequently give rise to sedentary behaviour, such as prolonged TV viewing, workplace sitting, and time spent in automobiles. Thus, developed countries should take actions to discourage sedentary behaviour, and developing countries should pay attention to the connection between sedentary behaviour and health consequences.

There are several biologic mechanisms through which sedentary behaviour in general could increase the risk of colorectal cancer. Sedentary behaviour has been shown to increase blood glucose levels and to decrease insulin sensitivity (Healy et al, 2007). Increased blood glucose and decreased insulin resistance are both thought to promote colorectal cancer carcinogenesis (Giovannucci, 2001; Giovannucci, 2007). Sedentary behaviour has also been linked to an increased risk of diabetes and obesity (Hu et al, 2003), both of which are established risk factors for colorectal cancer (Lynch, 2010). Other proposed mechanisms by which sedentary behaviour may increase the risk of colorectal cancer include increasing levels of proinflammatory factors, decreasing levels of anti-inflammatory factors, and vitamin D (Feskanich et al, 2004).

Our meta-analysis has several strengths. First, this is the first meta-analysis to systematically quantify the strength of association between sedentary behaviour and colon and rectal outcomes. Second, our study supports the hypothesis that sedentary behaviour is probably an independent risk factor for the colon cancer. Third, as sedentary behaviours are increasingly prevalent in modern society, the results of our study not only can act as an aetiology explanation but also can increase public awareness.

A few limitations of our meta-analysis should be acknowledged. First, the included studies were conducted among different population groups, and the measurement and categorisation of the sedentary behaviour were highly heterogeneous. Thus, the results of this meta-analysis should be interpreted cautiously. Second, the study relied on job title-based and self-reported engagement in sedentary behaviours, which was likely to cause the misclassification of exposure, and may underestimate the reported associations. Third, the limited information provided in the included studies precluded the possibility of dose-response analysis.

On the basis of our findings, we put forward some suggestions for future studies. First, further research using more objective measures of sedentary behaviour and energy expenditure may make observed effects more accurate and conceivable. In addition, more prospective and interventional studies are needed to explore the underlying mechanisms and to determine the cause and effect relationships that link sedentary behaviour and colorectal cancer.

In conclusion, our meta-analysis suggests that sedentary behaviour is associated with an increased risk of colon cancer. The increased risk of rectal cancer associated with sedentary behaviour in subgroup analyses warrants further studies. Given that sedentary behaviours are increasingly prevalent and pervasive in modern society, the result of our meta-analysis is greatly important to both cancer aetiology and public health education.