Association between smoking and colorectal cancer in Eastern Mediterranean Regional Office (EMRO): A systematic review and meta-analysis

Abstract

Background:

Smoking poses a significant risk for colorectal cancer (CRC), considered the third leading reason for cancer-related deaths worldwide. However, there has been limited research on the relationship between smoking and CRC in the Eastern Mediterranean Regional Office (EMRO). Therefore, a meta-analysis was conducted to combine available data and gain a comprehensive understanding of the relationship between smoking and CRC in EMRO.

Methods:

Two independent researchers searched PubMed, Scopus, and Web of Science until December 2022. The included studies were checked for risk of bias administering the Newcastle-Ottawa scale. Heterogeneity was evaluated using I2 statistics and the Cochrane test. Publication bias was determined through funnel plot analysis and Egger’s regression test. Additionally, a meta-regression analysis explored the impact of a country’s Human Development Index (HDI) on the relationship between smoking and CRC.

Results:

The final analysis included 26 studies, revealing a significant association between smoking and CRC (OR = 1.40; 95% CI: 1.11 - 1.78; P = 0.004). Moreover, smoking had a more pronounced adverse effect on CRC in countries with higher HDIs compared to those with lower HDIs (OR = 1.30; 95% CI: 0.99 - 1.71; P = 0.054).

Conclusions:

Our findings underscore the importance of implementing smoking cessation programs and policies in EMRO countries, as they demonstrate a positive relationship between smoking and the risk of CRC. Furthermore, the results suggest that a country’s level of human development may influence the association between smoking and CRC. Further research is needed to investigate this potential connection and develop targeted public health interventions.

INTRODUCTION

Colorectal cancer (CRC) is a malignant condition that develops in the colon or rectum. It develops from polyps of abnormally growing cells in the rectum or colon lining. These polyps have the potential to migrate to other areas of the body and turn cancerous over time.[1,2] Changes in bowel habits, blood in the stool, abdominal pain, and unexplained weight loss are some of the signs of CRC.[3] Age is a risk factor for CRC development, with most instances occurring in adults over 50 years.[4,5] Inflammatory bowel disease, having a family history of the condition, and eating a lot of red and processed meat are additional risk factors for CRC.[6]

The incidence of CRC is increasing, and it is one of the leading causes of cancer-related mortality globally.[7] According to statistics, CRC was the cause of 9.4% of cancer-related fatalities and 10% of newly diagnosed cancer cases.[8] According to a recent study, the number of CRC cases and fatalities could increase to 3.2 million and 1.6 million, respectively, by 2040.[9] The incidence of CRC has been rising in several regions of the world recently, particularly in the Eastern Mediterranean Region (EMRO).[10,11] With an incidence rate of 5.8%, CRC is the second leading cause of mortality among EMRO nations.[12] Twenty two nations make up the EMRO, including Yemen, Afghanistan, Bahrain, Djibouti, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Pakistan, Palestine, Qatar, Saudi Arabia, Somalia, Sudan, Syria, Tunisia, and the United Arab Emirates.[13]

Smoking has been established as a major risk factor for the development of CRC, despite the presence of other risk factors.[14] Smoking is a risk factor that can be reduced and is linked to numerous health risks.[15,16] It has been linked to a number of different malignancies, including bladder, gastrointestinal, and lung cancer.[17-20] Tobacco smoke contains substances that can damage DNA, encourage cell growth, and prevent apoptosis, all of which can result in the growth of malignant cells.[21]

However, there is little evidence of a connection between smoking and CRC in the EMRO region. While some investigations[22,23] have demonstrated a connection between smoking and CRC, other studies[24-26] have found no connection at all. We conducted a thorough meta-analysis study to assess the connection between smoking and CRC in EMRO nations in light of the study’s contentious findings.

MATERIALs AND METHODs

The current study is an evaluation of the possible impact of smoking on CRC in the EMRO region by a systematic review and meta-analysis. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist’s reporting requirements for disclosing the results of our investigation.[27]

Search strategy and study selection

The following keywords were used to conduct a thorough search on PubMed, Scopus, and Web of Science from inception to December 2022: “Iran” “Afghanistan”, “Bahrain”, “Djibouti”, “Egypt”, “Iraq”, “Jordan”, “Kuwait”, “Lebanon”, “Libya”, “Morocco”, “Yemen”, “Oman”, “Pakistan”, “State of Palestine”, “Palestine”, “Qatar”, “Saudi Arabia”, “Somalia”, “Sudan”, “Syria”, “Tunisia”, “United Arab Emirates”, “Factor, Risk”, “Relate factors”, “Social Risk Factors”, “Associated factors”, “Colorectal cancer”, “Colorectal Carcinomas” and “Colorectal Neoplasms”. Additionally, we looked through the included papers’ reference lists. Two independent researchers (MH. K and E. AS) were employed to reduce the possibility of bias. A third researcher (S.H.) handled any disagreements.

Inclusion and exclusion criteria

All the studies with the following criteria were included: Case–control studies done in one of the EMRO countries with at least one relevant piece of information about smoking status in both case and control groups. English papers with available full text. Articles from other countries that EMRO got excluded. Studies selecting their control group from patients with benign tumors and studies evaluating CRC and colon polyps together in the case group were excluded. In addition, Systematic reviews, meta-analyses, narrative reviews, randomized clinical trials, editorials, and commentaries were not included. Case–control studies were selected based on their suitability for exploring the relationship between exposures and outcomes, particularly in the context of rare diseases or long-term exposures. Moreover, case–control studies provided readily available data on exposure and outcome measures, which facilitated the comprehensive data extraction required for this meta-analysis. In addition, case–control studies offered higher statistical power compared to other study designs when examining infrequent outcomes like CRC. The decision to include only case–control studies was made to maximize the accuracy and reliability of the results of the study.

Data extraction procedure

Two investigators (MH. K and E.AS) extracted the included studies’ data and entered them into a predesigned Excel Sheet. The extracted data included: First author’s name, year of publication, country, sample size, basic characteristics of patients, and status of smoking (Yes/No and Current/Former/Never). Any discrepancies were resolved by the third researcher at this stage (S.H).

Quality assessment of studies

Newcastle-Ottawa (NCO) quality assessment scale checklist was used to assess the quality of included articles. It has several questions in three sections for assessing articles’ quality regarding selection, comparability, and exposure. The tool aims to examine the methodological quality of the studies and show probable errors in each literature.

Statistical analysis

Version 3 of the Comprehensive Meta-Analysis (CMA) software was used to evaluate the current study. The Cochrane test and I2 statistic were used to look at heterogeneity (the significance level was P < 0.1 and I2 > 50%). The random effects model was employed when there was heterogeneity; otherwise, the fixed effect model was applied. For the meta-analysis, we used odds ratios (OR) with 95% confidence intervals to summarise and assemble the data. We performed subgroup analysis based on the human developmental index (HDI) level and smoking status (current/former/never) in order to look into the causes of heterogeneity. Additionally, the effect of a greater HDI rate on the association between smoking and CRC was examined using meta-regression. A sensitivity analysis was done to determine how each study affected the combined outcome. To examine any potential publication bias, Egger’s regression test was utilized along with a funnel plot inspection. To determine whether the sample size was sufficient, power analysis was also used.

RESULTS

Study selection

Figure 1 shows the detailed study selection process. After a search of the databases, as mentioned earlier, 2,547 articles were identified, of which 745 were duplicated. From the remaining 1,802 studies, which went for the title and abstract screening, 1,725 did not meet the inclusion criteria and were excluded. A total of 77 studies were screened for a full-text assessment, and 51 were deleted at this stage. Finally, 26 studies were entered into the meta-analysis.

Figure 1.

Study selection process

Characteristics of the included studies

Table 1 displays details of all included studies. Twenty-six studies from EMRO countries with a total sample size of 13,022 patients were included in our study. Among them, 13 studies were from Iran,[26,28-39] five studies were from Jordan[24,25,40-42] and the eight remaining were from Morocco,[23] Saudi Arabia,[43] Oman,[44] Palestine,[45] Kuwait,[46] Pakistan,[22] Egypt,[47] and Qatar.[48] The sample size within the studies ranged from 72 to 2,906. Figure 2 shows the quality of studies based on NCO checklist.

Table 1.

Characteristics of included studies

First Author, year	Journal	Country	Study Period	Sample Size	Male/Female	Smoking Status report	Quality score
Alsheridah, 2018[46]	BMC Cancer	Kuwait	2016-2017	309	168/141	Yes/No	8
Rafiee, 2020[26]	European Journal of Cancer Prevention	Iran	NA*	370	199/171	Yes/No	6
Bener, 2010[48]	Asian Pacific Journal of Cancer Prevention	Qatar	2008-2009	428	249/179	Yes/No	6
Khatabeh, 2018[40]	Asian Pacific Journal of Cancer Prevention	Jordan	2016-2017	300	147/153	Yes/No	8
Simonian, 2018[28]	Middle East Journal of Cancer	Iran	2014-2015	437	205/232	Yes/No	7
Azzeh, 2017[43]	BMC Public Health	Saudi Arabia	2014-2015	301	NA	Yes/No	6
Shivappa, 2017[29]	Asian Pacific Journal of Cancer Prevention	Iran	2008-2010	213	105/108	Current/former/Never	8
Arafa, 2011[41]	Asian Pacific Journal of Cancer Prevention	Jordan	2008-2009	440	236/204	Current/former/Never	6
Khan, 2015[22]	Asian Pacific Journal of Cancer Prevention	Pakistan	2011-2015	222	105/117	Yes/No	9
Abolhassani, 2019[30]	Ecotoxicology and Environmental Safety	Iran	2014-2015	72	43/29	Yes/No	7
D’asheesh, 2021[31]	Journal of Gastrointestinal Cancer	Iran	2014-2019	600	NA	Yes/No	4
Khodaverdi, 2021[32]	BMC Cancer	Iran	NA	80	40/40	Yes/No	6
Farahani 2020[33]	Journal of Gastrointestinal Cancer	Iran	2016-2018	170	90/80	Current/former/Never	6
Abu Mweis, 2015[42]	European Journal of Cancer Prevention	Jordan	2010-2012	407	187/220	Yes/No	6
Azizi, 2015[34]	Asian Pacific Journal of Cancer Prevention	Iran	2013-2014	414	220/194	Current/former/Never	7
Golshiri, 2016[39]	International Journal of Preventive Medicine	Iran	NA	200	122/78	Yes/No	8
El Kinany, 2020[23]	European Journal of Nutrition	Morocco	2009-2017	2906	1432/1474	Current/former/Never	9
Tayyem, 2013[24]	Asian Pacific Journal of Cancer Prevention	Jordan	2010-2012	503	262/241	Yes/No	9
Ghrouz, 2021[45]	Nutrition and Cancer	Palestine	NA	210	115/95	Yes/No	7
Tayyem, 2016[25]	SAGE Journal	Jordan	2010-2012	501	248/253	Yes/No	6
Mahfouz, 2014[47]	Cent Eur J Public Health	Egypt	2010-2011	450	216/234	Current/Never	7
Safaee, 2019[35]	International Journal of Cancer Management	Iran	NA	1724	877/847	Current/Never	7
Mafiana, 2018[44]	Asian Pacific Journal of Cancer Prevention	Oman	2016	279	141/138	Current/former/Never	8
Moazzen, 2020[36]	Annals of Global Health	Iran	2004-2015	405	0/405	Current/former/Never	9
Naghibzadeh Tahami, 2016[37]	Iranian Journal of Public Health	Iran	2012-2014	525	390/135	Current/former/Never	9
Abbasnezhad, 2021[38]	Asia-Pacific Journal of Clinical Oncology	Iran	2012-2015	556	254/302	Current/former/Never	9

*=Not Available

Figure 2.

Newcastle-Ottawa (NCO) quality assessment scale

Main results

Results obtained from the meta-analysis revealed a significant association between smoking and CRC (OR = 1.40; 95% CI: 1.11-1.78; P = 0.004) [Figure 3a]. The results were accompanied by significant heterogeneity (I2 = 84.11). Sensitivity analysis was conducted by removing each study to assess its impact on the final result. The result of sensitivity analysis showed no significant change in the pooled result by removing each of the studies [Figure 3b]. Data set analysis did not indicate any signs of publication bias (Egger’s test P = 0.3614; Begg’s test P = 0.6277) [Figure 4a]. In addition, the result of power analysis showed that the mean number of sample size was adequate for our main result [Figure 4b].

Figure 3.

Forest plot of the relationship between smoking and colorectal cancer (a). The results of Sensitivity analysis (b). Forest plot of the subgroup analysis based on current and former smoking (c). Forest plot of the subgroup analysis based on human development index (d)

Figure 4.

Funnel plot of the publication bias assessment (a). Power analysis result (b)

Subgroup analysis

Smoking history-based subgroup analysis revealed a negative correlation between current smoking and CRC, although the connection was not statistically significant (OR = 0.94; 95% CI: 0.59-1.51; P = 0.821). Former smoking increased the risk of CRC but it was not significant (OR = 1.53; 95% CI: 0.96-2.45; P = 0.071) [Figure 3c].

Subgroup analysis based on HDI level showed that smokers in very high HDI level countries were at higher risk of CRC compared to non-smokers, but this relationship was marginally not significant (OR = 1.30; 95% CI: 0.99-1.71; P = 0.054). Smokers in high HDI level countries were significantly at risk of CRC compared to non-smokers (OR = 1.40; 95% CI: 1.01-1.92; P = 0.038) [Figure 3d]. We could not report the risk of CRC among smokers in low and medium-level countries due to the inadequate number of included countries.

Result of meta-regression

In addition to performing subgroup analysis based on a countrys HDI levels, we conducted a meta-regression analysis to find the correlation of the HDI impact on the relationship between smoking and CRC. The results of the meta-regression showed no statistically significant difference between the final result of the meta-analysis and the increased rate of HDI (Coefficient = -0.2876, P value = 0.8898) [Figure 5].

Figure 5.

The results of meta-regression analysis based on human development index of countries

DISCUSSION

Our key finding demonstrated that, in EMRO nations, smoking dramatically elevated CRC risk by 40%. The connection between smoking and CRC was the subject of yet another meta-analysis by Botteri et al.[49] According to their findings, smoking increased the absolute risk of CRC by a significant amount (10.8 cases per 100,000 person-years). Additionally, they stated that smoking at larger doses, particularly beyond 30 years, could greatly increase CRC. Smoking significantly raised the incidence of CRC, according to another meta-analysis research by Tsoi et al.[50] on 20 prospective cohort studies from America, Europe, and Asia; however, this effect was more pronounced in men than in women and in rectal cancer than in colon cancer. In a meta-analysis of 5,229 individuals, Ordóñez-Mena et al.[51] found that smoking can have a significant impact on CRC survival in both current and former smokers. Additionally, they discovered that giving up smoking could increase CRC patient survival rates. Case-control research conducted by Sharpe et al.[52] on 4,000 Canadian males found a link between smoking and CRC. Additionally, they also found a weak positive association between smoking and proximal colon cancer justifying the increasing prevalence of proximal colon cancer due to increased smoking consumption.

Smoking has been shown to have an impact on the site of tumors in CRC in earlier investigations, and to increase the risk of developing left-sided CRC in a dose-response manner in a multi-center analysis of 46,166 Chinese patients with CRC. Additionally, the study demonstrated that smoking has a detrimental effect on the prognosis of CRC on both the right and left sides.[53] Gram et al.[54] conducted a multicenter cohort study in the United States that tracked 188,052 CRC patients for a total of 16.7 years. They discovered that although female smokers were more likely to develop right-sided CRC, male ones were more likely to develop left-sided CRC. In addition, studies have demonstrated a link between smoking and particular molecular subtypes of CRC. Heavy smokers were found to have a higher risk of the CpG island methylator phenotype (CIMP) subtype, according to a large-scale case-control investigation by Wang et al.[55]

According to our research, there was no appreciable difference in CRC risk between current smokers and non (never) smokers. Other research findings were debatable in this regard. Current smoking and CRC were found to be significantly correlated as was shown in a study by Ordóñez -Mena et al.[51] and Botteri et al.,[56] but not by Freedman et al.[57] Additionally, despite being just non-significant, our study showed a favorable correlation between former smoking and an elevated risk of CRC. A favorable correlation between prior smoking and the incidence of CRC was also discovered by Walter et al.[58] In addition, compared to never-smokers, former smokers had a higher rate of CRC, according to Tsoi et al.[50] However, Freedman et al.[57] discovered a negligible correlation between prior smoking and CRC. Because there was high heterogeneity in the subgroup analysis, the findings of our study could be revisited by other research and should be treated with caution.

The carcinogens in tobacco, such as benzopyrene, chloroethylene, nitrosamine, polycyclic aromatic hydrocarbons, nitroso toluene, and nickel, are responsible for smokers’ increased risk of developing CRC.[59] Additionally, in vivo and human investigations on CRC have demonstrated carcinogenesis brought on by polycyclic aromatic hydrocarbons and heterocyclic amines.[60,61] The main molecular pathways involved in colorectal carcinogenesis, including P53 mutations, BRAF mutations, MSI positivity, and CIMP positivity, are also associated to smoking in sporadic CRC.[62,63]

There are a number of reasons why our study’s findings differ from those of other investigators. The outcomes of the research are impacted by the varying levels of smoking exposure for present and former smokers. The outcome of the study may also be impacted by different cigarette products with varying tobacco content. We carried out our meta-analysis in EMRO nations with a variety of ethnicities. Variation may also result from racial differences. Along with smoking, other characteristics can affect CRC, such as dietary preferences, lifestyle, and health conditions.

To investigate socioeconomic variances between nations and possible links to CRC risk, we conducted a subgroup analysis based on the HDI of countries. The three social facets of any country that make up HDI are health, knowledge, and standard of living. Our findings showed that smokers had a reduced risk of CRC in very-high-level nations than in high-level countries. Differences in the likelihood of CRC between high and very-high HDI-level nations may be attributable to factors like improved community health, improved hygiene standards, improved screening programs, higher income levels, or even the type of cigarettes smoked.

Our research has some drawbacks. First, some of the countries in the EMRO region did not provide reports. Second, the sample size for both current and past smoking was quite small. Third, there were not sufficient studies conducted in middle- and low-HDI countries to compare them to high- and very-high-level nations. Additionally, we lacked sufficient data to take into account in our study the duration of smoking (packs/year), smoking status in men and women separately, a brand of cigarettes, and the impact of smoking on colon and rectal cancer individually.

To summarize, our results showed a positive association between smoking and the risk of CRC in EMRO countries. Smokers residing in countries with high HDI levels face a greater risk of CRC in comparison to smokers residing in countries with very high HDI levels. Although former smoking had a positive relationship with CRC development, this effect was not statistically significant. Our study underscores the importance of implementing smoking cessation programs and policies in these regions to mitigate the burden of CRC.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.