Colorectal polyp distribution in relation to age: meta-analysis

Abstract

Background

An increase in early-onset colorectal cancer has been observed globally, despite a decline in colorectal cancer incidence rates in many Western countries. Screening strategies for younger individuals are particularly challenging, as there are limited data on polyps in young individuals to guide clinicians. This study aimed to systematically review the current evidence surrounding polyp distribution in different age groups.

Methods

A literature search was performed in PubMed, Scopus, MEDLINE, Embase (OVID), Web of Science, and Cochrane Review databases using the keywords ‘age’ and ‘polyp’, and Medical Subject Heading terms ‘age of onset’, ‘age factors’, ‘age distribution’, and ‘age groups’, with no restrictions on publication year. Articles published in English that described the distribution of polyps (adenomatous, advanced adenomatous polyps, and sessile serrated lesions) or individuals with polyps according to different age groups were considered for inclusion. Younger patients were defined as those aged < 50 years. The outcomes of interest were the number of patients with polyps in the left and/or right colon, or the number of polyps per side of the colorectum in different age groups.

Results

From 12 470 articles, 24 met the eligibility criteria for the systematic review, and 12 were suitable for meta-analysis. Among younger people, 46.5% had right-sided and 75.9% had left-sided polyps. In comparison, 70.8% of the older group had right-sided and 61.9% had left-sided polyps. Meta-analyses of studies showed a greater proportion of younger people than older people with at least one left-sided polyp (mean difference 0.06, 95% confidence interval (c.i.) 0.03 to 0.09; P < 0.001). There was also a greater proportion of left-sided polyps in younger people (odds ratio 0.77, 95% c.i. 0.59 to 1.01; P < 0.001).

Conclusion

Patients aged < 50 years have a greater tendency towards having polyps in the left colon, compared with people ≥ 50 years of age, similar to the distribution of early-onset colorectal cancer. This has implications for the methodology of screening and investigation of symptoms in those aged < 50 years.

This systematic review and meta-analysis aimed to review the evidence describing the distribution of colorectal polyps between younger and older individuals. The evidence identified suggests that younger people aged < 50 years have an increased propensity for left-sided polyps compared with older people. The pattern presented is consistent with colorectal cancer patterns in this age group, and may inform future screening and endoscopic pathways for people aged < 50 years.

Introduction

Despite a global decrease in incidence, colorectal cancer (CRC) continues to be a significant source of morbidity, mortality, and clinical challenges. In 2020, CRC constituted more than 1.9 million new cancer cases and 0.9 million cancer-related deaths per annum globally, with Australia and Aotearoa New Zealand ranking third highest in global incidence¹. The combined incidence of CRC for Australia and New Zealand is 20.0–23.2 per 100 000 for colon cancer and 7.7 to 13.6 per 100 000 for rectal cancer¹. An area of increasing concern is early-onset CRC (EOCRC).

Although no universally accepted definition exists, EOCRC is generally considered to refer to CRC diagnosed in individuals aged < 50 years. EOCRC accounts for 10% of CRC cases and is the third leading cause of cancer-associated death in patients aged < 50 years². In contrast to late-onset CRC, the incidence of EOCRC has risen steadily in the past several decades, ranging from a 1.6 to 7.9% increase per annum between 1990 and 2016 across age groups < 50 years^3,4. The largest increase in incidence has concerningly been seen in the 20–29-year age range^3,5. Clinicians and researchers have postulated that EOCRC represents a pathology that is biologically distinct from late-onset CRC, and is incompletely understood. Clinicopathologically, EOCRC has a rectal predominance and is more likely to present with advanced pathological stage, with a propensity for mucinous and signet-ring phenotypes^5–9. Furthermore, numerous studies^10–12 have shown that the majority of EOCRCs are sporadic and not explained by germline mutations that cause cancer syndromes.

Although the literature has demonstrated the left-sided and rectal predominance of EOCRC, the distribution of the premalignant polyp in this group has been largely assumed to follow cancer patterns, but this assumption lacks evidence^13,14. Establishing polyp distribution in young people may provide some insight into the causative mechanisms of EOCRC. Emerging evidence has suggested that proximal and distal tumours have biologically distinct features. Proximal tumours are associated with increasing age, tend to be bulky, exophytic masses, and have higher rates of lymph node involvement than distal tumours, which are typically constricting, infiltrative masses leading to obstruction^15–17. Furthermore, there is also a distinction in the distribution of colorectal adenomas, which are more prevalent in the distal colon, and sessile serrated lesions, which are more frequently found in the proximal colon^18–21. Adenomas and sessile serrated lesions have distinct molecular pathways leading to carcinogenesis, and these distinctions may provide clues to the development of EOCRC^22–24. Additionally, establishing distributive trends of polyps in young people may guide both screening and investigative strategies in this age group.

Early detection and screening strategies for colorectal polyps and cancers in individuals aged < 50 years pose a conundrum for healthcare providers. The American Cancer Society²⁵ has recommended lowering the screening age for colorectal polyps and cancers from 50 to 45 years in average-risk populations. Many national screening programmes, such as that of New Zealand, still do not include people aged < 50 years owing to resource constraints. In contrast, others, such as that in Australia, have initiated this change to meet current guidelines^26–29. Flexible sigmoidoscopy has been proposed as a less resource-intensive initial screening tool for younger people. However, a cohesive description of the distribution and characteristics of polyps in the younger age group is currently lacking; this is needed to inform the efficacy and safety of flexible sigmoidoscopy in this group. The aims of this work were to systematically review the available literature describing polyp distribution in individuals of different age groups, and to compare the rates of left and right colorectal polyps between younger and older individuals.

Methods

This systematic review of published literature was conducted according to the PRISMA guidelines³⁰.

Search strategy

The literature search was carried out through PubMed, Scopus, MEDLINE (OVID), Embase (OVID), Web of Science, and Cochrane Review databases. The search was performed using a combination of keywords, Medical Subject Headings (MeSH) terms, and title/abstract searches. The keywords used were ‘age’ and ‘polyp’. The MeSH terms used were ‘age of onset’, ‘age factors’, ‘age distribution’, and ‘age groups’. The title/abstract search term used was ‘polyp’. The MeSH term ‘polyp’ was not used as it produced too many irrelevant articles on preliminary searches. Similarly, using a combination of MeSH terms including ‘colon’, ‘rectal’, ‘colorectal’, and ‘polyp’ produced a search that was insufficiently inclusive of desired articles. Hence, a title/abstract search for the term ‘polyp’ was used. No filters were used. Literature analysis was conducted in March 2025 without any restriction on year of publication of articles.

Articles were included in this review if they described the prevalence of colorectal polyps (including adenomatous, advanced adenomatous polyps, and sessile serrated lesions) and their anatomical distribution within different age groups, and were published in English.

Exclusion criteria were: studies that did not describe patient age or age groups, polyp distribution or location, or polyp distribution in relation to age groups; studies that included patients with inflammatory bowel disease, cancer, or polyposis syndromes in their cohorts who could not be separated from non-high-risk groups for separate analysis; case reports, editorials, letters or conference presentations; studies that described only a paediatric population; studies that described flexible sigmoidoscopy only; and studies that described only non-adenomatous polyps.

Titles and abstracts generated by the search were extracted into a Microsoft^® Excel (Microsoft, Redmond, WA, USA) workbook, and reviewed independently by two researchers. Duplicates were deleted, and titles and abstracts were screened for inclusion. The researchers independently reviewed the full article of any study that met the eligibility criteria.

Outcomes of interest

Patients were divided into younger and older age groups. Younger age groups included individuals aged < 50 years, whereas the older age group comprised individuals aged ≥ 50 years. The right colon was defined as being proximal to the splenic flexure, and the left colon as being distal to the splenic flexure (including rectum). The outcomes of interest identified in the selected articles were descriptions of the number of patients with polyps in the left and/or right colon, or the number of polyps per side of the colorectum in different age groups, or a description of the risk of right or left polyps within age groups. A meta-analysis was undertaken for articles that included cohorts of both younger and older age groups.

Risk of bias

Articles considered for inclusion in this review were assessed for bias using the Joanna Briggs Institute (JBI) scale for cross-sectional and cohort studies by two independent reviewers.

Statistical analysis

The primary meta-analysis was carried out based on the number of people who had at least one left-sided polyp, comparing younger versus older patients, using articles describing the number of people with right- and left-sided polyps. The metafor package for R (R Foundation for Statistical Computing, Vienna, Austria) was used for meta-analysis³¹. Sensitivity analysis was undertaken using leave-one-out analysis.

The proportion of the older group with at least one left-sided polyp was subtracted from the proportion of the younger group with at least one left-sided polyp, and this was considered the difference in proportions. The standard error for the difference in two independent proportions was calculated for each study. A forest plot was produced, and an Egger’s plot was used to look for publication bias.

A secondary analysis was performed of studies describing the number of polyps, wherein a cross-tabulation of age group by left versus right side was analysed using articles describing the number of polyps per side of the colorectum. The conv.wald function was used for meta-analysis of articles that did not have raw numbers in each cross-tabulated cell, but had a calculatable odds ratio. Once again, Forest plots and Egger’s plot and test were used^32,33.

Results

Database searches identified 12 470 articles. Twenty-four articles were included in this systematic review (Fig. 1).

Fig. 1.

PRISMA diagram showing selection of articles for review and meta-analysis

IBD, inflammatory bowel disease.

The most common reasons for exclusion of articles were a lack of description of age groups (24 articles), polyp distribution (18), or polyp distribution in relation to age groups (23). In one article³⁴, right-sided was defined as the caecum and ascending colon, left was defined as the transverse colon to sigmoid colon, whereas the rectum was included in a separate category. Attempts were made to contact the authors to acquire the data set so that the polyp description could be realigned with this review’s definition, but no response was received, and so the article was excluded from analysis.

Table 1 shows the characteristics of the selected articles. Of the 24 articles included, 15^{36–38,40,42–47,49,50,53,55,57} described the number of patients who had right- and left-sided polyps, according to age group. The remaining nine articles^{32,35,41,48,51,52,56,58,59} described the number of polyps in either the right or left colon within each age group. Three^32,37,43 of the 24 articles provided only odds ratios and not raw numbers for analysis. Twelve articles^{32,35,39,40,42,44,46–48,53,56,57} included cohorts in the older and younger age groups, and had adequate data for inclusion in the meta-analysis. The remaining articles described only patients in either the younger or older age groups.

Table 1.

Characteristics of articles included for review

Reference	Year	Study design	Country	Age group (years)	Outcome described	Polyp types	No. of colonoscopies	No. of polyps
Vatn and Stalsberg35	1982	Necropsy Retrospective cross-sectional	Norway	All ages	No. of polyps per side	Adenoma, HP	445*	2743
Nicholson et al.36	2000	Retrospective cross-sectional	Australia	> 40 with family history and > 50	No. of patients per colon side	Adenoma	138
McCashland et al.37	2001	Retrospective cross-sectional	USA	All ages	No. of patients per colon side (odds ratio)	Adenoma (> 9 mm)	30 727	3268
Imperiale et al.38	2002	Retrospective cross-sectional	USA	40–49	No. of patients per colon side	Adenoma, HP	191
Okamoto et al.39	2017	Prospective cohort	Japan	All ages	No. of polyps per side	Adenoma	2942	2585
Yamaji et al.40	2006	Prospective cohort	Japan	All ages	No. of patients per colon side	Adenoma	1472
Strul et al.41	2006	Retrospective cross-sectional	Israel	40–80	No. of polyps per side	Adenoma	195	145
Pendergrass et al.42	2008	Necropsy retrospective cross-sectional	USA	All ages	No. of patients per colon side	Adenoma	2951*
Lieberman et al.43	2008	Retrospective cross-sectional	USA	All ages	No. of patients per colon side (odds ratio)	Adenoma (> 9 mm)	85 525
Bai et al.44	2010	Retrospective cross-sectional	China	10–95	No. of patients per colon side	Adenoma	1012
Min et al.45	2012	Retrospective cross-sectional	Korea	> 45	No. of patients per colon side	Proximal SP	926
Sohrabi et al.46	2014	Retrospective cross-sectional	Iran	> 40	No. of patients per colon side	Adenoma, SP, HP	1208
Hemmasi et al.47	2015	Prospective cohort	Iran	40–59	No. of patients per colon side	Adenoma	106
Visovan et al.32	2014	Retrospective cross-sectional	Romania	All ages	No. of polyps per side (odds ratio)	Adenoma, SP, HP	2436	3642
Klein et al.48	2016	Retrospective cross-sectional	Austria	> 18	No. of polyps per side	Adenoma, SP, HP	704	1767
Laird-Fick et al.49	2016	Retrospective cross-sectional	USA	50–100	No. of patients per colon side	Adenoma, SP	13 881
Chablani et al.50	2017	Prospective cohort	USA	> 50	No. of patients per colon side	Adenoma	138
Pommergaard et al.51	2016	Retrospective cross-sectional	Denmark	40–74	No. of polyps per side	Adenoma	1215	2149
Zhou et al.52	2017	Retrospective cross sectional	China	> 65	No. of polyps per side	Adenoma, SP	297	246
Filho et al.53	2017	Retrospective cross-sectional	Brazil	All ages	No. of patients per colon side	Adenoma, HP	456
Segev et al.54	2020	Retrospective cross-sectional	USA	16–50	No. of polyps per side	Adenoma, SP, HP	5283	8358
Chen et al.55	2019	Retrospective cross-sectional	China	18–49	No. of patients per colon side	Adenoma, SP	4485
Conway et al.56	2019	Retrospective cross-sectional	USA	40–59	No. of polyps per side	Adenoma, SP	677	753
Vithayathil et al.57	2020	Retrospective cross sectional	USA	all ages	No. of patients per colon side	Adenoma, SP, HP	36 620

^* Describes number of post-mortem examinations in study. SP, serrated polyps; HP, hyperplastic polyps.

All selected articles included traditional adenomas, except one⁴⁵ that described only proximal sessile serrated polyps. Eight articles include serrated polyps as well as adenomas, whereas two reported only on adenomas over 9 mm in size (Table 1). Notably, 2 included articles were autopsy studies of 240 subjects with colorectal polyps⁴² and 445 subjects, which yielded 2743 polyps³⁵.

Tables 2 and 3 show the number of individuals with left and right colonic polyps and the total number of polyps in the left and right colon, within age groups, respectively. However, in one study³⁶, the authors described polyps in individuals aged > 50 years or > 40 years with a family history of CRC. Thus, the group aged > 40 years, with a family history of CRC, was excluded from analysis (118 people), as it crosses the age group definition of the present study. In another study⁵¹, the cohort ranged from 30 to 80 years; however, the age division between the younger and older group was 60 years. The younger age group (aged < 60 years) had 285 and 660 adenomas in the right and left colon, respectively, but was not included in the meta-analysis. Two articles^37,43 describing the number of people with left- and right-sided polyps reported only the odds of older age groups having right- versus left-sided polyps, compared with individuals aged < 50 years.

Table 2.

Articles reporting number of people with right and left colonic polyps

Reference	No. of colonoscopies	No. of people		Age < 50 years		Age ≥ 50 years		Odds ratio for having right- versus left-sided polyps*
		< 50 years	≥ 50 years	Right colon	Left colon	Right colon	Left colon
Nicholson et al.36	138		20			8 (40.0%)	15 (75.0%)
McCashland et al.37	30 727	311	2957					< 50 years: 1.00 (reference) 50–59 years: 1.04 (0.75, 1.43) 60–69 years: 1.44 (1.08, 1.94) ≥ 70 years: 1.95 (1.47, 2.59)
Imperiale et al.38	191	191		92 (48.2%)	143 (74.9%)
Yamaji et al.40	1472	563	692	347 (61.6%)	338 (60.0%)	491 (71.0%)	384 (55.5%)
Pendergrass et al.42	2951	29	211	17 (58.6%)	12 (41.4%)	171 (81.0%)	40 (19.0%)
Lieberman et al.43	85 525	248	5386					< 50 years: 1.00 (reference) 50–59 years: 1.12 (0.85, 1.47) 60–69 years: 1.44 (1.09, 1.90) 70–79 years: 1.48 (1.10, 1.98) ≥ 80 years: 1.51 (1.00, 2.28)
Bai et al.44	1012	238	774	105 (44.1%)	149 (62.6%)	356 (46.0%)	479 (61.9%)
Min et al.45	926	16	910			49 (4.8%)
Sohrabi et al.46	1208	39	135	9 (23.1%)	30 (76.9%)	40 (29.6%)	103 (76.3%)
Hemmasi et al.47	106	40	67	10 (25.0%)	32 (80.0%)	21 (31.3%)	50 (74.6%)
Laird-Fick et al.49	13 881		13 881			9702 (69.9%)	5642 (40.6%)
Chablani et al.50	138		138			106 (76.8%)	32 (23.2%)
Filho et al.53	456	51	405	16 (31.4%)	40 (78.4%)	213 (52.6%)	275 (67.9%)
Chen et al.55	4485	4485		1866 (41.6%)	3965 (88.4%)
Vithayathil et al.57	36 620	3234	33 386	2022 (62.5%)	2611 (80.7%)	24 755 (74.1%)	24 361 (73.0%)

Values are n (%), except *values in parentheses are 95% confidence intervals.

Table 3.

Articles describing number of polyps in right and left colon

Reference	Total no. of polyps	Age < 50 years		Age ≥ 50 years		Odds ratio for right colonic versus left colonic polyps*
		Right colon	Left colon	Right colon	Left colon
Vatn and Stalsberg35	2746	3 (37.5%)	5 (62.5%)	783 (58.2%)	562 (41.8%)
Okamoto et al.58	2386	98 (39.8%)	148 (60.2%)	896 (41.9%)	1243 (58.1%)
Strul et al.41	145			54 (37.5%)	90 (62.5%)
Visovan et al.32	2000					< 50 years: 1.00 (reference) 50–70 years: 2.1 (1.31, 3.34) > 70 years: 1.54 (1.05, 2.27)
Klein et al.48	581	13 (59.1%)	9 (40.9%)	351 (62.9%)	207 (37.1%)
Pommergaard et al.51	1078			390 (36.2%)	688 (63.8%)
Zhou et al.52	246			82 (33.3%)	164 (66.7%)
Segev et al.54	8358	3793 (45.4%)	4565 (54.6%)
Conway et al.56	506	106 (34.2%)	204 (65.8%)	71 (36.4%)	124 (63.6%)

Values are n (%) unless otherwise stated, except *values in parentheses are 95% confidence intervals.

Risk of bias

JBI critical appraisal scores for all articles are summarized in Tables S1 and S2. The majority of articles reporting cross-sectional studies were good, with 12 of 21 achieving scores of 8. The remaining articles lacked in the domain of identifying and controlling for confounders, achieving scores of 6 or 7. These articles are at risk of inaccurate association between age and polyp distribution, owing to potential confounders that were unaccounted for.

Three cohort studies were assessed and had JBI scores of 7, 9, and 10. Assessment and strategies to address incomplete follow-up were not applicable to two articles, because of the nature of the study design. Although they were designed as cohort studies, colonoscopy outcomes were measured at the time of recruitment and so no follow-up issues were encountered. One article did not account for confounders and is at risk of biased outcomes.

Polyp pattern in younger and older individuals

In articles^{36,38,40,42,44–47,49,50,53,55,57} describing the number of people with left and right colorectal polyps that provided raw numbers of people within groups, there was a total of 9644 people in the younger age group and 50 721 in the older age group. Among younger people, 46.5% had right-sided polyps and 75.9% had left-sided polyps. In comparison, 70.8% of the older group had right-sided polyps and 61.9% had left-sided polyps. In articles^{35,39,41,48,51,52,54,56} that described the number of polyps per side of the colorectum and provided adequate cross-tabulated data, there were 8944 polyps described in younger people and 5705 in older people. Among younger people, the proportions of right- and left-sided polyps were 44.9 and 55.1% respectively. Among older people, the distribution of polyps was 46.0 and 54.0% respectively. The majority of articles reported that a higher proportion of young people had at least one left-sided compared with right-sided polyp, with the exception of two articles^40,42 (Table 2). Considering the number of polyps, all articles reported that most of the polyps in young people were situated in the left colorectum, except for one article⁴⁸ (Table 3).

In the older patient cohort, three studies^40,42,49 showed a marked prevalence of right-sided compared with left-sided polyps. Conversely, the remaining studies showed similar a distribution or left-sided predominance of polyps. However, of note, many of these are lower-volume studies. Two studies^37,45 reported an increased odds of prevalence of right-sided polyps with increasing age (Table 2). With regard to polyp burden, two studies^35,48 showed higher right-sided compared with left-sided volume in the older age group. All other studies describing polyp numbers showed higher polyp numbers in the left colon. Another study³² showed increasing odds of colorectal polyps being found in the right compared with the left colon in older age groups, when younger patients were used as the reference (50–70 years: odds ratio 2.10, 95% confidence interval (c.i.) 1.31 to 3.34; > 70 years: odds ratio 1.54, 1.05 to 2.27).

Meta-analysis

Meta-analysis showed that a greater proportion of young people had at least one left-sided polyp compared with the older age group (mean difference 0.06, 95% c.i. 0.03 to 0.09; P < 0.001) (Fig. 2a). The meta-analysis revealed low heterogeneity across the included studies (I² = 24.1%, τ² = 0.0004, H² = 1.32), with the Q-test indicating no statistically significant heterogeneity (Q[6] = 8.33, P = 0.22). These findings suggest that the effect sizes were relatively homogeneous, and the H² value of 1.32 indicates that the variability among observed effect sizes was only 32% higher than would be expected by sampling error alone, consistent with low heterogeneity. The overall pooled effect was statistically significant (estimate 0.06, 95% c.i. 0.03 to 0.09; P < 0.001). Sensitivity analysis revealed that the omission of a particular study caused the pooled estimate to become non-significant. There was no evidence of publication bias (Egger’s test: z = 0.32, P = 0.75) (Fig. 2b)

Fig. 2.

Forest plot showing difference in proportion of patients with left-sided polyps between younger and older age groups, and funnel plot for articles describing the number of people with left and right colorectal polyps

a Forest plot and b funnel plot. Differences in proportions are shown with 95 per cent confidence intervals.

The results of assessment of number of polyps per side showed that a greater proportion of polyps were found in the left colorectum in younger people compared with older people (odds ratio 0.77, 95% c.i. 0.59 to 1.01; P < 0.055) (Fig. 3a). This meta-analysis included five studies (4 were excluded owing to missing data). The article by Visovan et al.³² did not have raw numbers available for inclusion in each cross-tabulated cell but did have a calculatable odds ratio, which could be inserted via the conv.wald function³³. Included articles showed moderate heterogeneity (I² = 46.1%, τ² = 0.0393, H² = 1.88). The H² value of 1.88 indicates that the observed variability was 88% higher than expected from sampling error alone, supporting a moderate level of between-study variation. The Q-test for heterogeneity was not statistically significant (Q[4] = 6.97, P = 0.14), although its low power limits strong conclusions. The pooled effect estimate was borderline non-significant (estimate 0.77, 95% c.i. 0.59 to 1.01; P = 0.055), suggesting a possible trend. Sensitivity analysis revealed that the most influential study was the one by Visovan et al.³² wherein the P value rose to 0.31 upon its omission. A funnel plot was constructed to check for publication bias displayed in Fig. 3b, and there was no evidence of publication bias (z = –0.45, P = 0.65). Notably, three^32,35,58 of the included articles reported a significantly larger number of polyps, ranging from 2000 to 2743 polyps, which had considerable weighting on the outcomes of the analysis.

Fig. 3.

Forest plot and funnel plot for articles describing number of polyps per side of the colorectum in younger and older patients

a Forest plot and b funnel plot. Odds ratios for proportion of polyps found in the left (L) colorectum in younger people (Y) compared with older people (O) are shown with 95 per cent confidence intervals. R, right.

Discussion

This meta-analysis showed that, compared with older people, younger people were more likely to have left-sided colorectal than right-sided colonic polyps. Additionally, there was a trend towards a higher burden of polyps in the left colon compared with the right colon in young people. These results are consistent with cancer patterns seen in younger people.

EOCRC has distinct clinicopathological features compared with late-onset disease, notably in its anatomical distribution. Approximately 65% of EOCRCs occur in the rectum, with 18% in the left and 17% in the right colon⁵⁴. In contrast, a right-sided shift is typical with advancing age^60–62. These differences suggest that EOCRC may represent a unique pathological process rather than simply a temporal variant of conventional CRC. Although it is assumed that polyp distribution mirrors cancer location, few studies have addressed this relationship directly. The findings of the present review suggest a predominance of left-sided polyps in individuals aged < 50 years; however, right-sided lesions remain relatively frequent in this cohort.

The role of polyps in the adenoma–carcinoma sequence is well established in colorectal carcinogenesis, with lifetime CRC risk estimated at 4–5%⁶³. Although screening and surveillance protocols have been standardized for older populations, guidance for younger individuals remains unclear. Although the overall CRC incidence is declining globally as a result of screening and risk factor modification, the incidence of EOCRC is rising for largely unexplained reasons^4,11,64–66. Germline mutations, such as those associated with Lynch syndrome, account for only 5–35% of cases, leaving the majority as sporadic^11,67–69.

EOCRC is frequently diagnosed at an advanced stage, with tumours exhibiting poor prognostic features, including poor differentiation, mucinous or signet-ring morphology, and perineural and vascular invasion^14,70–73. These features suggest both an intrinsically aggressive tumour biology and diagnostic delay. Symptom misattribution is common; rectal bleeding and abdominal pain are often overlooked or misdiagnosed in young adults, contributing to diagnostic delays of up to 6 months^70,74. Younger patients may experience significantly longer times to presentation than their older counterparts⁷⁵.

Molecular and environmental influences on EOCRC risk are increasingly being recognized. Sporadic EOCRC demonstrates frequent TP53 and SMARCB1 mutations, but less involvement of KRAS, BRAF, and APC, compared with late-onset cancers^76–78. Established risk factors include male sex, Caucasian ethnicity, family history of CRC, obesity, hyperlipidaemia, and alcohol consumption⁷⁹. Obesity has been associated with a 14% increase in colon cancer risk per 5 kg/m² body mass index increment, although this is not observed for rectal cancer⁸⁰. For serrated polyps, risk factors include smoking, alcohol use, high-fat diets, and obesity; protective factors include non-steroidal anti-inflammatory drug or aspirin use, and high dietary intake of folate, calcium or fibre⁸¹. Alterations in the gut microbiota may also contribute to carcinogenesis, though causality remains speculative^82,83.

Age-related differences in polyp type and distribution were observed. Conventional adenomas predominate in older patients, whereas the prevalence of serrated lesions remains stable across age groups^21,84. Young individuals are more likely to develop sessile serrated polyps than traditional adenomas. Increased polyp size, multiplicity, and proximal location are associated with age and Black ethnicity⁸⁵. Synchronous neoplasia also differs by age; younger patients with colon cancer have higher rates of synchronous advanced adenomas compared with older patients, but lower rates when rectal cancer is present⁶⁶.

Timely access to colonoscopy remains challenging in many health systems, including Aotearoa New Zealand, particularly in the public sector. Balancing limited endoscopic resources with the rising EOCRC burden presents a dilemma. Flexible sigmoidoscopy may offer a pragmatic initial investigation for symptomatic young adults, given the predominance of distal disease⁵⁴. However, up to 47% of polyps in patients aged < 50 years described in this review are right-sided only, raising concerns about missed pathology if colonoscopy is withheld.

In older populations, flexible sigmoidoscopy alone is insufficient, given the rising prevalence of proximal lesions with age. A past study⁶¹ reported that right- and left-sided lesions occur with nearly equal frequency in individuals aged > 50 years, which underscores the importance of full colonoscopic evaluation in this group.

In response to the increasing EOCRC incidence, the American Cancer Society²⁵ has reduced the recommended starting age for screening from 50 to 45 years, with similar changes proposed in the UK and New Zealand. The greatest relative increase in EOCRC incidence has been in the 20–39-year age group, although the absolute risk remains low. Improved risk stratification tools are required to optimize early detection and avoid unnecessary procedures³. Faecal immunochemical testing may assist in selecting younger individuals for endoscopic assessment, but further validation is required.

The majority of studies presented here originated from the USA (9 studies), followed by six articles from Asia (China, Japan, Korea), four from Europe (Austria, Denmark, Norway, Romania), three from the Middle East (Iran, Israel), and the remainder from Australia and Brazil. Hence, there is significant representation of polyp patterns from the US population, which may limit the generalizability of the analysis presented. Furthermore, the 4 largest studies, with cohorts ranging from 13 881 to 85 525 participants, all originated from USA, and are significantly larger than any of the remaining studies, which further limits applicability outside this population. This review highlights the substantial prevalence of polyps and neoplastic potential in young patients. A heightened clinical suspicion and low threshold for endoscopic evaluation are warranted in symptomatic individuals aged < 50 years. Distal endoscopy may be a reasonable first-line investigation; however, the prevalence of right-sided lesions in this group requires further exploration to avoid underdiagnosis. Future studies should clarify polyp distribution patterns in EOCRC, determine the limitations of flexible sigmoidoscopy, and define molecular and environmental risk factors specific to younger patients.

Several limitations of this review warrant consideration. The majority of included studies were retrospective, introducing the possibility of selection bias. There was considerable heterogeneity in polyp reporting criteria, including inconsistencies in lesion size thresholds and the omission of serrated polyps in earlier studies. Consequently, the findings may not fully represent the distribution of sessile serrated lesions. Two of the included studies were autopsy-based, which may have introduced artefactual changes or sampling bias. Additionally, eight studies evaluated only a single age cohort as defined in this review, limiting the ability to perform direct comparative analyses and potentially skewing the estimates of polyp distribution. The inherent limitations of autopsy-based studies, including post-mortem artefacts and non-representative sampling, further restrict the generalizability of these findings.

Overall, this review has highlighted a clear predominance of left-sided colorectal polyps in individuals aged < 50 years, mirroring the anatomical distribution seen in EOCRC. However, right-sided polyps remain a significant minority in younger patients, underscoring the heterogeneity of lesion location in this group. Increasing age is associated with a progressive shift toward a greater prevalence of right-sided polyps and neoplasia. This age-related variation in polyp distribution has important implications for screening and diagnostic strategies, suggesting that, although distal endoscopic approaches may be appropriate as initial investigations in younger populations, comprehensive colonoscopic evaluation remains essential to detect right-sided pathology, particularly in older individuals. Future research should aim to further delineate the mechanisms driving these age-dependent distribution patterns to inform more precise, age-tailored CRC prevention and screening protocols.

Funding

This study was funded by the Department of Surgery and Critical Care, University of Otago, Christchurch, New Zealand.

Author contributions

Sai Tim Yam (Data curation, Methodology, Project administration, Writing—original draft, Writing—review & editing), Jared McLauchlan (Validation), Andrew McCombie (Formal analysis, Investigation, Methodology), Rachel Purcel (Methodology, Supervision, Writing—review & editing), John Pearson (Formal analysis, Investigation, Supervision), and Frank Frizelle (Conceptualization, Supervision, Writing—review & editing)

Disclosure

The authors declare no conflict of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

Data for this review will be available upon reasonable request to the corresponding author.