Early-onset colorectal cancer as an emerging disease of metabolic dysregulation

Abstract

Colorectal cancer (CRC) is one of the most common malignancies and the second leading cause of cancer-related death worldwide. Early-onset CRC (EOCRC), diagnosed in adults under the age of 50 years, has emerged as a pressing public health concern owing to its alarming rise in incidence since the 1990s. This trend, observed in the USA and at least eight other high-income countries, starkly contrasts with the declining incidence rates of late-onset CRC (age 50 years and above), largely attributed to early disease detection and lifestyle changes. Concurrent with the rising number of cases of EOCRC, the burden of metabolic diseases, particularly obesity and type 2 diabetes mellitus (T2DM), has surged among young populations. Despite well-documented links between metabolic dysfunction and late-onset CRC, understanding the precise role of obesity and T2DM in the pathogenesis of EOCRC remains in its infancy. This narrative Review synthesizes evidence on the relationship of obesity and T2DM with EOCRC, focusing on pathophysiological mechanisms and the mediating roles of diet and lifestyle factors. It also discusses potential clinical and public health strategies to address obesity and T2DM for EOCRC prevention, highlighting knowledge gaps and future research directions.

Introduction

Colorectal cancer (CRC) is an invasive cancer that affects the large bowel (also defined as the colon and rectum). CRC remains the third most common cancer and the second leading cause of cancer-related death worldwide, accounting for over 1.9 million incident cases and more than 900,000 deaths annually^1,2. The global burden of CRC is closely tied to socioeconomic development, with overall incidence rates rising sharply in low-income and middle-income countries while stabilizing or even declining in high-income nations over the past decades^3–5. Favourable overall incidence trends in the USA, Canada, Israel, Australia, New Zealand and several European countries are largely attributed to the widespread implementation of screening programmes and early detection of CRC in individuals aged 50 years and older^6,7. By contrast, the incidence of early-onset CRC (EOCRC), typically defined as CRC diagnosed before the age of 50 years, has risen substantially since the 1990s, particularly in high-income regions. Because many young adults are in the prime of their professional lives and maintain considerable caregiver responsibilities, a CRC diagnosis at a young age leads to uniquely profound emotional, psychological and economic consequences^8,9. Thus, understanding the basis for the alarming rise in cases of EOCRC is a high priority.

An estimated 75–84% of EOCRC cases arise with no prior family history or known genetic predisposition^10,11. Although hereditary CRC predisposition syndromes, such as Lynch syndrome, are known to increase the risk of CRC at a younger age than normal, their molecular pathogenesis differs from that of the vast majority of EOCRC^11,12. Based on tumour gene expression, CRCs are classified into four Consensus Molecular Subtypes (CMS1–4), each with distinct biological features¹³. CMS2 (arising from the canonical adenoma–carcinoma sequence) remains the dominant subtype across all age groups, but CRC cases diagnosed in younger individuals show a relatively higher frequency of CMS1 (immune-rich, microsatellite instability (MSI)-high tumours) and a lower prevalence of CMS3 (‘metabolic’ tumours) and CMS4 (‘mesenchymal’ tumours), probably as a result of more cases with Lynch syndrome (CMS1) and fewer cases with KRAS mutations (CMS3), compared with late-onset CRC cases¹⁴. However, because the incidence of Lynch syndrome has remained stable over the past decades, we will focus here on non-inherited forms of EOCRC, for which environmental and lifestyle factors are clearly implicated in driving the rise in incidence.

Concurrent with the rise in EOCRC incidence, the prevalence of metabolic disorders, particularly obesity and type 2 diabetes mellitus (T2DM), has surged among young adults aged under 50 years. Although associations between obesity, T2DM and overall CRC are well documented, the specific roles of obesity and T2DM in the pathogenesis of EOCRC are only beginning to emerge. Thus, this narrative Review examines the potential pathophysiological mechanisms related to obesity, T2DM and EOCRC. Although additional pathways might link metabolic dysfunction with EOCRC, we focus on insulin resistance, chronic inflammation, and altered gut microbial composition and function, areas with the largest bodies of evidence integrating findings from epidemiology studies, preclinical research and translational investigations. Clarifying the relationship between metabolic dysfunction and EOCRC is critical to informing effective prevention strategies. We also discuss potential clinical and public health strategies to address obesity and T2DM for EOCRC prevention, and highlight knowledge gaps and future research directions. This Review is timely given the urgent unmet need to address the dual burdens of CRC and metabolic diseases among young generations.

Descriptive epidemiology of EOCRC in the USA and worldwide

Increasing incidence

Although CRC remains a critical concern worldwide, its increasing incidence among young generations (those aged under 50 years) presents new challenges, particularly in high-income regions where Western lifestyles and industrialization predominate. In 2023, EOCRC accounted for approximately 13% of the 153,000 incident cases of CRC and 7% of the 52,550 CRC-related deaths in the USA^6,15. Between 1988 and 2015, the age-adjusted incidence of EOCRC surged by 65%, with a continued annual increase of 2.4% from 2012 through 2021 (refs. 16–18). Globally, EOCRC incidence has increased in 27 of 50 regions between 2007 and 2017, with the highest average annual percentage changes observed in New Zealand (3.97%), Chile (3.96%), Puerto Rico (3.81%) and England (3.59%)⁷. Individuals born around 1990 face a substantially higher age-specific risk of EOCRC than those born earlier^5,17, reflecting shifts in risk factor prevalence and distribution.

Demographic and clinical characteristics

EOCRC exhibits distinct demographic and clinical characteristics compared with late-onset CRC¹¹ (Table 1). In the USA, non-Hispanic white populations have experienced the steepest increases in EOCRC incidence over the past three decades, while rates among Black populations have remained relatively stable, despite their historically higher incidence of overall CRC than white populations^17,18. Patients with EOCRC are more likely than those with late-onset CRC to present with advanced stages at diagnosis, possibly owing to delayed diagnosis or more aggressive progression¹¹. Tumour location also differs between the two groups, with EOCRC more commonly occurring in the distal colon and rectum, whereas late-onset CRC more frequently arises in the proximal colon and distal colorectum^6,12. As described above, the vast majority of EOCRC tumours exhibit a gene expression profile consistent with the CMS2 subtype; this subtype is characterized by high levels of chromosomal instability, which underlies the canonical adenoma to carcinoma sequence. Compared with late-onset CRC, EOCRC tumours are more likely to harbour mutations in TP53, CTNNB1 and PTEN, and less likely to harbour mutations in APC, KRAS and BRAF^19,20. However, these differences are modest, and the overall spectrum of the most commonly mutated genes in CRC is broadly similar across age groups, particularly when controlling for anatomic location^11,12.

Table 1.

Epidemiologic, clinical, molecular and pathological characteristics of early-onset and late-onset colorectal cancer

Characteristic	Early-onset cases	Late-onset cases
Epidemiologic
Age of onset	Diagnosed under 50 years of age	Diagnosed at 50 years or older
Epidemiologic trends	Rising incidence, particularly in high-income countries since the 1990s; a high and relatively stable incidence in Black individuals and a rapid rise among non-Hispanic white individuals	Stable or declining incidence in many regions
Modifiable risk factors	Obesity, T2DM, Western dietary patterns and foods (for example, red and/or processed meat), vitamin D, sedentary behaviour, smoking, alcohol use and antibiotic use; early and more prolonged exposure is likely to influence EOCRC risk	Obesity, T2DM, Western dietary patterns and foods (for example, red and/or processed meat), vitamin D, sedentary behaviour, smoking and alcohol use, and antibiotic use
Clinical
Anatomical distribution	Predominantly occurs in the rectum and distal colon, with over 70% presenting in the left colon	Evenly distributed between proximal and distal sites
Stage at diagnosis	Frequently present at advanced stages (III or IV), potentially owing to delayed diagnosis, aggressive tumour biology and symptomatic presentation	More often detected at earlier stages (I/II) owing to routine screening
Symptoms	Rectal bleeding, abdominal pain and changes in bowel habits, often misattributed to benign conditions	Might present with similar symptoms to EOCRC but more likely to be detected through asymptomatic screening
Molecular
Genetic predisposition	High prevalence (15–25%) of hereditary syndromes (for example, Lynch syndrome, germline mutations); majority of cases (75–85%) are sporadic	Low prevalence (10–15%) of hereditary syndromes; majority of cases (85–90%) are sporadic
Microsatellite instability and somatic mutations	High rates of MSI-H owing to Lynch syndrome, unique microbial signatures and distinct mutations (for example, TP53, CTNNB1)	Commonly associated with chromosomal instability, sporadic MSI-H owing to MLH1 hypermethylation, and frequent APC, KRAS and BRAF mutations
Pathological
Tumour differentiation	More likely to exhibit poorly differentiated and signet-ring cell histology	Predominantly moderately differentiated adenocarcinomas

EOCRC, early-onset colorectal cancer; MSI-H, microsatellite instability-high; T2DM, type 2 diabetes mellitus

Obesity and T2DM: trends and associations with EOCRC

Rising obesity and T2DM rates

The incidence of EOCRC has risen in parallel with increasing rates of metabolic dysfunction, particularly obesity and T2DM, among young adults since the 1970s^21–23. According to the National Health and Nutrition Examination Survey data, obesity prevalence increased from 13.4% to 41.2% between 1976 and 2018 among adults in the USA under the age of 50 years^24,25 (Fig. 1). The US Panel Survey of Income Dynamics data reported that the highest increase was seen in individuals aged 40–49 years²⁶. Although more optimal measures of adiposity exist^27–29, in this Review, we consider the WHO definition of obesity as a body mass index (BMI) ≥30 kg/m² or ≥25 kg/m² for Asian people for consistency with the literature³⁰.

Fig. 1 |. Trends in the prevalence of obesity and diabetes mellitus and the incidence rate of colorectal cancer among people younger than 50 years in the USA from 1976 to 2018.

Weighted proportions of obesity and diabetes mellitus were estimated using data from the National Health and Nutrition Examination Survey (NHANES). For the prevalence of obesity, estimates in 1976–1980 and 1988–1994 were derived from Flegal et al.²⁴. For the prevalence of diabetes mellitus (type 1 and type 2), estimates in 1980 and 1988–1994 were derived from Geiss et al.³¹. The prevalences of obesity and diabetes mellitus since 1999 were estimated using data collected from NHANES participants aged 20–49 years from 1999–2000 to 2017–2018, accounting for complex survey design. The observed incidence rates of early-onset colorectal cancer (EOCRC) for each time point were obtained from the National Cancer Institute SEER*Explorer.

Similar to the rise in obesity rates, the incidence and prevalence of T2DM among adults aged 20–44 years nearly doubled between 1980 and 2012 in the USA, as did the incidence of T2DM among children and adolescents from the early 2000s to 2018 (ref. 31). This increase in the youth incidence of T2DM is close to that of type 1 diabetes mellitus and parallels the rise in childhood obesity^32–34 (Fig. 1), suggesting that early-onset T2DM might no longer primarily be a condition of β-cell dysfunction and abnormal levels and/or distribution of adipose tissue (lipodystrophy) but rather might reflect similar pathophysiological mechanisms of insulin resistance to those that are seen in late-onset T2DM.

Evidence linking obesity and T2DM with an increased risk of EOCRC

Although these trends suggest a potential role for obesity and T2DM in influencing the risk of EOCRC, it is important to note that ecological data alone do not prove causality. Nonetheless, emerging evidence indicates that obesity and T2DM might be associated with an elevated risk of EOCRC^35–38 (Tables 2 and 3).

Table 2.

Associations between obesity and risk of early-onset colorectal cancer

Author (reference)	Study design, country	Sample size, age at baseline, sex	Risk factor	Comparison groups	Association with EOCRC (95% CI)	Multivariable adjustment
Kim et al., 2016192	Retrospective cohort study (Kangbuk Samsung Health Study), South Korea	564 advanced colorectal neoplasms (including 25 EOCRC cases) among 59,782 participants; <50 years; male and female	Obesity	BMI ≥25 kg/m2, Yes versus No	OR: 1.23 (1.03–1.47)	Age, sex, smoking, alcohol, obesity, triglyceride ≥150 mg/dl, HDL-C <40 mg/dl for men and <50 mg/dl for women, LDL-C, diabetes mellitus, hypertension, insulin, regular exercise, CRC family history and CEA
Kantor et al., 2016193	Prospective cohort study (Swedish Military Conscription Register), Sweden	885 CRC cases among 239,658 participants; 16–20 years; male	Obesity	BMI ≥30 kg/m2 versus normal BMI in late adolescence (16–20 years)	HR: 2.38 (1.51–3.76)	Age at conscription, erythrocyte volume fraction, household crowding, health status, systolic blood pressure, diastolic blood pressure, muscular strength, physical working capacity and cognitive function
Levi et al., 2017194	Prospective cohort study (Israeli Jewish adolescents), Isreal	1,089 EOCRC cases among 1,794,570 participants; 16–19 years; male and female	Obesity	Baseline BMI ≥30 kg/m2 versus normal BMI	HR: 1.11 (0.61–2.03) (diagnosed at ≤35 years)	Not specified for stratification analysis; year of examination, age at examination, sex, country of birth and socioeconomic status were included in the primary analysis
					HR: 1.58 (1.02–2.44) (diagnosed at 36–45 years)
Liu et al., 201838	Prospective cohort study (The Nurses’ Health Study II), USA	114 EOCRC cases among 85,256 participants; 25–42 years; female	Obesity	BMI ≥30 kg/m2 versus normal BMI	RR: 1.93 (1.15–3.25)	Height, history of diabetes mellitus, smoking pack-years, physical activity, alcohol intake, regular use of aspirin, NSAID, multivitamin use, menopausal status, menopausal hormone use, dietary intake (total calories, red meat, fibre, folate, calcium and AHEI-2010), lower endoscopy within the past 10 years due to screening or other indications and family history of CRC
				BMI ≥23 kg/m2 versus normal BMI at 18 years	RR: 1.63 (1.01–2.61)a
Dash et al., 2020195	Prospective cohort study (Black Women’s Health Study), USA	113 EOCRC cases among 528,419 person years; <50 years; female	Obesity	BMI ≥30 kg/m2 versus normal BMI	RR: 0.97 (0.55–1.71)	Age, education, smoking, alcohol intake, family history of CRC in a first-degree relative, CRC screening, NSAID use, total energy intake, red meat intake, fruit and vegetable intake, menopausal status and physical activity
Jin et al., 2022196	Prospective cohort study, Korea	8,320 EOCRC cases among 5,672,153 participants; 20–49 years; male and female	Obesity	25≤ BMI <30 kg/m2 versus normal BMI	HR: 1.19 (1.12–1.25)	Age, sex, smoking status, alcohol consumption, physical activity and low-income status
			Severe obesity	BMI ≥30 kg/m2 versus normal BMI	HR: 1.45 (1.31–1.61)
			Waist circumference	Male 90–100 cm, female 85–95 cm versus male <80 cm, female <75 cm	HR: 1.28 (1.20–1.37)
				Male ≥100 cm, female ≥95 cm versus male <80 cm, female <75 cm	HR: 1.53 (1.34–1.74)

AHEI–2010, Alternative Healthy Eating Index 2010; BMI, body mass index; CEA, carcinoembryonic antigen; CI, confidence interval; CRC, colorectal cancer; EOCRC, early-onset colorectal cancer; HDL-C, high-density lipoprotein cholesterol; HR, hazard ratio; LDL-C, low-density lipoprotein cholesterol; NSAID, non-steroidal anti-inflammatory drug; OR, odds ratio; RR, relative risk

^aFurther adjusted for weight change since 18 years

Table 3.

Associations between T2DM and risk of early-onset colorectal cancer

Author (reference)	Study design, Country	Sample size, baseline age, sex	Risk Factor	Comparison groups	Association with EOCRC (95% CI)	Multivariable adjustment
Li et al., 202145	Nested case-control (claims-based IBM MarketScan Commercial Database), USA	6,001 EOCRC cases versus 52,104 frequency matched controls; 18–49 years; male and female	T2DM	Yes versus no	OR: 1.24 (1.09–1.41) (overall)	Age, sex, duration of insurance enrolment, region, prescription drug coverage, employment status, residence, health plan, CCI without T2DM and cancer and any of the following conditions between 91 days and 2 years before the index dates: IBD, obesity, family history of gastrointestinal cancer, screening colonoscopy, other colonoscopy and fecal occult blood test
					OR: 1.23 (1.03–1.46) (Men)
					OR: 1.26 (1.04–1.52) (women)
					OR: 1.13 (0.94–1.36) (controlled)
					OR: 1.37 (1.12–1.67) (uncontrolled)
					OR: 1.59 (1.08–2.35) (complicated)a
Jin et al., 2022196	Prospective cohort study, Korea	8,320 EOCRC cases among 5,672,153 participants; 20–49 years; male and female	Fasting glucose	High (100 mg/dl or on medications for diabetes mellitus) versus normal	HR: 1.08 (1.03–1.13)	Age, sex, smoking status, alcohol consumption, physical activity and low-income status
Luo et al., 202448	Prospective cohort study (UK Biobank and Kailuan Cohort), UK and China	357 EOCRC cases among 1,744,135 person years (UK Biobank); 40–55 years; male and female	Blood glucose	Elevated glucose >7.0 mmol/l versus normal level ≤7.0 mmol/l	HR: 1.61 (1.07–2.44) (UK Biobank)	Age, sex, race, education, Townsend deprivation index, assessment centre, CRC screening history, family CRC history of parents and siblings, standard polygenic risk score for CRC and diabetes mellitus medication uses and healthy lifestyle index (UK Biobank)
		112 EOCRC cases among 1,366,586 person years (Kailuan Cohort); <55 years; male and female			HR: 1.65 (0.80–3.42) (Kailuan Cohort)	Age, sex, education, income status, diabetes mellitus medication uses (Kailuan Cohort)

CCI, Charlson Comorbidity Index; CI, confidence interval; CRC, colorectal cancer; EOCRC, early-onset colorectal cancer; HR, hazard ratio; IBD, inflammatory bowel disease; OR, odds ratio; RR, relative risk; T2DM, type 2 diabetes mellitus

^aComplicated cases were defined based on ICD-9-CM diagnosis codes, including T2DM with renal, ophthalmic, neurological manifestations, peripheral circulatory disorders, other specified manifestations or unspecified complication.

Among the earliest data focusing on EOCRC, an analysis of 85,256 women aged 25–44 years in the Nurses’ Health Study II found that individuals with obesity (BMI ≥30 kg/m²) had almost twice the risk of EOCRC than those with a BMI in the normal range (18.5–22.9 kg/m²)³⁸. Although subsequent findings from retrospective cross-sectional or case–control studies have been mixed, probably owing in part to reverse causality, several reports suggest that the obesity–CRC association is more pronounced in individuals under 50 years^39,40. A meta-analysis of six prospective cohort studies conducted in the USA, Sweden and Israel reported an 88% (95% CI 40–154%) increased EOCRC risk among individuals with obesity³⁷.

Molecular epidemiology evidence supports the relevance of obesity to EOCRC risk. One report demonstrated that obesity is associated primarily with CRC arising via the canonical adenoma–carcinoma chromosomal instability pathway⁴¹, which is the most prevalent molecular subtype in EOCRC. Heterogeneity in the obesity–CRC association according to age has also been observed. For example, obesity is more strongly associated with CRC among premenopausal women than among postmenopausal women⁴². Notably, however, the association of BMI with CRC risk is stronger for the colon than for the rectum and is more pronounced in men than in women⁴³. By contrast, EOCRC demonstrates a predominance in tumours of the distal colon and rectum in both men and women. However, the discrepancies in the predilection for tumour subsite and sex observed for EOCRC do not negate a potential role for metabolic dysfunction in EOCRC as most prior studies of site-specific and sex-specific associations primarily included older populations^11,37.

Compared with obesity, the association between T2DM and the molecular subtypes typical of EOCRC is less consistent. T2DM has been linked to BRAF-mutated CRC tumours in cohorts that were enriched for late-onset CRC⁴⁴. BRAF mutations are more frequent in late-onset CRC than in EOCRC, despite CMS1 (its associated molecular subtype) being relatively more prevalent than CMS3 and CMS4 in EOCRC^14,20,44. Epidemiological evidence specifically linking T2DM and EOCRC is emerging (Table 3). A claims-based nested case–control study in the USA found that individuals with T2DM had a 24% (95% CI 9–41%) higher risk of EOCRC than those without T2DM, adjusting for obesity and other confounders⁴⁵. Notably, the risk was more pronounced in individuals with poorly controlled or complicated T2DM compared with those with well-controlled T2DM. A pooled analysis of 13 population-based studies reported a similar but statistically insignificant association, while a Swedish nationwide cohort study suggested that T2DM might be associated with a higher risk of EOCRC than late-onset CRC^46,47. Elevated blood levels of glucose, a hallmark of T2DM, correlated with a 61% increased risk of EOCRC in the UK Biobank cohort compared with a 14% increased risk of late-onset CRC⁴⁸. Notably, the effect size of the T2DM–EOCRC association appears modest compared with the association observed between obesity and EOCRC but is similar to the effect size observed for late-onset CRC^49,50. This limited association might reflect heterogeneity in T2DM pathophysiology, particularly in early-onset cases, which are increasingly characterized by metabolic dysregulation and insulin resistance as well as variability in disease duration, exposure to pharmacotherapies such as metformin, and limited statistical power owing to the relative rarity and frequent misdiagnosis of T2DM among young adults.

Although further mechanistic, longitudinal and subtype-specific studies are needed to clarify the direction and strength of these associations across populations and metabolic phenotypes, and to better assess causality, current data suggest that obesity and T2DM might both contribute to an increased risk of EOCRC. An improved understanding of these relationships will be essential in identifying individuals at high risk and informing targeted prevention strategies.

Pathophysiological mechanisms linking metabolic dysfunction and EOCRC

The endocrine system orchestrates cellular growth and metabolism by integrating systemic nutrient status with cellular activities through molecules like insulin and other signalling molecules⁵¹. Metabolic disorders that disrupt endocrine homeostasis might contribute to colorectal tumorigenesis through dysregulated insulin signalling, chronic low-grade inflammation, and alterations in gut microbial composition and function (Figs. 2 and 3).

Fig. 2 |. Schematic representation of potential mechanisms linking obesity and T2DM to early-onset colorectal cancer.

Key pathways include insulin resistance and hyperinsulinaemia, chronic inflammation, and altered microbiome in the gut. Excess adiposity promotes insulin resistance, leading to chronic hyperinsulinaemia and impaired glucose metabolism. Insulin stimulates cell proliferation and inhibits apoptosis through the phosphatidylinositol 3-kinase (PI3K)–AKT–mTOR and RAS–mitogen-activated protein kinase (MAPK) signalling pathways, creating a pro-tumorigenic environment. Hyperglycaemia further contributes to carcinogenesis by promoting the production of reactive oxygen species, causing DNA damage and aberrant gene expression. Obesity and type 2 diabetes mellitus (T2DM) induce low-grade inflammation, characterized by elevated levels of cytokines and altered adipokine secretion. Chronic inflammation can drive colorectal cancer by promoting DNA damage through oxidative stress and dysfunction of the intestinal epithelial barrier. Macrophage polarization towards the pro-inflammatory M1 phenotype in adipose tissue promotes cytokine release, contributing to insulin resistance and systemic inflammation. Obesity and T2DM alter the composition of the gut microbiota, with reduced numbers of butyrate-producing bacteria and increased numbers of opportunistic pathogens, stimulating oncogenic signalling and promoting inflammation and suppression of antitumour immune responses. Microbial metabolites, such as short-chain fatty acids (SCFAs) and secondary bile acids, modulate inflammation and gut barrier function. Impaired barrier integrity allows bacterial components to infiltrate into surrounding tissues, activating inflammatory and immune responses and promoting tumorigenesis. Additionally, gut dysbiosis associated with obesity or T2DM might also influence colorectal tumorigenesis through gut–brain axis signalling or the insulin–PI3K pathway. Collectively, these interconnected mechanisms highlight how metabolic dysfunction, chronic inflammation and altered gut microbial composition might accelerate colorectal tumorigenesis, with potential age-specific differences contributing to the rising incidence of early-onset colorectal cancer. CRP, C-reactive protein; IGF, insulin-like growth factor; TNF, tumour necrosis factor.

Fig. 3 |. Directed acyclic graphs for hypothesized pathways to EOCRC and late-onset CRC.

This visualization reflects how early-life exposures might influence colorectal cancer (CRC) risk differently between young adults and older individuals. Potential confounders or mediators, including diet, physical activity, medication use, reproductive factors (for example, pregnancies, menopausal hormone use) and comorbidities, might differ across life stages and influence both exposure and outcome. EOCRC, early-onset CRC; T2DM, type 2 diabetes mellitus.

Insulin resistance and hyperinsulinaemia

Insulin is a pivotal regulator of glucose metabolism and cell proliferation, bridging systemic nutrient availability with cellular activities^51–53. This delicate homeostasis is vulnerable to metabolic dysfunction, and obesity is an established risk factor for insulin resistance and T2DM⁵⁴. Notably, beyond energy storage, adipose tissue functions as an endocrine organ that secretes multiple bioactive substances such as metabolic substrates and adipokines, which influence metabolism and inflammation. In obesity, the increased release of these molecules (which include non-esterified fatty acids, glycerol, hormones and pro-inflammatory cytokines) from excess adipose tissue impairs glucose regulation and leads to chronic hyperinsulinaemia and progressive islet β-cell dysfunction, ultimately resulting in T2DM⁵⁴.

Evidence from both human and animal studies suggests that these metabolic disturbances can create a pro-tumorigenic environment in the colon⁵⁵. Hyperglycaemia can generate reactive oxygen species (ROS), which cause DNA damage and promote tumour progression⁵⁶. Meanwhile, insulin can promote cell proliferation and inhibit apoptosis through the insulin–phosphatidylinositol 3-kinase (PI3K)–AKT–mTOR and RAS–mitogen-activated protein kinase (MAPK) signalling pathways^51,52. Circulating insulin might directly stimulate colorectal epithelial cell growth and tumour angiogenesis through these pathways^52,53. Insulin also increases insulin-like growth factor 1 (IGF1) biosynthesis and bioavailability by suppressing IGF-binding proteins, and IGF1 and IGF2 promote tumorigenesis via autocrine and paracrine signalling⁵⁷. Two rigorous prospective studies provided evidence for a relationship between elevated levels of circulating IGF1 and the risk of CRC^58,59. However, whether IGF1 is an obesity-related hormone remains controversial, and the relevance of free IGF1 and its effect on the tumour microenvironment warrants further investigation. Notably, younger individuals with obesity or T2DM might experience prolonged exposure to insulin compared with their older counterparts, which might thus amplify the effects of insulin signalling on cell proliferation and subsequent tumour formation over the course of a lifetime. In addition, obesity-related and T2DM-related insulin resistance disrupts hepatic glucose and lipid metabolism, contributing to hepatic insulin resistance, a key driver of systemic metabolic dysfunction and inflammation that might further promote colorectal tumorigenesis^60,61.

Additional endocrine and metabolic pathways, such as those involved in specific insulin-resistant conditions like polycystic ovary syndrome and autoimmune disorders, might further contribute to systemic metabolic dysfunction^62,63. However, these mechanisms are not yet well characterized in the context of EOCRC and are beyond the scope of this Review.

Chronic inflammation

As mentioned above, obesity and T2DM can contribute to CRC development by inducing chronic low-grade inflammation, characterized by elevated levels of circulating pro-inflammatory cytokines, such as IL-6 and tumour necrosis factor (TNF), and acute-phase proteins such as C-reactive protein (CRP)^61,64. Chronic inflammation can drive CRC by promoting DNA damage through oxidative stress and dysfunction of the intestinal epithelial barrier^65,66. In response to inflammation, tissue-resident and recruited immune cells, including macrophages and neutrophils, release ROS and nitrogen species, which can cause DNA damage in intestinal epithelial cells and initiate tumorigenesis⁶⁷. Inflammation also disrupts the integrity of the epithelial barrier, which exposes intestinal stem cells to environmental mutagens such as microbial toxins, which increase the generation of ROS and induction of DNA damage⁶⁸. Additionally, chronic intestinal inflammation might drive excessive tissue regeneration, fostering the proliferation and clonal expansion of tumour-initiating cells, as well as the dedifferentiation of non-stem cells into stem-like cells to promote tissue repair^69,70. Inflammation can also affect signalling induced by key cytokines (such as those mediated by TNF and IL-1 receptor signalling), further driving tumour initiation and progression^71,72.

While excess adiposity increases the levels of pro-inflammatory markers and acute-phase proteins, a weight loss intervention could reduce them^73,74. Notably, visceral adipose tissue secretes more pro-inflammatory adipokines, such as TNF, and less adiponectin (an anti-inflammatory molecule) compared with subcutaneous adipose tissue^64,75. A meta-analysis of observational studies showed that visceral adipose tissue was associated with a statistically significant increase in the risk of CRC precursors after adjusting for subcutaneous adipose tissue, suggesting a potential role for visceral adipose tissue-driven inflammation in colorectal carcinogenesis⁷⁶.

Obesity-related and T2DM-related inflammation is further exacerbated by the infiltration of macrophages into adipose tissue. The polarization of these innate immune cells to a pro-inflammatory, M1 type leads to the release of cytokines, further contributing to insulin resistance and systemic inflammation⁶¹. The overload of glucose and free fatty acids in the context of obesity and/or T2DM triggers oxidative stress and activates inflammasomes (for example, NOD-like receptor protein 3 (NLRP3)) in adipose tissue and pancreatic cells^61,77. This activation stimulates downstream inflammatory pathways (for example, cytokines IL-1β and IL-18) and further oxidative stress, which can cause DNA damage and promote colorectal tumorigenesis. Moreover, hyperglycaemia can promote tissue damage through the accumulation of glycated biomolecules and advanced glycation end products, thereby also triggering inflammatory responses⁷⁸.

Chronic low-grade inflammation might be particularly relevant to EOCRC given the earlier and more prolonged exposure to proinflammatory conditions associated with obesity and T2DM that begins in childhood or adolescence. This extended duration of metabolic inflammation might accelerate DNA damage, tissue remodelling and immune dysregulation, potentially contributing to earlier tumour initiation. Moreover, early-life inflammatory exposures might also leave lasting epigenetic and immune imprints that predispose individuals to colorectal carcinogenesis later in adulthood.

Altered gut microbiome

The gut microbiota has a crucial role in maintaining metabolic homeostasis and immune function. Notably, obesity and T2DM are appreciably associated with gut microbial dysbiosis⁷⁹. Overall, an estimated 25% of species associated with CRC are associated with cardiometabolic risk⁸⁰. Certain oral pathogens, such as Fusobacterium nucleatum, that might promote CRC are also enriched in individuals with obesity or uncontrolled T2DM^66,81–83. Fusobacterium nucleatum can drive tumorigenesis by stimulating β-catenin signalling in colonic epithelial cells as well as by upregulating the expression of oncogenes and genes encoding pro-inflammatory cytokines (for example, TNF and IL-17)^84–87. Colonization of the gut by this species of bacterium can promote CRC tumour growth by recruiting immunosuppressive cells such as tumour-associated macrophages, neutrophils and myeloid-derived suppressor cells⁸⁸. The bacterium might also bind to TIGIT, an inhibitory receptor expressed on several types of immune cells, thereby suppressing the function of these cells and facilitating tumour progression⁸⁹. Other bacteria linked to metabolic health can also influence CRC. For example, the presence of Prevotella copri in the gut has been associated with chronic systemic inflammation, regardless of the presence of anti-inflammatory factors (such as fibre)⁹⁰. A global metagenomic data analysis of 56,989 individuals across 32 countries also showed that the presence of the microscopic parasite Blastocystis was positively associated with a healthier cardiometabolic profile and negatively associated with obesity compared with the absence of this parasite, potentially reducing the risk of CRC⁹¹.

Metabolites derived from the gut microbiota can regulate immune and inflammation responses and influence tumorigenesis⁹². Short-chain fatty acids (SCFAs), such as butyrate, exhibit anti-inflammatory properties by inhibiting histone deacetylase in neutrophils, which leads to the decreased production of TNF and nitric oxide and to inhibition of NF-κB signalling⁹³. SCFAs can also downregulate the production of pro-inflammatory mediators, including IL-6, IL-12 and nitric oxide, in intestinal macrophages, and enhance the immunosuppressive functions of FOXP3⁺ regulatory T cells by inhibiting histone deacetylase^94,95. Notably, however, obesity is associated with a reduced abundance of butyrate-producing bacteria as well as with an increased presence of opportunistic pathogens (such as F. nucleatum), which might impair mucosal immunity and contribute to a pro-inflammatory, tumourpromoting environment⁸². These effects can be particularly relevant to EOCRC given the early (and thus prolonged) exposure beginning in childhood or adolescence. Phytate is a naturally occurring plant compound that is typically considered beneficial for its antioxidant and anticancer properties. In contrast to SCFAs, which inhibit histone deacetylase, microbial metabolites of phytate, particularly inositol 3-phosphate generated by commensal bacteria, such as Escherichia coli, can stimulate the activity of histone deacetylase and promote epithelial cell proliferation following intestinal damage, a process that can support tissue repair and confer protection against CRC⁹⁶.

Secondary bile acids are produced by the action of gut microbes on primary bile acids. Certain secondary bile acids, such as lithocholic acid, can stimulate the overexpression of pro-inflammatory cytokines (for example, IL-8) and induce endothelial cell proliferation and tube-like formation in the tumour microenvironment, potentially contributing to CRC progression⁹⁷. Obesity and T2DM can further exacerbate these effects by disrupting bile acid metabolism through altered insulin signalling and gut microbiota composition^98,99. Dysregulation of bile acid–microbiota crosstalk increases the levels of secondary bile acids, which exert detrimental effects on the architecture and function of the colonic epithelium through mechanisms that include oxidative DNA damage and inflammation, contributing to colorectal tumorigenesis¹⁰⁰.

Changes in microbial composition and functionality associated with obesity or T2DM might also influence colorectal tumorigenesis through gut–brain axis signalling or the insulin–PI3K pathway. SCFAs and bile acids influence the secretion of anorexigenic hormones, including peptide YY and glucagon-like peptide 1 (GLP1), thereby promoting gut–brain axis signalling to regulate immunity, intestinal transit time, inflammation and metabolism¹⁰¹. Persistent exposure to metabolic dysfunction, which alters gut microbiota and their metabolites, increases susceptibility to EOCRC. Moreover, impaired gut–brain signalling might further exacerbate dietary and metabolic risk factors early in life. Imidazole propionate is a microbial metabolite found at elevated levels in patients with T2DM; these high levels can promote insulin resistance and potentially enhance tumorigenesis through the insulin–PI3K pathway¹⁰². Obesity and T2DM have also been linked to increased intestinal permeability, thereby facilitating microbial translocation across the gut barrier and activation of pro-inflammatory signalling cascades and immune responses that foster tumour development⁸².

It remains unclear whether a unique microbial signature is specific to EOCRC and/or whether this could be affected by obesity and/or T2DM. Emerging evidence suggests that, although late-onset CRC is often associated with reduced bacterial diversity, patients with EOCRC show increased microbial diversity compared with patients with late-onset CRC¹⁰³. Flavonifractor plautii has been identified as a dominant bacterial species in EOCRC, whereas Streptococcus contains the key phylotype in late-onset CRC¹⁰³. Functional analysis revealed that EOCRC has unique features of bacterial metabolism that are characterized by the dominance of DNA binding and RNA-dependent DNA biosynthetic processes, indicating increased cell proliferation and invasion ability compared with those in the age-matched controls¹⁰³. A 2025 study found that mutational signatures associated with colibactin, a toxic metabolite generated by the pks⁺ E. coli strain, were enriched in EOCRC. However, the study did not find an association between the presence of colibactin-related signatures and the detection of pks⁺ E. coli in EOCRC tumour tissues, suggesting that early-life exposure to colibactin-producing bacteria may influence oncogenic pathways that leave lasting mutational imprints, even after the bacteria are cleared¹⁰⁴. Whether metabolic dysfunction is specifically related to an overabundance of gut microbial colibactin production is unclear. Further research is needed to delineate the microbial contributions to EOCRC pathogenesis.

Collectively, these findings underscore the complex interplay between metabolic dysfunction, insulinaemia, chronic inflammation and altered gut microbiome in colorectal tumorigenesis. Although many pathways are shared between late-onset CRC and EOCRC, distinct features of EOCRC warrant further investigation to identify age-specific drivers and inform targeted prevention strategies.

Association between contributors to metabolic health and EOCRC

Dietary patterns and foods

The sharp rise in cases of EOCRC, particularly in high-income countries, parallels a shift towards Western dietary patterns, characterized by energy-dense foods, refined carbohydrates, sugars, and red and/or processed meats. Randomized trials and meta-analyses of prospective studies have linked such a Western-style diet to obesity and T2DM, which, in turn, elevate the risk of CRC probably through insulin resistance, chronic inflammation and altered gut microbiome, as outlined above^105,106. Beyond the specific studies discussed below, we summarize additional evidence examining the potential link between dietary intake and EOCRC risk in Table 4.

Table 4.

Associations between contributors to metabolic health and early-onset colorectal cancer

Author (reference)	Sample size, baseline agea	Risk factor	Comparison groups/unit increment	Association with EOCRC (95% CI)	Multivariable adjustment
Nguyen et al., 2018141	118 EOCRC cases among 89,278 participants; 20–49 years	Sedentary behaviour	7.1–14 versus ≤ 7 hours/week	RR: 1.12 (0.72–1.75)	Height, family history of CRC, diabetes melitus, screening lower endoscopy within past 10 years, lower endoscopy owing to other indications within past 10 years, smoking, alcohol intake, regular use of aspirin, NSAID, race, multivitamin, menopausal status, menopausal hormone use, dietary intake (total calories, red meat, fibre, folate, calcium, AHEI-2010), physical activity and BMI
			>14 versus ≤ 7 hours/week	RR: 1.69 (1.07–2.67)
Hur et al., 2021123	109 EOCRC cases among 95,464 participants, 20–49 years	SSB	≥2 servings/day versus non-drinkers	HR: 2.18 (1.10–4.35) (adulthood)	Age, follow-up cycle, total calorie intake, race, height, BMI, menopausal status and menopausal hormone use, family history of CRC, pack-years of smoking, physical activity, regular use of aspirin or NSAIDs, current use of multivitamins, alcohol intake, red and processed meat, dietary fibre, total folate and total calcium, AHEI-2010 without SSBs and alcohol, lower endoscopy due to screening or for other indications within the past 10 years
				HR: 3.41 (1.08–10.8) (adolescence)
Zheng et al., 2021109	375 early-onset high-risk adenomab cases among 29,474 participants, 20–49 years	Western diet	High versus low score	OR: 1.67 (1.18–2.37)	Age, total caloric intake, time period of endoscopy, number of reported endoscopies, time in years since the most recent endoscopy, and reason for the current endoscopy, height, BMI, family history of CRC, menopausal status, menopausal hormone use, personal history of T2DM, pack-years of smoking, physical activity, current use of multivitamin, regular use of aspirin and regular use of NSAIDs
		Prudent diet	High versus low score	OR: 0.69 (0.48–0.98)	Same as above
		DASH diet	High versus low score	OR: 0.65 (0.45–0.93)	Same as above + alcohol intake
		Mediterranean Diet	High versus low score	OR: 0.55 (0.38–0.79)	Same as above
		AHEI-2010	High versus low score	OR: 0.71 (0.51–1.01)	Same as above
Yue et al., 2021197	111 EOCRC cases among 94,217 participants, <50 years	Lifestyle index for hyperinsulinaemia	Per 75% increment	HR: 1.86 (1.12–3.07)	Age, follow-up cycle, total energy intake, alcohol consumption, height, race, BMI, family history of CRC, history of diabetes mellitus, smoking pack-years, regular use of aspirin, regular use of NSAIDs, multivitamin use, menopausal status and menopausal hormone therapy, and history of lower endoscopy

AHEI–2010, Alternative Healthy Eating Index 2010; BMI, body mass index; CI, confidence interval; CRC, colorectal cancer; EOCRC, early-onset colorectal cancer; HR, hazard ratio; NSAID, non-steroidal anti-inflammatory drug; OR, odds ratio; RR, relative risk; SSB, sugar-sweetened beverage; T2DM, type 2 diabetes mellitus.

All data derived from the The Nurses’ Health Study II, USA, a prospective cohort study

^aStudy participants were all female individuals

^bHigh-risk adenoma includes adenoma at or greater than 1 cm, or with tubulovillous or villous histology or high-grade dysplasia, or 3 or more adenomas

Western diets can promote gut dysbiosis and chronic intestinal inflammation. A study of three large cohorts in the USA showed that a dietary pattern linked to an excess of sulfur-metabolizing bacteria was associated with an increased risk of CRC, particularly in the distal colon¹⁰⁷. Sulfur-metabolizing bacteria reduce dietary sulfur to produce hydrogen sulfide, a genotoxin that induces inflammation and DNA damage¹⁰⁸. Another study analysed data from the Nurses’ Health Study II and reported that a Western diet was associated with an increased risk of early-onset, high-risk colorectal adenomas, especially in the distal colon and rectum¹⁰⁹. A high-fat diet, which contributes to the development of obesity and T2DM, is strongly associated with the CMS2 molecular subtype of EOCRC^110,111. Both Western and high-fat diets elevate markers of insulinaemia and inflammation (for example, C-peptide, CRP, IL-6 and TNF), which are likely to promote tumorigenesis. A loss-of-function missense mutation in the transcriptional regulatory protein HNF1A is present in many patients with EOCRC, and mice that express this particular variant develop colonic polyps when fed a high-fat diet (but not on a chow diet) via a mechanism involving decreased expression of Cdx2 and increased β-catenin activation¹¹².

Specific food components also contribute to CRC, either directly or through obesity or T2DM¹¹³. Ultra-processed foods (UPFs), defined as foods that have been extensively processed using multiple industrial techniques and that typically contain manufactured ingredients and additives, can comprise up to 60% of the total daily energy intake, particularly in high-income countries^114–116. Randomized controlled trials and systematic reviews have linked a high intake of UPFs with excess weight gain, obesity and T2DM^117–119. High levels of consumption of UPFs have been associated with an average BMI increase of 0.02 kg/m² annually among children and adolescents and, during pregnancy, with an increased risk of obesity in offspring, both potentially contributing to EOCRC in adulthood^120,121. In a new study (available as an abstract), high levels of UPF intake (mean ± standard deviation: 9.9 ± 2.2 servings per day) were found to be associated with a 40% increased risk of early-onset conventional adenomas among women younger than 50 years compared with a low level of UPF intake (3.3 ± 0.7 servings per day)¹²². The consumption of high amounts of UPFs was found to be associated with a 29% increased risk of CRC compared with a low intake in a large prospective cohort of over 46,000 men in the USA¹²³.

Sugar-sweetened beverages (SSBs) are particularly concerning. Women who consumed two or more daily servings of SSBs had over twice the risk of developing EOCRC as those consuming less than one serving per week¹²⁴. Excessive quantities of refined sugars drive obesity and T2DM by accelerating de novo lipogenesis and inducing insulin resistance and inflammation¹²⁵. Fructose, a major component of SSBs (and often also highly prevalent in UPFs), disrupts intestinal barrier function, increasing gut permeability and triggering dysbiosis and endotoxaemia, which activates hepatic macrophages to produce TNF, further driving lipogenesis^125–128. Interestingly, in APC-mutant mice, which are predisposed to developing CRC, high-fructose corn syrup increased the size and grade of tumours, even in mice without obesity or metabolic syndrome, suggesting a direct effect on tumour growth¹²⁹.

Conversely, diets rich in fruits, vegetables and whole grains can reduce the risk of EOCRC. The Mediterranean diet, rich in plant-based foods and containing moderate amounts of fish and dairy, is associated with a low risk of obesity and T2DM. A randomized controlled trial showed that adherence to a Mediterranean diet increased the levels of beneficial gut bacteria, such as Faecalibacterium prausnitzii and Roseburia, while reducing levels of pro-inflammatory species like Ruminococcus gnavus and Ruminococcus torques^130–132. Dietary fibre offers protection against CRC, with stronger associations observed for early-onset cases (particularly for rectal cancer) compared with late-onset CRC^47,133, by improving the control of blood levels of glucose, reducing inflammation and promoting good gut health^134,135. Fibre also functions as a physical barrier that reduces the exposure of epithelial cells to carcinogens by binding to and facilitating the excretion of carcinogens, supporting gut barrier integrity and reducing inflammation through the production of SCFAs via microbial fermentation^136,137. Both the Mediterranean diet and increased fibre intake were found to be associated with the absence of Prevotella copri, a bacterium linked to persistent inflammation^90,138.

The intake of dietary fibre and high-quality carbohydrates in high-income countries remains below recommended levels, while UPF and red and/or processed meat consumption continue to rise, particularly among younger adults^115,139,140. This shift is likely to be contributing to the growing EOCRC burden, emphasizing the need for public health interventions to promote healthy eating habits.

Physical activity and other factors

Factors that increase the risk of CRC in older adults, such as physical inactivity, are increasingly prevalent among younger populations and might be contributing to the rising rates of EOCRC (Table 4). Sedentary behaviour promotes obesity and is linked to a higher incidence of EOCRC than active behaviour, as prolonged sitting impairs muscle function and promotes insulin resistance, largely mediated through visceral adipose tissue^141,142. Excessive alcohol consumption induces oxidative stress, and moderate-to-heavy drinkers have a higher EOCRC risk than do light drinkers¹⁴³. Ethanol impairs lipid oxidation, disrupts appetite regulation and increases hepatic de novo lipogenesis, especially when combined with high-sugar or high-fat diets, and is therefore likely to promote weight gain and increase EOCRC risk^144–146. Antibiotic use might have a direct association with EOCRC risk. Early-life antibiotic use can disrupt the gut microbiota, promoting colonization by pro-tumorigenic bacteria, such as F. nucleatum, and thus increasing EOCRC risk^147,148. Antibiotic-induced dysbiosis might also impair gut barrier function and modulate host immune responses. A case–control study found that the use of broad-spectrum antibiotics was associated with a 13% higher risk of EOCRC compared with non-use¹⁴⁹. However, in a separate case–control study, the authors reported no association between antibiotic use and EOCRC risk¹⁵⁰. Investigations with rigorous study designs and large sample sizes are needed to further evaluate the type and dosage of antibiotic use and the risk of EOCRC.

Considering the prolonged time-span required to shape metabolic diseases and for carcinogenesis, early-life exposures to dietary, lifestyle and other risk factors are likely to be critical determinants of EOCRC. For example, women with a BMI of ≥23 kg/m² at age 18 had a 60% higher risk of developing EOCRC than those with a BMI of 18.5–20.9 kg/m² (ref. 38), underscoring the importance of early-life metabolic trajectories.

Clinical and public health approaches for EOCRC prevention

As the rising incidence of EOCRC mirrors trends in obesity, T2DM and Westernized lifestyles, targeted screening for metabolic and lifestyle risk factors, along with interventions to reduce adiposity and T2DM, could optimize EOCRC prevention (Fig. 4 and Supplementary Box 1). This section summarizes promising strategies worthy of further investigation rather than advocating for immediate implementation. These strategies require further validation through longitudinal and mechanistic studies, particularly in young populations, before any formal recommendations are made.

Fig. 4 |. Clinical and public health approaches for the prevention of EOCRC.

This figure depicts intervention strategies across the continuum of early-onset colorectal cancer (EOCRC) development, highlighting opportunities for prevention and risk mitigation at different life stages. Targeted screening for metabolic and lifestyle risk factors, along with interventions to reduce obesity and type 2 diabetes mellitus (T2DM), could optimize EOCRC prevention. GLP1RAs, glucagon-like peptide 1 receptor agonists.

Screening and risk prediction

Guidelines in the USA recommend starting CRC screening at 45 years old^151,152, yet approximately half of all cases of EOCRC occur before this age¹¹. Hence, earlier screening for younger populations who are at high risk could improve outcomes. Current guidelines recommend early screening by colonoscopy every 5 years for individuals with a first-degree relative diagnosed with CRC or advanced adenomas before the age of 60 years, or two first-degree relatives diagnosed at any age, starting at age 40 or 10 years before the youngest diagnosis, whichever comes first¹⁵³. Interestingly, it has been shown that patients with T2DM reach a 10-year cumulative risk of 0.35% for CRC by the age of 36 years, in comparison to the 0.35% risk at the benchmark age of 45 years old in those without T2DM, suggesting that individuals with T2DM or with a higher overall lifestyle risk might also benefit from earlier screening¹⁵⁴. However, these estimates were from cohorts predominantly comprising women, with limited case numbers and residual confounding. Thus, additional evidence is needed before any clinical recommendations can be made.

While family history is primarily considered for risk stratification, a study combining environmental and genetic (CRC-associated single-nucleotide polymorphism) factors along with family history predicted CRC risk and starting age of screening with greater accuracy than a model with family history alone¹⁵⁵. Developing models that integrate metabolic and lifestyle factors might improve early detection of EOCRC and better inform personalized risk estimates. Moreover, emerging evidence supports the combination of gut microbiome biomarkers and faecal immunochemical tests to improve the sensitivity and specificity of risk prediction for both adenomas and CRC^156,157. This modality holds promise for enhancing risk prediction and screening for EOCRC as more specific bacteria are being identified.

Lifestyle interventions

Tailored interventions that promote weight loss, diet quality and physical activity have demonstrated efficacy in mitigating the risk of obesity and T2DM, both of which are associated with EOCRC. Intensive diet and/or lifestyle counselling over 6 months can effectively achieve at least a 5% weight loss and improve HbA_1c levels^158,159. Such programmes are accessible through primary care, community settings, commercial platforms and technology-empowered formats¹⁵⁹. A systematic review found that community-based lifestyle interventions modelled on the Diabetes Prevention Program achieved an average of 4% weight loss after 1 year¹⁶⁰. Commercial programmes could lead to greater weight loss compared with self-help or usual-care participants^161–164.

Technology has improved access to these lifestyle interventions. Telehealth, smartphone applications and wearable devices offer flexibility and convenience and have been integrated into routine health care and lifestyle interventions. More intensive modalities, such as phone-based programmes, could provide equivalent efficacy to in-person programmes, while more passive approaches (for example, online learning modules, text, and/or e-mail messaging and smartphone applications) appear less effective¹⁵⁹. Tools such as accelerometers and smart scales (which record and track various aspects of body composition), combined with regular feedback with a trained interventionist, could help to address adherence challenges, improving weight loss and blood sugar control¹⁵⁹.

Personalized nutrition programmes that integrate multi-omics, biomarkers and artificial intelligence algorithms show promise for the precision prevention of EOCRC. By combining genetic, epigenetic and metabolic factors, these programmes offer individualized risk assessments and targeted interventions. The PREDICT studies demonstrated that postprandial metabolic responses to food depend on gut microbiome profiles and lifestyle^165–167, leading to the development of an app-based precision nutrition programme, which uses omics profiling, postprandial responses and food logs to personalize dietary recommendations^165,168. An 18-week randomized controlled trial using such an approach showed statistically significant improvements in body weight, waist circumference, triglyceride levels, HbA_1c, diet quality and gut microbial composition compared with standard dietary and/or lifestyle counselling¹⁶⁸. Large-scale efforts, including the All of Us Nutrition for Precision Health Initiative, might offer the unique opportunity to examine the risk of EOCRC according to individual metabotypes, a method of classifying people into similar metabolic groups based on lifestyle, demographic and omics factors that are core to precision nutrition efforts. Although promising, challenges such as long-term adherence, costs and data privacy must be addressed for broader implementation.

Pharmacological interventions

Pharmacological treatments for obesity and T2DM, including GLP1 receptor agonists (GLP1RAs), metformin and sodium–glucose co-transporter 2 (SGLT2) inhibitors, promote weight loss and improve glycaemic control and might have implications for CRC prevention. Epidemiological data show that GLP1RAs, which are highly effective at inducing weight loss and glucose regulation, are associated with a reduced risk of CRC compared with other pharmacological regimens, such as insulin, metformin, SGLT2 inhibitors and other antidiabetics, among patients with T2DM, overweight or obesity¹⁶⁹. However, whether this association reflects causality or extends to EOCRC specifically remains unclear. There is limited evidence regarding whether GLP1RAs act more effectively in younger individuals than in older ones, or whether early intervention confers increased long-term cancer protection, considering that such interventions offer enhanced benefits by targeting prolonged exposure to metabolic dysfunction. A 12-month, phase III, randomized controlled trial showed that, at a low dose, metformin reduced the risk of metachronous adenomas by 40% in Japanese adults without T2DM who had previously had colorectal adenomas or polyps resected¹⁷⁰. By contrast, a phase IIa trial in 32 adults in the USA with obesity and colorectal adenomas found that a 12-week metformin treatment regimen did not alter the levels of rectal mucosa biomarkers such as phosphorylated ribosomal protein S6 or Ki-67, which can be used to assess CRC progression¹⁷¹. These discordant results could reflect differences in treatment duration, study populations or end points. Of note, metformin induces only modest weight loss and insulin suppression in comparison with GLP1RAs^172,173. Although these findings are promising, further mechanistic studies and well-powered randomized controlled trials, particularly in young, at-risk populations, are needed to clarify the role of these pharmacological agents in the prevention of EOCRC.

Anti-inflammatory drugs, such as aspirin, might also prevent EOCRC by reducing inflammation associated with metabolic disorders¹⁷⁴. Aspirin inhibits prostaglandin synthase 2, lowering the levels of prostaglandin E2, which promotes inflammation, proliferation and tumorigenesis, while also modulating Wnt, cAMP–PKA and β-catenin signalling pathways¹⁷⁴. Although data supporting the use of aspirin for the prevention of CRC in young adults are limited, a case–control study of veterans in the USA found that aspirin users, particularly those aged 40–49 years, had a lower risk of EOCRC compared with non-users¹⁷⁵. From an implementation standpoint for the prevention of EOCRC, initiating an aspirin regimen for young adults could be attractive, given that costs are low and established side effects, such as gastrointestinal bleeding, are strongly related to older age.

Integration into public health policies

Public health policies targeting the root causes of obesity and T2DM are key to curbing the increasing incidence of EOCRC. Sugary drink taxes have been associated with a 10–20% reduction in SSB consumption, a known driver of obesity and T2DM, with cost-effectiveness analyses suggesting greater benefits for younger populations than for older populations^176–179. By addressing notable dietary contributors to metabolic disease, such measures could have downstream effects in lowering EOCRC incidence in groups that are deemed to be ‘at risk’. However, the success of these measures depends on comprehensive implementation, public acceptance and overcoming industry opposition and socioeconomic barriers. Collaborative efforts among researchers, clinicians, policy-makers and consumers are essential to optimizing these population-level prevention strategies.

Enhanced food labelling, particularly for UPFs, SSBs and junk foods, represents another pivotal approach with benefits extending beyond reducing the risk of EOCRC. Clear front-of-package labels indicating excessive sugars, fats and sodium have been shown to encourage healthy dietary choices¹⁸⁰. A meta-analysis of 60 studies across 11 countries found that food labelling led to a 6.6–13.0% reduction in total energy and fat intake and a 13.5% increase in vegetable consumption¹⁸¹. These policies could help young populations to adopt healthy diets, potentially lowering their metabolic risk and reducing EOCRC incidence. However, the effectiveness of labelling depends on consumer literacy and the regulation of misleading marketing practices. Ensuring that these approaches and interventions reach vulnerable and low-income communities, where obesity and T2DM are more prevalent, is essential for a broad public health impact.

Importantly, public health policies that address risk factors that are shared across multiple chronic diseases, rather than focusing solely on EOCRC, will maximize the public health impact. Promoting healthy diets and reducing metabolic dysfunction could contribute to the prevention of obesity, T2DM, cardiovascular disease and various cancers, supporting long-term population health.

Research gaps

Several key research gaps exist. Most mechanistic insights are extrapolated from studies of late-onset CRC and might not capture the unique biology of EOCRC. It is unclear whether inflammatory cascades, immune responses and altered gut microbiome are more pronounced in younger individuals with obesity or T2DM than in their older counterparts. Comprehensive profiling of gut microbial communities and functional pathways specific to EOCRC as well as understanding of potential age-related differences in inflammatory drivers are lacking, and yet are essential for identifying relevant mechanisms and therapeutic targets for EOCRC prevention and treatment.

The role of Western dietary patterns in EOCRC remains underexplored. Existing studies often lack integration of meal timing, food processing and cumulative dietary effects. Findings for individual foods or food groups remain heterogeneous, partly owing to reliance on case–control studies with limited statistical power and residual confounding, variation in population characteristics, and differences in dietary assessment methods. Red and processed meats — well-known dietary risk factors for CRC — have not been adequately examined in EOCRC-specific studies. However, their plausible biological mechanisms and associations with late-onset CRC justify further investigation of their role in younger populations.

Further research should clarify how specific dietary components, gut microbiota and host metabolism interact in EOCRC development. The mechanisms by which UPFs and SSBs contribute to EOCRC warrant further investigation. Emerging evidence suggests that dietary factors like coffee, dark chocolate and yoghurt might support weight management and blood glucose control and reduce the risk of CRC^182–191; whether these benefits extend to EOCRC requires further exploration. Systematic investigations are needed to understand the dynamic interactions between dietary factors, microbial metabolites (for example, SCFAs, secondary bile acids and hydrogen sulfide) and host metabolic responses in young populations at risk for EOCRC. Although high-fat diets are associated with certain molecular subtypes of CRC, studies exploring such associations in EOCRC are scant. Similarly, the interaction between genetic predispositions and diet-induced obesity and T2DM is underexamined. Moreover, early-life exposures might influence metabolic health trajectories related to EOCRC, yet data on these exposures remain limited. Research on socioeconomic variations in metabolic health and EOCRC is also sparse.

Environmental exposures, such as endocrine-disrupting chemicals, including bisphenol A, phthalates and other persistent organic pollutants, might also contribute to EOCRC¹⁹². These compounds can interfere with hormonal balance, metabolic signalling and gut microbial ecology. Integrated studies across environmental, metabolic and microbial domains are warranted.

Screening can reduce CRC incidence and mortality by enabling the removal of precancerous lesions before malignant transformation or the early detection and treatment of CRC. However, evidence on optimal screening modalities for young populations at high risk is limited. Although lifestyle and pharmacological interventions show promise for managing obesity and T2DM, their efficacy in reducing EOCRC risk is not well established. Long-term adherence, cost-effectiveness and accessibility of personalized, technology-driven interventions require further study. Mechanistic studies and randomized controlled trials are needed to validate the role of pharmacological agents in EOCRC prevention. Public health measures should be evaluated for scalability, sustainability and impact across populations.

Addressing these gaps is essential to advancing our understanding of the pathophysiological mechanisms underlying EOCRC and for the development of targeted, evidence-based prevention strategies. Integrating multi-omics approaches, microbiome science and targeted interventions will be pivotal in achieving these goals.

Future directions

Tackling the rising burden of EOCRC requires multifaceted strategies that combine cutting-edge research, targeted interventions and scalable policy efforts focused on three main priorities. First, longitudinal cohort studies that follow individuals from infancy through mid-adulthood are needed to capture exposures during critical windows when metabolic alterations, dietary patterns and gut microbiota changes might influence the risk of EOCRC. Second, advances in characterizing exposomes, such as profiling of the microbiome, metabolome and proteome, can identify novel risk factors for EOCRC and their interactions with metabolic dysfunction. Whole-genome sequencing, including that of normal tissue, will be useful in identifying and validating mutational signatures, such as the ones caused by the mutagen colibactin. Third, mechanistic studies are needed to elucidate the biological pathways that link obesity, T2DM and EOCRC. Future studies integrating microbiome, epigenomic and metabolic phenotyping analyses in both biomarker-driven human studies and animal models are urgently needed. These efforts require interdisciplinary initiatives, such as the Cancer Grand Challenges Team PROSPECT, launched in 2024, which is dedicated to understanding and reversing the causes of EOCRC, with a central focus on metabolic dysfunction.

Taken together, these research directions could inform precision medicine frameworks to identify high-risk subgroups for targeted interventions. Risk prediction models that incorporate metabolic, microbiome, dietary and genetic data might enhance early detection and risk stratification. Personalized approaches that integrate multi-omics techniques and biomarkers could further identify individuals at risk and tailor interventions accordingly. In the long term, longitudinal studies and randomized controlled trials evaluating multimodal interventions, including personalized diet, physical activity and pharmacological agents (for example, GLP1RAs, metformin and aspirin), will be needed. Health economics studies to evaluate the long-term effect of public health policies on mitigating the burdens of metabolic diseases and EOCRC, including in low-income and middle-income countries, will also be critical. Finally, continued attention to social determinants, including access to high-quality foods, safe environments and preventive health care, that influence both metabolic health and EOCRC risk is needed to alleviate health disparities.

Conclusions

The alarming rise in cases of EOCRC since the 1990s, particularly in high-income countries, parallels increases in the incidence of obesity and T2DM among young adults and represents a growing public health challenge. Metabolic and endocrine disruptions, especially those resulting in obesity or T2DM, might have critical roles in EOCRC development. Dysregulated insulin signalling promotes tumour growth via the PI3K–AKT–mTOR and RAS–MAPK signalling pathways. Chronic low-grade inflammation, driven by adipokines and pro-inflammatory cytokines, is another key contributor, with visceral adipose tissue exacerbating inflammation, oxidative stress and DNA damage, thus promoting carcinogenesis. Alterations in the gut microbiota, including reduced levels of beneficial bacteria and increased numbers of pathogenic species, such as colibactin-producing bacteria, could contribute to EOCRC by disrupting metabolic homeostasis, promoting insulin resistance, triggering chronic inflammation and oncogenic signalling, and suppressing immune responses. Dietary shifts towards Western patterns characterized by energy-dense UPFs might link obesity and T2DM with EOCRC via the aforementioned mechanisms, whereas the Mediterranean diet and fibre-rich foods enhance gut health and reduce inflammation. Long-term exposure to a detrimental diet beginning in early life is likely to shape EOCRC risk trajectories.

Strategies to prevent EOCRC merit further investigation. Current age-based CRC screening guidelines might overlook a considerable proportion of EOCRC cases. Integrating metabolic, lifestyle and microbial biomarkers into risk models could improve early detection. Tailored lifestyle interventions, including diet and physical activity, remain effective for managing obesity and T2DM. Technology-based approaches (for example, wearable devices and artificial intelligence-powered personalized nutrition) hold promise but face challenges related to cost and adherence. Pharmacological agents like metformin and aspirin show potential for EOCRC prevention, with GLP1RAs gaining attention for their effectiveness in managing obesity and T2DM. However, the long-term effects of these pharmacological agents, particularly GLP1RAs, on EOCRC risk remain unclear, necessitating rigorous studies to confirm efficacy and elucidate mechanisms. Public health policies, such as sugary drink taxes and food labelling, could help to reduce the risk of metabolic disease but require consistent implementation for broad impact.

By addressing modifiable risk factors and leveraging clinical and public health strategies, it should be possible to reduce the burden of EOCRC and improve outcomes for young populations. Further research is needed to clarify the complex interplay between metabolic health and EOCRC, ultimately enabling the development of evidence-based strategies for its early detection and prevention.

Supplementary Material

Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41574-025-01159-z.

Key points.

• The incidence of early-onset colorectal cancer (EOCRC) is increasing globally, particularly in high-income countries, aligning with rising obesity and type 2 diabetes mellitus (T2DM) rates among young adults.

• Obesity and T2DM could contribute to EOCRC through insulin resistance and/or hyperinsulinaemia, chronic inflammation and altered gut microbiome, which create a pro-tumorigenic environment that accelerates colorectal carcinogenesis.

• Western diets high in ultra-processed foods, refined sugars and saturated fats exacerbate obesity, T2DM and EOCRC risk by disrupting insulin signalling, promoting chronic inflammation and altering the gut microbiome.

• Current colorectal cancer screening guidelines might miss young adults at high risk; integrating metabolic factors, microbiome biomarkers and lifestyle data could improve individual risk assessments for EOCRC.

• Tailored dietary programmes, physical activity, glucagon-like peptide 1 receptor agonists and metformin show potential for reducing EOCRC risk, although long-term efficacy and mechanisms require further study.

• Multi-omics profiling, mechanistic studies and randomized controlled trials focusing on age-specific metabolic pathways, microbial signatures and socioenvironmental factors are essential for targeted EOCRC prevention.