Family History and Colorectal Cancer Risk in IBD: Should the Number of Affected Relatives Change Surveillance?

Two affected relatives or one early-onset cancer: Which history matters more?

Consider two patients with longstanding ulcerative colitis undergoing colorectal cancer surveillance.

The first has a father diagnosed with colorectal cancer at 46 years. The second has a mother and a brother who developed colorectal cancer after age 50.

Which patient should be considered at greater risk, assuming their other IBD-related characteristics are comparable?

Existing IBD surveillance frameworks give particular attention to colorectal cancer in a first-degree relative diagnosed at a young age. Yet the number of affected relatives may contain additional prognostic information that is not fully captured by age at diagnosis alone.

A nationwide Swedish cohort study published in Gastroenterology on 3 September 2026 provides important evidence addressing this question.

Investigators followed more than 124,000 patients with inflammatory bowel disease (IBD), comparing their colorectal cancer incidence with that of more than 1.2 million matched individuals from the general population.

The most pronounced increase in colorectal cancer incidence occurred among patients with IBD who had at least two first-degree relatives affected by colorectal cancer.

By contrast, the observed increase among patients with a relative diagnosed before age 50 was smaller and did not reach statistical significance.

These findings raise a clinically relevant question: should the number of affected relatives receive greater emphasis in IBD surveillance risk stratification?

The answer requires careful interpretation of absolute risk, relative risk and the limitations of observational evidence.

How Swedish nationwide registers were used to examine familial cancer risk

Åsa H. Everhov and colleagues conducted a nationwide, register-based matched cohort study using Swedish healthcare and population data from 1996 through 2023.

The investigators identified 124,387 patients with IBD and compared them with 1,213,641 individuals from the general population.

Each patient with IBD was matched with up to ten comparators based on age, sex, geographical area and calendar year.

The IBD cohort comprised 74,122 patients with ulcerative colitis, 41,062 with Crohn's disease and 9,203 with IBD unclassified at diagnosis.

The median age at IBD diagnosis was 41 years, approximately 49% were women, and median follow-up was 11 years.

Patients with previous colorectal cancer or colectomy were excluded.

Family history was defined using colorectal cancer diagnoses among first-degree relatives: parents, siblings or children.

Two dimensions of familial risk were evaluated:

• The number of first-degree relatives diagnosed with colorectal cancer.

• Whether a first-degree relative developed colorectal cancer before age 50.

Importantly, family history was updated during follow-up as relatives received new cancer diagnoses.

The primary outcome was incident colorectal cancer, identified through nationwide cancer registration or colorectal cancer recorded as a cause of death.

This design enabled researchers to compare cancer incidence across different family-history categories within IBD and against matched general-population groups.

However, family history was an observed exposure, not a randomized intervention. The study could establish associations but could not demonstrate that any particular familial characteristic directly caused colorectal cancer.

The underlying cancer burden in IBD

During follow-up, 1,882 patients with IBD developed colorectal cancer.

This corresponded to an incidence rate of 1.17 cases per 1,000 person-years.

Among matched general-population comparators, 14,177 colorectal cancers occurred, producing an incidence rate of 0.88 cases per 1,000 person-years.

Approximately 6% of individuals in each cohort had at least one first-degree relative diagnosed with colorectal cancer.

These figures provide essential context.

The overall colorectal cancer incidence was higher in the IBD cohort than in the matched general population, but the absolute difference was considerably smaller than the relative contrasts sometimes emphasized in discussions of cancer risk.

An incidence rate describes how frequently new events occur over accumulated observation time. It is not equivalent to an individual's lifetime probability of developing cancer.

For example, 1.17 events per 1,000 person-years represents a population incidence rate, not a prediction that every patient has a fixed 0.117% annual probability regardless of age, disease characteristics or duration.

This distinction becomes particularly important when interpreting family-history subgroups.

Multiple affected relatives identified the highest-risk group

The strongest finding concerned patients with at least two first-degree relatives who had colorectal cancer.

Among 406 patients with IBD in this category, colorectal cancer occurred at a rate of 3.84 cases per 1,000 person-years.

For patients with IBD and no affected first-degree relative, the corresponding rate was 1.15 cases per 1,000 person-years.

The difference was clinically substantial.

Colorectal cancer (CRC) incidence among patients with inflammatory bowel disease (IBD) varied according to family history, particularly the number of affected first-degree relatives.

Patients with no affected first-degree relatives had a CRC incidence of 1.15 cases per 1,000 person-years.

In contrast, those with at least two affected first-degree relatives experienced a substantially higher incidence of 3.84 cases per 1,000 person-years, representing more than three times the observed rate among patients without an affected first-degree relative.

Patients with at least one relative diagnosed with CRC before age 50 had an incidence of 1.57 cases per 1,000 person-years.

These findings suggest that family history, particularly having multiple affected first-degree relatives, may be an important consideration in CRC risk assessment among patients with IBD. However, these incidence differences alone do not establish causality or justify changes to surveillance intervals without considering other established risk factors and guideline recommendations.

Family history in patients with IBDCRC incidence per 1,000 person-yearsNo affected first-degree relative1.15At least two affected first-degree relatives3.84At least one relative diagnosed before age 501.57

Patients with at least two affected relatives had approximately 2.59 times the colorectal cancer incidence rate of those without a family history.

The absolute incidence-rate difference was 2.69 additional cases per 1,000 person-years (95% confidence interval [CI] 0.60–4.78).

The confidence interval excluded zero, supporting a statistically detectable increase in absolute incidence.

The finding suggests that familial cancer burden may contribute important information beyond the presence or absence of any affected relative.

Nevertheless, the subgroup contained relatively few patients, and its estimate was less precise than the overall cohort findings.

The study does not establish that the optimal clinical threshold is exactly two relatives, nor does it determine an appropriate colonoscopy interval for this group.

It identifies a population in which the observed cancer burden warrants further investigation.

Why an early-onset family history produced a less pronounced signal

Current IBD surveillance recommendations place particular emphasis on a first-degree relative diagnosed with colorectal cancer before age 50.

The Swedish analysis examined this characteristic directly.

Patients with IBD and at least one first-degree relative diagnosed before age 50 had a colorectal cancer incidence rate of 1.57 cases per 1,000 person-years.

Compared with patients without affected first-degree relatives, the absolute difference was 0.42 additional cases per 1,000 person-years.

The 95% confidence interval ranged from −0.47 to 1.31, crossing zero.

The corresponding relative rate was approximately 1.47, but this difference was not statistically significant.

Only 12 colorectal cancer events occurred in this family-history subgroup.

This is an important limitation.

Failure to demonstrate statistical significance does not establish the absence of risk, particularly when event numbers are small.

Nor does it prove that early-onset cancer in a first-degree relative is irrelevant to surveillance planning.

The investigators suggested several potential explanations for the smaller-than-expected association, including differences between contemporary nationwide populations and older specialist hospital cohorts, improved IBD management and changing colorectal cancer incidence over time.

These explanations remain plausible interpretations rather than mechanisms directly established by the analysis.

The evidence therefore questions how familial risk should be weighted, but does not justify abandoning established risk criteria.

Absolute risk and relative risk tell different parts of the story

One of the study's most informative findings emerged when patients with IBD were compared with general-population individuals who had similar family histories.

The investigators reported that family history was associated with a broadly similar absolute increase in colorectal cancer incidence in both populations.

Yet the relative association was weaker among patients with IBD.

For any family history, the reported relative association was approximately 1.31 in IBD, compared with 2.09 among general-population comparators.

This difference is understandable because the baseline colorectal cancer incidence was already higher in the IBD population.

A similar absolute increase can correspond to a smaller relative increase when the starting incidence is higher.

The comparison should not be interpreted as evidence that family history matters less clinically in patients with IBD.

Indeed, the absolute incidence among patients with multiple affected relatives was the highest observed within the IBD family-history categories.

The investigators also reported that colorectal cancer incidence was comparable between patients with IBD and general-population individuals sharing the same family-history category, except among those without affected first-degree relatives.

In the absence of family history, IBD was associated with approximately 1.41 times the colorectal cancer incidence of matched comparators, with the excess largely driven by ulcerative colitis.

These findings illustrate why absolute incidence and relative measures should both be considered when evaluating surveillance strategies.

Relative measures describe the strength of association. Absolute measures better convey the observed additional burden of disease.

Neither, on its own, establishes how frequently colonoscopy should be performed.

Was the association consistent in ulcerative colitis and Crohn's disease?

The researchers examined the two major IBD subtypes separately.

The highest colorectal cancer incidence occurred among patients with at least two affected first-degree relatives in both ulcerative colitis and Crohn's disease.

For ulcerative colitis, the incidence rate in this category was 4.08 cases per 1,000 person-years.

For Crohn's disease, the corresponding rate was 3.15 cases per 1,000 person-years.

However, the Crohn's disease estimate was based on only three colorectal cancer events.

Only the increase observed in the ulcerative colitis subgroup reached statistical significance.

This distinction matters.

A consistent numerical pattern across disease subtypes provides supportive evidence that the association is not confined to one form of IBD.

However, sparse events prevent confident conclusions about whether familial risk operates similarly in ulcerative colitis and Crohn's disease.

The study was not designed to determine a different family-history-based surveillance interval for each subtype.

Nor does the analysis establish that the underlying biological relationship between familial susceptibility and colorectal carcinogenesis is identical across IBD phenotypes.

The most defensible interpretation is that multiple affected relatives were associated with the highest observed cancer incidence in both major IBD groups, with considerably greater statistical uncertainty in Crohn's disease.

Sensitivity analyses supported the pattern, but did not resolve uncertainty

A major strength of the study was its use of several sensitivity analyses to examine whether the principal findings depended on particular assumptions.

The investigators repeated analyses after excluding patients with possible Lynch syndrome, censoring follow-up at diagnosis of primary sclerosing cholangitis, incorporating high-grade dysplasia into relevant definitions and excluding the first year of follow-up.

The overall pattern remained: patients with multiple affected first-degree relatives generally had the highest colorectal cancer incidence among the family-history categories.

This consistency reduces concern that the main result was entirely attributable to one specific analytical decision.

However, sensitivity analyses do not eliminate confounding or compensate for small event numbers.

When investigators restricted the analysis to patients eligible for IBD surveillance, the rate ratio for those with at least two affected relatives remained the highest, but the increase was not statistically significant.

This finding is especially important for clinical interpretation because surveillance-eligible patients are the population most directly relevant to colonoscopy scheduling.

The available evidence supports further evaluation of familial burden as a surveillance risk factor, but not immediate implementation of a new surveillance algorithm.

What nationwide data cannot fully capture

The study's population size, lengthy follow-up and linkage of healthcare and family records are substantial strengths.

Unlike analyses restricted to tertiary referral centres, nationwide registry research can include patients with a wider range of disease severity and clinical characteristics.

Nevertheless, several limitations require attention.

First, the registers may not fully characterize the anatomical extent or location of IBD. Disease extent is relevant because colorectal cancer risk differs across inflammatory disease phenotypes.

Second, endoscopy data were incomplete.

The investigators could not consistently account for differences in colonoscopy frequency, surveillance quality or detection practices between comparison groups.

More frequent colonoscopy could affect the identification of cancer, while effective surveillance might prevent some cancers through detection and treatment of precancerous lesions.

Third, family history can be underestimated in younger patients whose relatives have not yet reached ages at which colorectal cancer commonly develops.

Fourth, national colorectal cancer screening was not fully implemented throughout Sweden during the study period. Differences in screening exposure could influence observed incidence.

Finally, the observational design cannot isolate familial genetic susceptibility from shared environmental factors, differences in healthcare use or other unmeasured influences.

Although the study included sensitivity analyses addressing possible Lynch syndrome, it did not provide a comprehensive germline genetic characterization of the cohort.

The findings should therefore not be interpreted as proof of a particular inherited mechanism.

Should the number of affected relatives influence surveillance decisions now?

The study raises a compelling question about how colorectal cancer surveillance strategies could become more individualized.

A family history is not simply positive or negative.

The number of affected first-degree relatives, their ages at diagnosis and the patient's other IBD-associated risk factors may collectively influence the clinical assessment.

From a practical perspective, the findings strengthen the rationale for documenting family history in greater detail and updating it over time.

A patient who initially reports no affected first-degree relatives may acquire a relevant family history during long-term follow-up.

However, recording more detailed risk information is different from changing surveillance intervals.

The investigators did not test whether more frequent colonoscopy in patients with multiple affected relatives improves clinical outcomes.

Equally, the study did not demonstrate that patients with a single relative diagnosed before age 50 can safely undergo less frequent surveillance.

In accompanying comments, the first author suggested that future guidelines consider giving greater priority to patients with multiple affected relatives, while explicitly cautioning against immediate de-escalation of surveillance in other groups.

Surveillance decisions must continue to consider disease extent, duration, primary sclerosing cholangitis, previous dysplasia and other established risk factors.

The central implication is therefore one of improved risk characterization, not an independently validated change in colonoscopy scheduling.

Building the next generation of individualized IBD surveillance

The findings identify several research priorities.

Future studies should examine whether incorporating the number of affected relatives improves colorectal cancer risk prediction beyond existing IBD surveillance criteria.

This requires validation in independent populations, including settings with different colorectal cancer screening practices, healthcare systems and ethnic or genetic backgrounds.

Larger studies are particularly important for patients with multiple affected relatives and for those with early-onset colorectal cancer in the family, because relatively few cancer events occurred in these subgroups.

More complete information about inflammatory burden, surveillance colonoscopy, dysplasia history and treatment exposure could improve understanding of how familial risk interacts with IBD-specific factors.

Studies should also evaluate whether revised risk-stratification models improve decisions that matter clinically: appropriate surveillance intensity, cancer detection, resource use and patient outcomes.

Without such evidence, statistically detectable differences in incidence cannot be assumed to translate directly into improved surveillance strategies.

The next step is not merely to reproduce the observed association, but to establish whether incorporating familial cancer burden produces more accurate and clinically beneficial decisions.

Clinical Takeaway: Family history in IBD is more than a yes-or-no question

This large Swedish cohort found that patients with IBD and at least two first-degree relatives with colorectal cancer experienced the greatest increase in colorectal cancer incidence among the family-history groups studied.

The association was more pronounced than that observed for having a first-degree relative diagnosed before age 50, although the latter subgroup had few events and considerable uncertainty.

The findings provide a strong rationale for evaluating the number of affected relatives as an additional component of individualized colorectal cancer risk assessment in IBD.

However, the study remains observational, and it does not establish new surveillance intervals or identify patients in whom reduced surveillance has been proven safe.

GastroAGI perspective: A detailed, regularly updated family history should be regarded as an important element of IBD cancer risk assessment. Multiple affected first-degree relatives may identify a subgroup deserving particular attention in future surveillance models, but changes to colonoscopy frequency require prospective validation and integration with established IBD-related risk factors.

The important question for future research is whether a more nuanced interpretation of family history can improve surveillance without increasing unnecessary procedures or leaving high-risk patients inadequately protected.

Primary Reference

Everhov ÅH, Kristjánsson K, Ludvigsson JF, Halfvarson J, Backman AS, Myrelid P, et al. Impact of Family History on Colorectal Cancer Risk in Inflammatory Bowel Disease and in Matched General Population Comparators. Gastroenterology. Published online 3 September 2026.

DOI: https://doi.org/10.1053/j.gastro.2026.08.029

Supporting approved source

Brunk D. Multiple affected relatives linked to higher CRC risk in IBD. GI & Hepatology News. 9 October 2026.

https://news.gastro.org/issues/2026/october-2026/multiple-affected-relatives-linked-to-higher-crc-risk-in-ibd/