Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Sep 10.
Published before final editing as: Cancer Epidemiol Biomarkers Prev. 2026 Aug 11:10.1158/1055-9965.EPI-25-1658. doi: 10.1158/1055-9965.EPI-25-1658

Association of high birth weight with risk of early-onset colorectal cancer

Amrita Mukherjee 1, Lanfang Xu 2, Darios Getahun 1,6, Jessica Chubak 3, Jane C Figueiredo 4, Kimberly L Cannavale 1, Alec D Gilfillan 1, Bechien U Wu 1,5, Chun R Chao 1,6
PMCID: PMC13557039  NIHMSID: NIHMS2205062  PMID: 42578973

Abstract

Background:

Early-onset colorectal cancer (eoCRC), particularly rectal cancer, is increasing. We evaluated the association between high birth weight and eoCRC in a nested case-control study.

Methods:

We included patients aged 15–49 years diagnosed with colorectal adenocarcinoma at Kaiser Permanente Southern California (2009–2021). Cancer-free controls were matched 10:1 on age, sex, and length of health plan membership. Record linkage with California Department of Public Health’s birth registry was performed. Conditional logistic regression was used to evaluate associations between high birth weight, measured using (1) macrosomia (birth weight >4,000grams), and (2) large for gestational age (LGA; birth weight above the 90th percentile for gestational age and sex), and overall eoCRC, colon, and rectal cancer. Two-stage model adjustments were performed: first adjusting for pre-specified potential confounders, then adjusting for potential intermediates.

Results:

Of 1,400 eligible cases and matched 13,608 controls, 471 eoCRC cases (mean diagnosis age: 41.1 years) and 1,931 controls with linked birth certificate were included. Adjusted odds ratio (OR) for eoCRC was 1.32 (95% CI: 0.95–1.84) for macrosomia and 1.38 (0.97–1.98) for LGA. Macrosomia and LGA were associated with a marginal, non-statistically significant elevated risk of rectal cancer [OR (95%CI): 1.69 (0.97–2.97) and 1.81 (0.96–3.41), respectively], but not colon cancer [OR (95%CI): 1.18 (0.78–1.80) and 1.20 (0.77–1.86) for macrosomia and LGA, respectively].

Conclusions:

An elevated risk of early-onset rectal cancer was suggested in individuals with high birth weight.

Impact:

The impact of intrauterine conditions on eoCRC risk should be further evaluated.

Keywords: early-onset colorectal cancer, colorectal cancer, birth weight, intrauterine conditions, etiology

Introduction

Colorectal cancer (CRC) is the second most common cause of cancer deaths in the United States (US);(1, 2) 52,900 deaths due to CRC are estimated in 2025.(2) Data from the US Surveillance, Epidemiology, and End Results (SEER) database suggest an overall decrease in CRC incidence and mortality over the past few decades.(3) However, these epidemiologic trends are not homogenous across all age groups.(4, 5) In patients aged 65 years and older, a decrease in CRC incidence rates has been reported since 1975, likely due to widely adopted screening and lifestyle changes.(5, 6) Although incidence rates have been stable in ages 50–64 since 2011, in patients aged <50 years, CRC incidence have been increasing by 2% per year. (4, 5, 7, 8) Patients with early-onset CRC (eoCRC, diagnosis age <50 years) are also more likely to be diagnosed at advanced stages, leading to substantial mortality.(4, 5, 8–10) These concerning trends in eoCRC are not only reported in the US but also in other high-income countries.(11) However, factors contributing to the rise of eoCRC remain to be fully understood.(12)

Associations between intrauterine conditions and health outcomes later in life have been documented in previous studies.(13, 14) Structural and functional changes in cells, tissues, and organ systems due to fetal programming are linked to long-term health consequences and increased susceptibility to certain conditions in adulthood, including hypertension, insulin resistance, dyslipidemia, cardiovascular diseases, and non-insulin-dependent diabetes.(14, 15) Epidemiologic studies have also demonstrated a link between high birth weight and risk of certain cancers in adolescent and young adults, including young-onset testicular cancer and pre-menopausal breast cancer,(16–21) and to a lesser degree, in older adults.(18, 22–25) Intrauterine programming of key hormonal axes, such as the hypothalamic-pituitary-adrenal axis and the insulin-like growth factor system is one of the plausible biological mechanisms behind increased carcinogenesis in individuals with high birth weight (Figure 1).(26, 27) High birth weight is also correlated with a greater reservoir of stem cells at risk for carcinogenesis (Figure 1).(28, 29) However, the association between birth weight and risk of eoCRC remains relatively unexplored.

Figure 1:

Figure 1:

Open in a new tab

Hypothesized model for association between high birth weight and early onset colorectal cancer (eoCRC). Conceptual model shows the potential causal pathways justifying our model selection in assessing the association between high birth weight and eoCRC.

*: We used race/ethnicity as a proxy for racism.

The average birth weight has been increasing in the US since the 1950 birth cohort,(30) possibly due, in part, to the increasing prevalence of obesity and diabetes among pregnant women.(31) Thus, the timing of the increase in birth weight meets the expected pattern for it to be a possible, partial explanation for the increase in the incidence of eoCRC, thereby warranting investigation. Our objective was to assess the association between high birth weight and risk of eoCRC.

Materials and Methods

Study design, setting and population

This nested case control study was conducted at Kaiser Permanente Southern California (KPSC), an integrated health care delivery system that serves more than 4.6 million racially/ethnically and socioeconomically diverse members who are broadly representative of the population in southern California.(32, 33) KPSC members diagnosed with in situ and invasive CRC between 2009 and 2021 at ages 15 to 49 years were eligible for inclusion. eoCRC cases were identified through KPSC’s SEER-affiliated cancer registry, except for cases diagnosed in 2021, which were identified using International Classification of Diseases, 10th Revision (ICD-10) codes and subsequently confirmed via chart review. Cases who had: (1) < two years of KPSC membership prior to diagnosis, (2) another cancer diagnosis (except non-melanoma skin cancer), HIV infection, or transplant history prior to the CRC diagnosis, (3) familial adenomatous polyposis (FAP), Lynch syndrome, or Peutz–Jeghers syndrome, or (4) histology other than adenocarcinoma (e.g., neuroendocrine or squamous cell histology) were excluded from the analysis. Patients with fewer than two years of KPSC membership were excluded because we assessed status of comorbidities and health behaviors from at least two years prior to CRC diagnosis/index date to reduce the possibility of protopathic bias.

Cases were matched at a 1:10 ratio to control subjects selected using incidence density sampling, based on exact age, sex, and length of KPSC membership (within 1 year, to ensure comparable length of available clinical history). Each control was assigned an index date which was the diagnosis date of their matched case. Controls were considered eligible if they were KPSC members on the reference data and did not meet exclusion criteria (1) – (3) described above for the cases. The primary analysis was limited to individuals whose birth certificate records were successfully identified via record linkage with the California Department of Public Health (CDPH) birth data (procedure described below). This study was approved by the KPSC’s Institutional Review Board; the requirement for informed consent was waived.

Data collection and variables of interest

Eligible cases and controls were linked to CDPH birth records from 1960 to 2005 to obtain information on birth characteristics of interest, including birth weight, gestational age, gestational length, race/ethnicity, sex at birth, age of mother, age of father, and pregnancy complications (available from 1978 onward) such as gestational diabetes and preeclampsia. Linkage of CDPH birth records to the study population was preformed using date of birth, sex, first name, last name, and middle name. Detailed information on linkage to CDPH birth records and the multi-stage matching process are presented in Figure 2 and in the Supplementary Methods and Materials file.

Figure 2:

Figure 2:

Open in a new tab

Flowchart illustrating linkage of Kaiser Permanente Southern California electronic health records (EHR) to California Department of Public Health’s (CDPH) birth registry. Study flowchart shows the multi-stage process used to link KPSC’s EHR to CDPH birth registry, and matched eoCRC cases and cancer-free controls included in the analysis.

High birth weight was defined using two alternative criteria: (1) neonatal macrosomia, defined as birth weight >4000 g;(34) and (2) large for gestational age (LGA), defined as birth weight above the 90th percentile for gestational age and sex.(35) All babies born at KPSC served as the population reference to calculate birth weight percentiles for both cases and controls. This method of using an internal population reference is preferred and well accepted.(36, 37) We included both macrosomia and LGA to evaluate the impacts of sheer birth weight (an absolute measure) vs. heavy weight given the gestational age (a relative measure). These definitions capture somewhat different individuals who can be considered heavy at birth, although with considerable overlapping.

Data on CRC characteristics, including date of diagnosis, histology, stage at diagnosis and anatomic site were collected from KPSC’s cancer registry. Other covariates of interest included age (matched) at diagnosis/index date, sex (matched), race/ethnicity, length of membership (matched), adult body mass index [BMI, categorized as underweight (<18.5 kg/m2), normal weight (18.5–24.9 kg/m2), overweight (25–29.9 kg/m2), and obese (≥30 kg/m2)], adult height (in sex-specific quartiles), diabetes, hypertension, smoking (ever, never, based on record from at least one encounter), inflammatory bowel disease (IBD), and Medicaid enrollment (ever, never). These data were obtained from KPSC’s EHR. For variables whose values changed over time (e.g., BMI, comorbidities), we used the most recent value recorded more than two years prior to the diagnosis/index date. For BMI, a small proportion of the study population had a value recorded at under age 20 years. For these subjects, we classified BMI based on the children’s BMI-for-age percentiles(38). Missing race/ethnicity data from the KPSC EHR were supplemented with information from CDPH records. Diabetes, hypertension, and IBD were identified using ICD-9 and ICD-10 diagnosis codes, supplemented by laboratory test results and/or medication use for hypertension and diabetes. Conditions used in the exclusion criteria (i.e., familial adenomatous polyposis (FAP), Lynch syndrome, or Peutz–Jeghers syndrome) were identified using genetic testing results.

Gestational diabetes is recorded in CDPH birth records from 1978–2005. CDPH birth records do not contain information on non-gestational diabetes. From 1978–1988, a single category inclusive of pre-eclampsia, eclampsia, and toxemia was available. From 1989–2005, eclampsia and pre-eclampsia were recorded as separate categories. However, data obtained from 1989–2005 showed extremely low prevalence of these two conditions in the study population (i.e., <1% for gestational diabetes and <2% for pre-eclampsia/eclampsia), suggesting significant under-reporting. As a result, we did not include these two conditions in the multivariable models.

Statistical analysis

The distribution of socio-demographic and clinical characteristics was compared between cases and controls using Chi-square tests. Demographic characteristics were also compared between cases with and without a birth record match; this was repeated among controls as well. Both exposure variables for high birth weight were modeled as dichotomous variables (yes/no). For the regression analysis, we constructed three stages of models: crude analysis using bivariate models (stage 1); multivariable model adjusting for pre-specified potential confounders (stage 2); and multivariable model adjusting for selected potential intermediates (stage 3). For stage 1, bivariate conditional logistic regression was used to evaluate the crude association between high birth weight and eoCRC. Separate models were built for each of the alternative definitions for high birth weight. For stage 2, we adjusted for the following potential confounders based on assumptions in Figure 1: maternal age, paternal age, smoking, Medicaid enrollment, and race and ethnicity. We used race/ethnicity as a proxy for racism (Figure 1). For stage 3, we estimated the association between high birth weight and eoCRC independent of causal pathways potentially mediated through the following factors (Figure 1): adult height, adult BMI, hypertension, and diabetes. We considered the stage 3 models our primary models of interest (rather than the stage 2 models, which estimated the total association between high birth weight and the outcome). We pre-specified that if stage 2 adjustment for confounders changed the adjusted estimate by ≥10% from the crude estimate for high birth weight, we would retain those confounders in the stage 3 model. However, if stage 2 adjustment did not change estimates of association by ≥10%, a more parsimonious model would be used, adjusting for the potential intermediates only to preserve power. Because stage 2 adjustment led to minimal changes to the effect estimates (<10%), suggesting minimal confounding effects by these covariates, we included only the intermediates in the stage 3 models.

In a sensitivity analysis, we excluded individuals with a history of IBD, given the high risk of CRC in IBD patients. Another sensitivity analysis restricted to cases diagnosed with eoCRC at age ≥ 30 years and their matched controls to allow a more consistent interpretation of the BMI variable, which would reflect only adult BMI in this sensitivity analysis rather than adult and adolescent BMI in the main analysis. In analyses of specific CRC anatomical site, we repeated the primary analysis separately for colon and rectal cancer. Family history of CRC was not recognized in the literature as a potential confounder and was not included in the final multivariable models. However, as a validation we performed models adjusting for family history of CRC and it was significantly associated with increased risk of eoCRC, as expected. We used Rubin’s multiple imputations to handle missing data.(39)

In addition, we explored potential effect modification by sex, calendar year at diagnosis, and race/ethnicity. Because sex and calendar year at diagnosis were matching factors, we conducted stratified analysis for males vs. females, and for those diagnosed during 2009–2014 vs. 2015–2021. For race/ethnicity, we evaluated effect modification by including two-way interaction terms with neonatal macrosomia and with LGA in the stage 3 models. All analyses were conducted in SAS statistical software Version 9.4; Cary, North Carolina, USA.

Data Availability:

Anonymized data that support the findings of this study may be made available from the corresponding author on reasonable requests from qualified researchers with documented evidence of training for human subjects protections.

Results

Characteristics of the study population

After applying the inclusion and exclusion criteria, 1,400 cases and their 13,608 matched controls were included for birth record linkage (Supplementary Figure 1). Thirty-five percent of both cases and controls were linked with birth registry data. After excluding controls without linked cases or cases without linked controls, a total of 471 eoCRC cases and 1,931 matched controls remained in matched units and were included in the analysis (Supplementary Figure 1).

EoCRC cases with a birth record match were more likely to be male (62.5% vs. 43.1%), belong to non-Hispanic Black (14.8% vs. 6.8%) and non-Hispanic White (40.0% vs 31.0%) race/ethnicity, and younger at eoCRC diagnosis (35.7% at < age 40 years vs. 19.1% at < age 40 years), compared to cases without a birth record match (Supplementary Table 1). Cancer-free controls with and without a birth record match demonstrated similar patterns (Supplementary Table 1). Table 1 shows the distribution of sociodemographic and clinical characteristics of the included eoCRC cases and their matched controls. The mean age (standard deviation) at diagnosis for CRC cases was 41.1 (6.7) years. Approximately two-thirds of the cases were male (63.7%). Forty percent of the cases were Hispanic, another 40.1% were non-Hispanic White, followed by 14.7% non-Hispanic Black, and 5.1% non-Hispanic Asians/Pacific Islander/Other. About two-thirds of the cases had colon cancer and one third had rectal cancer. About three quarters of eoCRC cases were diagnosed at regional or advanced stages (stages II-IV).

Table 1:

Sociodemographic and clinical characteristics of early-onset colorectal cancer (eoCRC) cases and matched controls, Kaiser Permanente Southern California, 2009–2021

eoCRC cases Controls Total p-value
(n=471) (n=1931) (N=2402)
Age at diagnosis/index date (years), Mean (SD) 41.1 (6.7) 39.0 (7.6) 39.4 (7.5) <0.01e
Sex assigned at birth, n (%) <0.01e
Female 171 (36.3%) 551 (28.5%) 722 (30.1%)
Male 300 (63.7%) 1380 (71.5%) 1680 (69.9%)
Race/Ethnicity, n (%) 0.32
Non-Hispanic Asian/Pacific Islander/Other 24 (5.1%) 107 (5.5%) 131 (5.4%)
Non-Hispanic Black 69 (14.7%) 224 (11.6%) 293 (12.2%)
Hispanic, any race 189 (40.1%) 821 (42.5%) 1010 (42.1%)
Non-Hispanic White 189 (40.1%) 779 (40.3%) 968 (40.3%)
Birth year, n(%)
1960–1969 128 (27.2%) 402 (20.8%) 530 (22.1%) <0.01e
1970–1979 229 (48.2%) 864 (44.7%) 1093 (45.5%)
1980–1989 93 (19.8%) 496 (25.7%) 589 (24.5%)
1990–1998 21 (4.5%) 169 (8.8%) 190 (7.9%)
Neonatal macrosomiaa, n (%) 0.27
No 414 (87.9%) 1731 (89.6%) 2145 (89.3%)
Yes 57 (12.1%) 200 (10.4%) 257 (10.7%)
Large for gestational age (LGA)b, n (%) 0.02
No 349 (88.1%) 1575 (91.9%) 1924 (91.2%)
Yes 47 (11.2%) 139 (8.1%) 186 (8.8%)
Missing 75 217 292
Father’s age at case’s/control’s birth (years), Mean (SD) 29.0 (7.0) 28.6 (6.6) 28.7 (6.7) 0.42
Mother’s age at case’s/control’s birth (years), Mean (SD) 25.6 (5.8) 25.8 (5.6) 25.8 (5.6) 0.33
Body mass indexc, n (%) 0.01
Underweight (<18.5 kg/m2) 1 (0.2%) 13 (0.8%) 14 (0.7%)
Normal weight (18.5–24.9 kg/m2) 69 (16.6%) 376 (21.9%) 445 (20.9%)
Overweight (25–29.9 kg/m2) 123 (29.6%) 561 (32.7%) 684 (32.1%)
Obesity (≥30 kg/m2) 222 (53.5%) 764 (44.6%) 986 (46.3%)
Missing 56 217 273
Hypertensiond, n (%) 0.00
No 380 (80.7%) 1673 (86.6%) 2053 (85.5%)
Yes 91 (19.3%) 258 (13.4%) 349 (14.5%)
Diabetesd, n (%) 0.02
No 418 (88.7%) 1780 (92.2%) 2198 (91.5%)
Yes 53 (11.3%) 151 (7.8%) 204 (8.5%)
Smokingd, n (%) 0.00
Ever 94 (22.8%) 521 (30.4%) 615 (28.9%)
Never 318 (77.2%) 1195 (69.6%) 1513 (71.1%)
Missing 59 215 274
Medicaid insurance, n (%) 0.19
No 443 (94.1%) 1844 (95.5%) 2287 (95.2%)
Yes 28 (5.9%) 87 (4.5%) 115 (4.8%)
Inflammatory bowel disease, n (%) <.01
No 451 (95.8%) 1920 (99.4%) 2371 (98.7%)
Yes 20 (4.2%) 11 (0.6%) 31 (1.3%)
CRC family history, n (%) <.01
No 440 (93.4%) 1891 (97.9%) 2331 (97.0%)
Yes 31 (6.6 %) 40 (2.1%) 71 (3.0%)
Anatomic site of CRC, n (%) NA
Colon 303 (64.33%) NA NA
Rectum & Rectosigmoid 168 (35.67%) NA NA
Stage at diagnosis, n (%)
Stage 0 (in situ) 33 (7.01%) NA NA NA
Stage I/Localized 84 (17.83%) NA NA
Stage II/III/Regional 239 (50.74%) NA NA
Stage IV/Distant 109 (23.14%) NA NA
Unknown/NA/NR 6 (1.27%) NA NA
Length of KPSC membership before diagnosis/index date(years), Mean (SD) 10.6 (8.0) 12.2 (8.9) 11.9 (8.8) <.01e
Open in a new tab

Column percentages are rounded to one decimal place and may not add up to 100%. NA: not applicable; SD: standard deviation.

a

Neonatal macrosomia, defined as birth weight >4000 g;

b

Large for gestational age (LGA), defined as birth weight above the 90th percentile for gestational age and sex;

c

Based on most recent value recorded more than two years prior to the diagnosis/index date.

d

Measured from at least 2 years prior to diagnosis or index date.

e

Distribution appears to be unbalanced due to variation in the number of controls matched to cases across matched sets after CDPH birth record linkage. Within a match set, the distributions of these variables were balanced.

High birth weight and risk of eoCRC

Bivariate and multivariable associations between high birth weight (neonatal macrosomia and LGA) and eoCRC are presented in Table 2. In the bivariate models, odds ratio (OR) estimates for neonatal macrosomia and LGA were 1.31 (95% confidence interval: 0.95–1.81) and 1.38 (0.97–1.98), respectively. Results were similar after adjustment for pre-specified potential confounders (maternal age, paternal age, smoking, Medicaid enrollment, and race/ethnicity), suggesting limited confounding. In the stage 3 models, adjusting for potential intermediate variables on the causal pathway (adult height, adult BMI, diabetes, and hypertension) did not materially change the estimates for the high birth weight variables from those from the stage 1 and 2 models (Table 2). OR estimates for neonatal macrosomia and LGA in stage 3 models were 1.32 (0.95–1.84) and 1.38 (0.97–1.98), respectively.

Table 2:

Bivariate (unadjusted) and multivariable (adjusted) association of high birth weight and eoCRC, according to different definitions of high birth weight

NEONATAL MACROSOMIA LGA
Stage 1: Crude association
Unadjusted
OR (95%CI)		 Unadjusted
OR (95%CI)
High birth weight 1.31 (0.95–1.81) 1.38 (0.97–1.98)
Stage 2: Association adjusted for potential confounders (maternal age, paternal age, smoking, Medicaid enrollment, and race/ethnicity)
Adjusted
OR (95%CI)		 Adjusted
OR (95%CI)
High birth weight 1.33 (0.96–1.85) 1.43 (0.99–2.06)
Stage 3: Association adjusted for potential intermediates (adult height, adult BMI, hypertension, and diabetes)
Adjusted
OR (95% CI)		 Adjusted
OR (95% CI)
High birth weight 1.32 (0.95–1.84) 1.38 (0.97–1.98)
Open in a new tab

OR (95%CI): Odds Ratio (95% Confidence Interval). Neonatal macrosomia, defined as birth weight >4000 g;

Large for gestational age (LGA), defined as birth weight above the 90th percentile for gestational age and sex.

In a sensitivity analysis restricted to patients without a history of IBD (cases=451, controls=1,812), the results remained similar to the main findings (Table 3a). In a second sensitivity analysis focused on anatomic site of eoCRC, high birth weight was not associated with early-onset colon cancer (n=303 for cases and 1,232 for controls) in any model (Table 3b). On the other hand, neonatal macrosomia and LGA were positively associated with early-onset rectal cancer [adjusted OR : 1.69 (0.97–2.97) and 1.81 (0.96–3.41), respectively] (Table 3b). These associations were, however, not statistically significant. In another sensitivity analysis restricted to patients aged ≥30 years at diagnosis, the overall findings remained similar to the main findings (Table 3c). Macrosomia and LGA were associated with an elevated risk of early-onset rectal cancer [adjusted OR: 1.84 (1.02–3.32) and 1.99 (1.01–3.90), respectively].

Table 3:

Sensitivity analysis for the adjusted association between high birthweight and eoCRC risk and analysis by eoCRC site

3A: EXCLUDING IBD PATIENTS 3B: BY eoCRC ANATOMIC SITE
eoCRC
(Cases=451. Controls=1812)		 Colon cancer
(Cases=303; Controls=1232)		 Rectal cancer
(Cases=168. Controls=699)
Adjusted
OR (95%CI)		 Adjusted
OR (95%CI)		 Adjusted
OR (95%CI)
Neonatal macrosomia * 1.31 (0.93–1.85) 1.18 (0.78–1.80) 1.69 (0.97–2.97)
LGA * 1.39 (0.96–2.02) 1.20 (0.77–1.86) 1.81 (0.96–3.41)
3C: PATIENTS AGED ≥30 YEARS AT DIAGNOSIS
eoCRC
(Cases=438. Controls=1689)		 Colon cancer
(Cases=283; Controls=1085)		 Rectal cancer
(Cases=155. Controls=604)
Neonatal macrosomia * 1.33 (0.95–1.88) 1.16 (0.75–1.79) 1.84 (1.02–3.32)
LGA * 1.48 (1.01–2.15) 1.26 (0.80–1.98) 1.99 (1.01–3.90)
Open in a new tab

OR (95%CI): Odds Ratio (95% Confidence Interval). Neonatal macrosomia, defined as birth weight >4000 g;

Large for gestational age (LGA), defined as birth weight above the 90th percentile for gestational age and sex.

*

Adjusted for adult height, BMI, hypertension, and diabetes.

In the stratified analysis by sex and calendar year of diagnosis, OR estimates were similar across subgroups (Supplementary Table 2). No effect modification by race/ethnicity was suggested by the evaluation of the interaction terms (interaction term p-value = 0.35 and 0.25 in the neonatal macrosomia and LGA model, respectively).

Discussion

In a racially and ethnically diverse population in southern California, a possible higher odds of eoCRC risk among those with high birth weight was suggested, although the confidence intervals included null. The association estimates for high birth weight did not change materially when accounting for adult height and metabolic risk factors like obesity, diabetes, and hypertension in the analysis. When colon and rectal cancers were examined separately, our finding suggested that high birth weight was likely associated with a higher risk of early-onset rectal cancer, but not early-onset colon cancer.

Our results differ somewhat from prior studies. A prospective cohort study using data from the Danish health registers suggested a positive association between high birth weight and colon cancer;(40) however, no association between high birth weight and rectal cancer risk was reported.(40) Positive associations between high birth weight and risk of CRC, as well as colon cancer were also reported in the Women’s Health Initiative Observational Study (WHI-OS) cohort.(41, 42) Findings from the ‘Etude Epidémiologique auprès des Femmes de la Mutuelle Générale de l’Education Nationale’ (E3N) cohort similarly suggest a positive association between high birth weight and CRC.(43) In contrast, a couple of other studies reported no association between high birth weight and CRC risk.(44, 45) In our study, there was a suggestion of positive association between high birth weight and overall eoCRC, although the confidence interval included the null. In contrast to prior studies,(40, 41) a positive association with early-onset rectal cancer was suggested in our study, especially among those diagnosed at age 30 years or older. It is possible that the association between high birth weight and average onset CRC may not be the same as that with eoCRC; however, we also cannot rule out the possibility of a chance finding given the limited number of cases in the stratified analysis by anatomical subsite. We do not currently have a specific hypothesis that may explain the suggested association between high birth weight and early-onset rectal cancer but not colon cancer.

While the literature on the association between high birth weight and average-onset CRC is limited, the literature on the association of birth weight and eoCRC is even more scarce.(46, 47) Murphy et. al. evaluated maternal characteristics and birth weight and risk of CRC using the Child Health and Development Studies cohort.(47) About half of their CRC cases were early-onset, albeit small in number (n=33). They reported an elevated risk of overall CRC in individuals weighing ≥4000gms versus 2500–3999gms at birth [adjusted HR (95% CI): 1.95 (0.86–4.38)]; however, they did not adjust for adult metabolic conditions in the offsprings and they did not evaluate CRC risk by cancer sub-sites.(47) To our knowledge, no prior studies have directly evaluated the association between birth weight and eoCRC. We expanded the literature by focusing on the evaluation of eoCRC. Prior studies, however, have reported on the association between high birth weight and risk of other early-onset cancer, such as breast and ovarian cancer.(48–52) High birth weight was also identified as a risk factor for thyroid cancer and synovial sarcoma in pediatric and young adult patients, respectively.(53, 54)

Based on the “Developmental Origins of Health and Disease” hypothesis, intrauterine environment can permanently set the structure and function of organ systems, and can lead to impairment of key physiologic and metabolic processes.(13) Birth weight is an indicator of fetal nutritional status and intrauterine development, which are potential risk factors in carcinogenesis.(42, 47, 55) Of the plausible biological mechanisms behind the increased risk of carcinogenesis in individuals with high birth weight, intrauterine programming of key hormonal axes, such as the hypothalamic-pituitary-adrenal axis and the insulin-like growth factor system is most commonly discussed in literature. (18, 22–27, 42) The correlation of high birth weight with a greater reservoir of stem cells also increases an individual’s risk for carcinogenesis,(28, 29) although our null findings on colon cancer are contrary to this hypothesis. In our study, the findings on high birth weight were stable with and without adjusting for adult metabolic conditions and adult height, suggesting that alterations of intrauterine hormone programming may have direct effect or other causal pathways through which it affects risk of eoCRC, if the association is true. Future studies in larger populations are warranted to further shed light on the role of intrauterine environment in the development of eoCRC.

We used two alternate definitions to define birth weight, and our findings were fairly consistent across both definitions. Macrosomia reflects sheer birth weight and may reflects more on the stem cell reserve hypothesis compared with the LGA definitions; while the LGA likely reflects more on the altered intrauterine hormonal programming hypothesis compared with macrosomia. However, infants who meet those definitions largely overlap; thus, we did not intend to use these definitions to compare specific causal mechanisms.

There are several limitations to consider when interpreting our findings. First, we did not have birth records available for those born outside the state of California (approximately 44% of the California population)(56). Further, we were only able to achieve a 35% match rate for birth record linkage, which was lower than expected. When we excluded cases with no matched controls and controls with no matched cases, the retention rate declined further (34% for cases vs 14% for controls). Due to the need for birth record linkage, our final study sample might not represent the source population in the distribution of age at diagnosis, sex, race/ethnicity, and calendar year of diagnosis. That said, we evaluated whether these factors may modify the associations and did not find strong evidence for such effect modification. Given that odds ratio estimates were somewhat similar across strata defined by age group (although we were unable to evaluate those < age 30 years at diagnosis separately), sex, race/ethnicity, and calendar year at diagnosis subgroups, the differential distribution of those covariates among those with a linked birth record may not significantly affect the generalizability of our estimates. However, those with and without a linked birth record also likely differ in the distribution of other unmeasured factors such as immigration status and geographic location of birth (California vs. non-California, which may reflect different obstetric practices). It is possible that some of these factors (such as immigration status) may modify the association between high birth weight and eoCRC risk and thus affect the generalizability of our estimate to the overall US population. Future studies from border geographic areas are needed to further shed light on the association between birth weight and eoCRC. Second, the sample size is limited after birth record linkage, and we may be under powered for analysis by cancer site. It is also possible that the observed associations were confounded and thus spurious. We did not have data on relevant maternal characteristics (e.g., gestational diabetes, health behaviors) during pregnancy or socioeconomic factors; thus, residual confounding by these variables is possible.

To our knowledge, this is one of the first epidemiologic studies to evaluate the association between high birth weight and risk of eoCRC. The strengths of our study included linkage between population based eoCRC cases and controls and birth certificates data, the adjustments for potential confounders as well as adult metabolic conditions, and the evaluations of associations for different eoCRC sites. Our observations of a potential association between high birth weight and eoCRC, primarily rectal cancer, call for further evaluation on this topic in large datasets with more detailed confounder information. If the association is confirmed, future studies could evaluate whether including high birthweight in models predicting eoCRC improves their accuracy.

Supplementary Material

1
2
3
4

Acknowledgements

The authors thank the patients of Kaiser Permanente for helping us improve care through the use of information collected through our electronic health record systems.

Funding

Research reported in this publication was supported by The National Cancer Institute of the National Institutes of Health under award number 1R01CA258160 (PI: CR Chao). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health

This work was supported by grant# 1R01CA258160 funded by the National Cancer Institute (PI; CR Chao).

Footnotes

Conflict of Interest Statement:

The authors declare there are no other potential conflicts of interest.

REFERENCES

  • 1.Saus E, Iraola-Guzmán S, Willis JR, Brunet-Vega A, Gabaldón T. Microbiome and colorectal cancer: Roles in carcinogenesis and clinical potential. Molecular Aspects of Medicine. 2019;69:93–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.SEER Cancer Statistics Factsheets: Common Cancer Sites. [cited 2026 May 11]. Available from: http://www.ama-assn.org/ama/pub/category/1736.html.
  • 3.SEER Cancer Stat Facts: Colorectal Cancer. [cited 2026 May 11]. Available from: https://seer.cancer.gov/statfacts/html/colorect.html.
  • 4.Bailey CE, Hu CY, You YN, Bednarski BK, Rodriguez-Bigas MA, Skibber JM, et al. Increasing disparities in the age-related incidences of colon and rectal cancers in the United States, 1975–2010. JAMA surgery. 2015;150(1):17–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Colorectal Cancer Facts & Figures 2023–2025. Atlanta. GA: American Cancer Society; 2023. Contract No.: September, 2025. [Google Scholar]
  • 6.Carr PR, Weigl K, Jansen L, Walter V, Erben V, Chang-Claude J, et al. Healthy Lifestyle Factors Associated With Lower Risk of Colorectal Cancer Irrespective of Genetic Risk. Gastroenterology. 2018;155(6):1805–15 e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Mauri G, Sartore-Bianchi A, Russo AG, Marsoni S, Bardelli A, Siena S. Early-onset colorectal cancer in young individuals. Molecular oncology. 2019;13(2):109–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Siegel RL, Jemal A, Ward EM. Increase in incidence of colorectal cancer among young men and women in the United States. Cancer Epidemiol Biomarkers Prev. 2009;18(6):1695–8. [DOI] [PubMed] [Google Scholar]
  • 9.You YN, Xing Y, Feig BW, Chang GJ, Cormier JN. Young-onset colorectal cancer: is it time to pay attention? Arch Intern Med. 2012;172(3):287–9. [DOI] [PubMed] [Google Scholar]
  • 10.Georgiou A, Khakoo S, Edwards P, Minchom A, Kouvelakis K, Kalaitzaki E, et al. Outcomes of Patients with Early Onset Colorectal Cancer Treated in a UK Specialist Cancer Center. Cancers (Basel). 2019;11(10). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Araghi M, Soerjomataram I, Bardot A, Ferlay J, Cabasag CJ, Morrison DS, et al. Changes in colorectal cancer incidence in seven high-income countries: a population-based study. Lancet Gastroenterol Hepatol. 2019;4(7):511–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hua H, Jiang Q, Sun P, Xu X. Risk factors for early-onset colorectal cancer: systematic review and meta-analysis. Front Oncol. 2023;13:1132306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Barker DJP. The Developmental Origins of Adult Disease. Journal of the American College of Nutrition. 2004;23(sup6):588S–95S. [DOI] [PubMed] [Google Scholar]
  • 14.Ozturk HNO, Turker PF. Fetal programming: could intrauterin life affect health status in adulthood? Obstet Gynecol Sci. 2021;64(6):473–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Lindsay KL, Buss C, Wadhwa PD, Entringer S. The Interplay Between Nutrition and Stress in Pregnancy: Implications for Fetal Programming of Brain Development. Biol Psychiatry. 2019;85(2):135–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Nigam M, Aschebrook-Kilfoy B, Shikanov S, Eggener S. Increasing incidence of testicular cancer in the United States and Europe between 1992 and 2009. World journal of urology. 2015;33(5):623–31. [DOI] [PubMed] [Google Scholar]
  • 17.Ahlgren M, Melbye M, Wohlfahrt J, Sorensen TI. Growth patterns and the risk of breast cancer in women. N Engl J Med. 2004;351(16):1619–26. [DOI] [PubMed] [Google Scholar]
  • 18.Michos A, Xue F, Michels KB. Birth weight and the risk of testicular cancer: a meta-analysis. International journal of cancer. 2007;121(5):1123–31. [DOI] [PubMed] [Google Scholar]
  • 19.Xue F, Michels KB. Intrauterine factors and risk of breast cancer: a systematic review and meta-analysis of current evidence. The Lancet Oncology. 2007;8(12):1088–100. [DOI] [PubMed] [Google Scholar]
  • 20.Park SK, Kang D, McGlynn KA, Garcia-Closas M, Kim Y, Yoo KY, et al. Intrauterine environments and breast cancer risk: meta-analysis and systematic review. Breast cancer research : BCR. 2008;10(1):R8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Xue F, Rosner B, Eliassen H, Michels KB. Body fatness throughout the life course and the incidence of premenopausal breast cancer. International journal of epidemiology. 2016;45(4):1103–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Ferguson P, Mohr L, Hoel D, Lipsitz S, Lackland D. Possible relationship between birth weight and cancer incidence among young adults. Annals of epidemiology. 2000;10(7):471. [DOI] [PubMed] [Google Scholar]
  • 23.Silva Idos S, De Stavola B, McCormack V. Birth size and breast cancer risk: re-analysis of individual participant data from 32 studies. PLoS medicine. 2008;5(9):e193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Tettamanti G, Ljung R, Mathiesen T, Schwartzbaum J, Feychting M. Birth Size Characteristics and Risk of Brain Tumors in Early Adulthood: Results from a Swedish Cohort Study. Cancer Epidemiol Biomarkers Prev. 2016;25(4):678–85. [DOI] [PubMed] [Google Scholar]
  • 25.Ahlgren M, Wohlfahrt J, Olsen LW, Sorensen TI, Melbye M. Birth weight and risk of cancer. Cancer. 2007;110(2):412–9. [DOI] [PubMed] [Google Scholar]
  • 26.Kajantie E, Phillips DI, Andersson S, Barker DJ, Dunkel L, Forsen T, et al. Size at birth, gestational age and cortisol secretion in adult life: foetal programming of both hyper- and hypocortisolism? Clinical endocrinology. 2002;57(5):635–41. [DOI] [PubMed] [Google Scholar]
  • 27.Kajantie E, Feldt K, Raikkonen K, Phillips DI, Osmond C, Heinonen K, et al. Body size at birth predicts hypothalamic-pituitary-adrenal axis response to psychosocial stress at age 60 to 70 years. The Journal of clinical endocrinology and metabolism. 2007;92(11):4094–100. [DOI] [PubMed] [Google Scholar]
  • 28.Strohsnitter WC, Savarese TM, Low HP, Chelmow DP, Lagiou P, Lambe M, et al. Correlation of umbilical cord blood haematopoietic stem and progenitor cell levels with birth weight: implications for a prenatal influence on cancer risk. British journal of cancer. 2008;98(3):660–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Capittini C, Bergamaschi P, De Silvestri A, Marchesi A, Genovese V, Romano B, et al. Birth-weight as a risk factor for cancer in adulthood: the stem cell perspective. Maturitas. 2011;69(1):91–3. [DOI] [PubMed] [Google Scholar]
  • 30.Johnson W, Choh AC, Soloway LE, Czerwinski SA, Towne B, Demerath EW. Eighty-year trends in infant weight and length growth: the Fels Longitudinal Study. J Pediatr. 2012;160(5):762–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Surkan PJ, Hsieh CC, Johansson AL, Dickman PW, Cnattingius S. Reasons for increasing trends in large for gestational age births. Obstetrics and gynecology. 2004;104(4):720–6. [DOI] [PubMed] [Google Scholar]
  • 32.Koebnick C, Langer-Gould AM, Gould MK, Chao CR, Iyer RL, Smith N, et al. Sociodemographic characteristics of members of a large, integrated health care system: comparison with US Census Bureau data. Perm J. 2012;16(3):37–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Davis AC, Voelkel JL, Remmers CL, Adams JL, McGlynn EA. Comparing Kaiser Permanente Members to the General Population: Implications for Generalizability of Research. Perm J. 2023;27(2):87–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Schwartz R, Teramo KA. What is the significance of macrosomia? Diabetes care. 1999;22(7):1201–5. [DOI] [PubMed] [Google Scholar]
  • 35.Hiersch L, Shinar S, Melamed N, Aviram A, Hadar E, Yogev Y, et al. Birthweight and large for gestational age trends in non-diabetic women with three consecutive term deliveries. Archives of gynecology and obstetrics. 2018;298(4):725–30. [DOI] [PubMed] [Google Scholar]
  • 36.Getahun D, Ananth CV, Peltier MR, Salihu HM, Scorza WE. Changes in prepregnancy body mass index between the first and second pregnancies and risk of large-for-gestational-age birth. American journal of obstetrics and gynecology. 2007;196(6):530.e1-8. [DOI] [PubMed] [Google Scholar]
  • 37.Getahun D, Lawrence JM, Fassett MJ, Strickland D, Koebnick C, Chen W, et al. The association between stillbirth in the first pregnancy and subsequent adverse perinatal outcomes. American journal of obstetrics and gynecology. 2009;201(4):378.e1–6. [DOI] [PubMed] [Google Scholar]
  • 38.Centers for Disease C, Prevention. BMI Categories. 2025. [Cited 2026 June 1]. Available from: https://www.cdc.gov/bmi/adult-calculator/bmi-categories.html.
  • 39.Lee KJ, Carlin JB. Multiple imputation for missing data: fully conditional specification versus multivariate normal imputation. Am J Epidemiol. 2010;171(5):624–32. [DOI] [PubMed] [Google Scholar]
  • 40.Smith NR, Jensen BW, Zimmermann E, Gamborg M, Sorensen TI, Baker JL. Associations between birth weight and colon and rectal cancer risk in adulthood. Cancer Epidemiol. 2016;42:181–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Spracklen CN, Wallace RB, Sealy-Jefferson S, Robinson JG, Freudenheim JL, Wellons MF, et al. Birth weight and subsequent risk of cancer. Cancer Epidemiol. 2014;38(5):538–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Su L, Hendryx M, Li M, Pichardo MS, Jung SY, Lane DS, et al. Birth weight, adult body size, and risk of colorectal cancer. Cancer Epidemiol. 2023;85:102407. [DOI] [PubMed] [Google Scholar]
  • 43.Morois S, Mesrine S, Besemer F, Josset M, Clavel-Chapelon F, Boutron-Ruault MC. Risks of colon and rectal adenomas are differentially associated with anthropometry throughout life: the French E3N prospective cohort. International journal of epidemiology. 2011;40(5):1269–79. [DOI] [PubMed] [Google Scholar]
  • 44.Yang TO, Reeves GK, Green J, Beral V, Cairns BJ, Million Women Study C. Birth weight and adult cancer incidence: large prospective study and meta-analysis. Ann Oncol. 2014;25(9):1836–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Cnattingius S, Lundberg F, Iliadou A. Birth characteristics and risk of colorectal cancer: a study among Swedish twins. British journal of cancer. 2009;100(5):803–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Syed AR, Thakkar P, Horne ZD, Abdul-Baki H, Kochhar G, Farah K, et al. Old vs new: Risk factors predicting early onset colorectal cancer. World journal of gastrointestinal oncology. 2019;11(11):1011–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Murphy CC, Cirillo PM, Krigbaum NY, Singal AG, Lee M, Zaki T, et al. Maternal obesity, pregnancy weight gain, and birth weight and risk of colorectal cancer. Gut. 2022;71(7):1332–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Mellemkjaer L, Olsen ML, Sorensen HT, Thulstrup AM, Olsen J, Olsen JH. Birth weight and risk of early-onset breast cancer (Denmark). Cancer causes & control : CCC. 2003;14(1):61–4. [DOI] [PubMed] [Google Scholar]
  • 49.Diaz-Santana MV, O’Brien KM, D’Aloisio AA, Regalado G, Sandler DP, Weinberg CR. Perinatal and postnatal exposures and risk of young-onset breast cancer. Breast cancer research : BCR. 2020;22(1):88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Innes K, Byers T, Schymura M. Birth characteristics and subsequent risk for breast cancer in very young women. Am J Epidemiol. 2000;152(12):1121–8. [DOI] [PubMed] [Google Scholar]
  • 51.Kang SJ. Birth Characteristics And Risk Of Early-Onset Ovarian Cancer: Yale; 2021. https://elischolar.library.yale.edu/ysphtdl/2061
  • 52.Barba M, McCann SE, Nie J, Vito D, Stranges S, Fuhrman B, et al. Perinatal exposures and breast cancer risk in the Western New York Exposures and Breast Cancer (WEB) Study. Cancer causes & control : CCC. 2006;17(4):395–401. [DOI] [PubMed] [Google Scholar]
  • 53.Deziel NC, Zhang Y, Wang R, Wiemels JL, Morimoto L, Clark CJ, et al. Birth Characteristics and Risk of Pediatric Thyroid Cancer: A Population-Based Record-Linkage Study in California. Thyroid. 2021;31(4):596–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Wiemels JL, Wang R, Feng Q, Clark CJ, Amatruda JF, Rubin E, et al. Birth Characteristics and Risk of Early-Onset Synovial Sarcoma. Cancer Epidemiol Biomarkers Prev. 2020;29(6):1162–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Bothou A, Zervoudis S, Tsatsaris G, Pappou P, Liadopoulou M, Iatrakis G, et al. Neonatal Birth Weight of the Woman as a Risk Factor for Breast Cancer in her Life: a Case-Control Bicentric Study. Mater Sociomed. 2021;33(2):119–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.US Census Bureau Population Estimates. [cited 2026 May 11]. Available from https://www.ppic.org/publication/californias-population/.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2205062-supplement-1.docx (14.8KB, docx)
2
NIHMS2205062-supplement-2.docx (18.8KB, docx)
3
NIHMS2205062-supplement-3.docx (17.1KB, docx)
4
NIHMS2205062-supplement-4.png (604.4KB, png)

Data Availability Statement

Anonymized data that support the findings of this study may be made available from the corresponding author on reasonable requests from qualified researchers with documented evidence of training for human subjects protections.

RESOURCES