
==== Front
J Nutr
J Nutr
The Journal of Nutrition
0022-3166
1541-6100
American Society for Nutrition

S0022-3166(24)00293-1
10.1016/j.tjnut.2024.05.016
Nutritional Epidemiology
Intake of Sugar and Food Sources of Sugar and Colorectal Cancer Risk in the Multiethnic Cohort Study
Kanehara Rieko rkanehar@ncc.go.jp
12⁎
Park Song-Yi 1
Okada Yuito 1
Iwasaki Motoki 3
Tsugane Shoichiro 24
Sawada Norie 2
Inoue Manami 5
Haiman Christopher A 6
Wilkens Lynne R 1†
Le Marchand Loïc 1
1 Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, United States
2 Division of Cohort Research, National Cancer Center Institute for Cancer Control, Tokyo, Japan
3 Division of Epidemiology, National Cancer Center Institute for Cancer Control, Tokyo, Japan
4 International University of Health and Welfare Graduate School of Public Health, Tokyo, Japan
5 National Cancer Center Institute for Cancer Control, Tokyo, Japan
6 Department of Population and Public Health Sciences, Keck School of Medicine and Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, CA, United States
⁎ Corresponding author. rkanehar@ncc.go.jp
† Lynne R. Wilkens is a member of Statistical Review Board for The Journal of Nutrition and played no role in the Journal’s evaluation of the manuscript.

23 5 2024
8 2024
23 5 2024
154 8 24812492
4 1 2024
10 5 2024
18 5 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Background

The influence of sugar intake on the risk of colorectal cancer (CRC) remains controversial, and there is a need to investigate the heterogeneity of effects among racial and ethnic groups.

Objectives

To examine the association of intake of simple sugars and their food sources with CRC risk according to race/ethnicity in a multiethnic cohort study.

Methods

We analyzed data from 192,651 participants who participated in the Multiethnic Cohort Study comprising African American, Japanese American, Latino, Native Hawaiian, and White older adults living in Hawaii and California with an average follow-up of 19 y. Intakes of total and specific types of sugars and sugary foods were estimated from a quantitative food frequency questionnaire completed by the participants in 1993–1996. We estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for CRC risk according to quintiles (Q) of sugar and food intakes using Cox models adjusted for potential confounders.

Results

As of December 2017, 4403 incident CRC cases were identified. Among all participants, multivariable-adjusted CRC HRs for Q2, Q3, Q4, and Q5 compared with Q1 for total sugars were 1.03 (95% CI: 0.94, 1.13), 1.05 (95% CI: 0.96, 1.16), 1.12 (95% CI: 1.01, 1.24), and 1.13 (95% CI: 1.01, 1.27), respectively. A similar positive association was observed for total fructose, glucose, fructose, and maltose but not for added sugars and sugary foods. The increased risk appeared to be limited to colon cancer and to be strongest among younger participants (i.e., 45–54 y at baseline); an association with CRC was observed for sugar-sweetened beverages in the latter group. Among racial and ethnic groups, increased risk of CRC was most apparent in Latinos.

Conclusions

In this diverse cohort, intakes of total sugar, total fructose, glucose, fructose, and maltose were associated with an increased risk of CRC, and the association was strongest for colon cancer, younger participants, and Latinos.

Keywords

colorectal cancer
dietary sugar
Multiethnic Cohort Study
middle-aged adults
food sources of sugar
Abbreviations

CI confidence interval

CRC colorectal cancer

GL glycemic load

HR hazard ratio

IGF-I insulin-like growth factor-I

MEC Multiethnic Cohort

NSAID nonsteroidal anti-inflammatory drug

QFFQ quantitative food frequency questionnaire

SSB sugar-sweetened beverage
==== Body
pmcIntroduction

More than 1.93 million new colorectal cancer (CRC) cases were diagnosed worldwide in 2020, making it the third most common of all cancers and the second leading cause of cancer death [1]. Obesity and hyperinsulinemia, which have been linked to excess sugar intake [2,3], have been reported to be risk factors for CRC [4,5]. However, the influence of sugar intake on the risk of CRC remains controversial, as reviewed in the latest report on nutrition and cancer by the World Cancer Research Fund and the American Institute for Cancer Research [6].

Several previous studies of the United States population have reported no association with CRC for intake of total sugar (total of mono- and disaccharides) [7,8], glucose [9], and sucrose [[8], [9], [10], [11], [12]]. However, in 2 of these previous studies, fructose intake was associated with an increased risk of CRC [10,12]. Moreover, a cohort study in Japan has suggested that a higher intake of total sugar was associated with an increased risk of rectal cancer among women, but there was no clear association between sugar intake and colon cancer or CRC overall [13]. These inconsistencies among previous studies may reflect variations in the overall sugar intake level or the preponderance of different sugar subtypes in the diet and differences in their health impacts among racial and ethnic groups. Therefore, investigating the heterogeneity of the association of the various simple sugars among individuals of different races and ethnic groups may have important etiologic and public health implications.

One of the risk factors for colorectal carcinogenesis is postulated to be insulin resistance, which can be induced by excessive de novo lipogenesis [4,5]. Consumption of glucose and disaccharides that include glucose (sucrose, maltose, lactose) directly leads to insulin secretion. In contrast, fructose has a unique metabolic pathway as it contributes to de novo lipogenesis and adiposity and is not regulated by the negative feedback from total energy intake [14,15]. On the other hand, a case-control study in the United Kingdom and an animal study reported that galactose in dietary fiber showed a protective effect on CRC [16,17]. To clarify the role of sugar intake on CRC, comprehensive investigations focusing on both total (total sugars, total fructose) and specific types of simple sugars (glucose, fructose, galactose, sucrose, maltose, and lactose) are needed. Furthermore, the effect of sugars on CRC may differ by food sources, for example, between processed foods, such as sugar-sweetened beverages (SSBs) and confectionaries, which include added sugars, and less processed or unprocessed foods such as fruits, which include naturally occurring sugar. Although fruit intake is thought to have multiple health benefits, it remains uncertain whether sugar from fruits is detrimental with regard to CRC risk [18]. Moreover, an animal study reported that high-fructose corn syrup intake promotes tumor development in the colon without obesity and metabolic dysfunction [19], suggesting a role of sugar intake may vary depending on the anatomical subsites of CRC.

Here, we aimed to examine the association between intake of simple sugars and their food sources and CRC risk overall and according to anatomical subsites and race/ethnicity in a large population-based multiethnic cohort study. We focused on total sugar (sum of simple sugars), total fructose (sum of fructose as monosaccharides and half of the sucrose), added sugar, and specific types of simple sugars (glucose, fructose, galactose, sucrose, maltose, and lactose) and their main food sources, including whole fruits, fruit juice, SSBs, and confectionaries.

Methods

Study population

The Multiethnic Cohort (MEC) study includes >215,000 Hawaii and California residents, aged 45 to 75 y at baseline, mainly consisting of African Americans, Japanese Americans, Latinos, Native Hawaiians, and Whites [20]. They entered the cohort by completing a 26-page mailed questionnaire between 1993 and 1996. In the questionnaire, participants answered a detailed dietary assessment and provided information on body weight, height, physical activity, smoking status, personal medical history, family history of cancer, and reproductive history.

In the current analysis, we excluded participants who were not assigned to 1 of the 5 racial/ethnic groups (n = 12,206), had prior CRC reported on the baseline questionnaire (n = 2272) or from tumor registries (n = 302), lack of follow-up (n = 23, case classification error [n = 14] or inconsistent dates of diagnosis and death [n = 9]), and reported implausible diets based on total energy intake (distance from the mean >3 times a robust SD based on a truncated normal distribution) or its components (distance from the mean >3.5 times robust SDs) (n = 8180), leaving a total of 192,651 cohort members for analysis (Supplemental Figure 1).

As explained in the invitation to participate, the return of a completed questionnaire indicated informed consent. This study was approved by the institutional review boards of the University of Hawaii and the University of Southern California.

Dietary assessment

A quantitative food frequency questionnaire (QFFQ) of >180 food items was used to estimate usual dietary intake during the previous year. Details of the QFFQ development and validation are described elsewhere [[20], [21], [22]]. Nutrient intakes were calculated using food composition data derived from the USDA [23,24], other research and commercial publications [[25], [26], [27]], and data from laboratory analyses (unpublished data), which was developed to ensure that all data on foods consumed by the 5 races/ethnicities was covered. We estimated the average daily intake of simple sugars (glucose, fructose, galactose, sucrose, maltose, lactose, total sugar, total fructose, and added sugar), food sources of sugars (whole fruits, fruit juice, SSBs, and confectionaries) (Supplemental Table 1), and other foods and nutrients [[20], [21], [22]]. Glucose, fructose, and galactose were estimated as those present as monosaccharides in foods, while sucrose, maltose, and lactose were estimated as those present as disaccharides. We calculated the sum of glucose, fructose, galactose, sucrose, lactose, and maltose consumption as total sugar intake; for total fructose intake, the formula (fructose intake) + (one-half of sucrose intake) was used [28]. Added sugar intake was defined as the amount of sugar added to foods and beverages during the manufacturing or cooking processes. Energy-adjusted nutrient and food intakes were computed by the density method to lessen the effect of measurement error [29]. Sugars, protein, fat, carbohydrate, and starch intakes were expressed as percentage energy. Other nutrients and foods were presented as per 1000 kcal.

Ascertainment of CRC

Incident cases of invasive adenocarcinoma of the large bowel, overall and by anatomical subsites (colon, proximal colon, distal colon, and rectum), were identified by linkage to the Surveillance, Epidemiology, and End Results Program statewide tumor registries in Hawaii and California. Deaths were identified by linkage to death certificate files in both states and the National Death Index. Participants were followed up for newly diagnosed cases and death ascertainment through 31 December, 2017.

Statistical analysis

Participants were divided into quintiles according to energy-adjusted intakes of each type of sugar (total sugar, total fructose, added sugars, glucose, fructose, galactose, sucrose, maltose, or lactose) or main food sources of sugar (whole fruits, fruit juice, SSBs, or confectionaries) based on the distributions in the entire cohort. The mean, SD, and proportion of participant characteristics are presented according to quintiles of total sugar intake.

We performed Cox proportional hazards regression of CRC with age as the time metric to assess the hazard ratios (HRs) and 95% confidence intervals (CIs) risk for the quintiles of energy-adjusted sugar or food consumption. The lowest category was used as the reference. Observation began at cohort entry and ended at the earliest of the CRC diagnosis, death, or study closure on 31 December, 2017. Potential confounders, which were ascertained in the baseline questionnaire, were included based on their clinical and biological plausibility and their having no strong multicollinearity. Model 1 was adjusted for race/ethnicity and sex as strata and age at cohort entry as a covariate. Model 2 was additionally adjusted for the following baseline covariates: BMI (kg/m2, quintiles), alcohol consumption, smoking status (never, past, or current), number of pack-years, moderate or vigorous physical activity (hours per day), education (less than college education or more), family history of CRC (yes or no), history of polyps of the intestines (yes or no), history of diabetes (yes or no), use of exogenous female hormones (never, past, or current, in women only), nonsteroidal anti-inflammatory drug (NSAID) use (never or ever), total energy intake, and intake of saturated fatty acids, n–3 polyunsaturated fatty acids, vitamin D, calcium, dietary fiber, and folate. To examine the linear trend across quintiles, we modeled the median sugar intake according to sex and race/ethnicity for each category as a continuous variable. We excluded from the models participants who were missing any of the covariates (n = 37,115). Because the results were similar in women and men, we combined the 2 sexes in the main tables. We also performed the analyses stratified by race/ethnicity (African American, Native Hawaiian, Japanese American, Latino, or White) and tumor anatomical subsite (colon, proximal colon, distal colon, or rectum). CRC cases with subsites other than those of interest were censored at the time of the CRC diagnosis. The differences in CRC-sugar associations in subsites (colon compared with rectum, proximal compared with distal colon) were tested using Lunn’s approach [30].

We further performed stratified analyses for the associations between total sugar, total fructose, fructose, glucose, maltose, or main food sources of sugar intake and CRC risk according to age at cohort entry, follow-up period, alcohol drinking status, smoking status, BMI, and history of diabetes. The differences between CRC-sugar associations in subgroups were evaluated by inclusion of interaction terms between the sugar intake variable (the median according to sex and race/ethnicity for each category) and subgroup membership.

In another sensitivity analysis, participants who were diagnosed with CRC during the first 3 y of follow-up were excluded to minimize reverse causation. Additionally, we examined the association assuming the missing covariate data were missing at random and imputed missing BMI, smoking status, pack-years, physical activity, education, history of polyps of the intestine, history of diabetes, family history of CRC, use of exogenous female hormones, and use of NSAIDs. The SAS MI procedure was used for 5 rounds of multiple imputations. We then combined estimations of HRs from each imputed dataset using Rubin’s rules and the SAS MIANALYZE procedure. All statistical tests were 2-sided, and all analyses were performed using SAS (version 9.4; SAS Institute).

Results

During a mean follow-up of 19.3 y, 4,403 new cases of CRC (2,262 male; 2,141 female) were identified. For each sex, participants in the highest quintile of total sugar intake tended to have a lower proportion of current smokers, a lower frequency of diabetes at baseline, and a lower intake of alcohol, protein, fat, and starch. In contrast, intakes of dietary fiber, calcium, vitamin D, vitamin B6, vitamin B12, and folate were higher in the highest than in the lowest quintile of total sugar intake (Table 1). Intakes of total fructose, glucose, fructose, and sucrose were strongly correlated with total sugar intake (Spearman’s correlation coefficients: r = 0.95, 0.87, 0.86, and 0.83, respectively), whereas the correlations with total sugar were modest for added sugars and whole fruits (r = 0.55 and 0.57, respectively) and lower for galactose, maltose, lactose, confectionary, fruit juice, and SSB intake (r = 0.27, 0.28, 0.34, 0.19, 0.40, and 0.33, respectively) (Supplemental Table 2). The percentages of the contributions of foods and food groups to total sugar intake by race/ethnic groups are shown in Supplemental Table 3.TABLE 1 Baseline characteristics of participants according to quintiles of total sugar intake (n = 192,651)

TABLE 1	n	Quintiles of total sugar intake	
Q1	Q2	Q3	Q4	Q5	
Median (interquartile ranges) oftotal sugar intake1, % energy/d		10.3 (8.6, 11.5)	14.3 (13.5, 15.1)	17.5 (16.7, 18.3)	21.1 (20.1, 22.2)	27.3 (25.1, 31.0)	
Number of subjects		38,531	38,529	38,530	38,531	38,530	
Number without missing data		31,503	31,518	31,403	30,996	30,116	
Sex, %	
 Men	86,741	57.4	49.4	44.6	39.3	34.4	
 Women	105,910	42.6	50.6	55.4	60.7	65.6	
Age, y		57.9 (8.6)2	59.4 (8.8)	60.4 (8.8)	61.1 (8.8)	60.9 (8.8)	
Area, %	
 Hawaii	91,180	53.4	50.1	47.8	45.5	39.9	
 Los Angeles	101,471	46.6	49.9	52.2	54.5	60.1	
Race/ethnicity, %	
 African American	32,765	13.5	15.0	16.5	18.3	21.7	
 Native Hawaiian	13,744	8.6	7.6	7.1	6.2	6.1	
 Japanese American	53,961	37.2	31.5	27.8	24.7	18.9	
 Latino	45,182	21.6	22.7	23.1	23.5	26.5	
 White	46,999	19.1	23.2	25.5	27.4	26.8	
Body mass index, kg/m2		27.1 (5.3)	26.8 (5.1)	26.5 (5.1)	26.3 (4.9)	26.4 (5.1)	
 <20 kg/m2, %	9,930	4.2	4.6	5.3	5.8	5.9	
 20 to <25 kg/m2, %	68,817	32.5	34.6	36.7	37.8	37.0	
 25 to <30 kg/m2, %	73,643	39.7	39.0	38.0	37.5	37.0	
 ≥30 kg/m2, %	37,699	22.4	20.7	18.8	17.6	18.4	
 Missing data, %	2,562	1.3	1.1	1.3	1.3	1.6	
Alcohol intake, g/d		20.0 (43.6)	9.3 (21.5)	6.8 (16.6)	5.0 (13.5)	3.5 (12.7)	
Smoking status, %	
 Never smoked	83,656	32.3	40.5	44.9	48.7	50.7	
 Past smoker	75,552	43.7	41.8	39.5	37.0	34.0	
 Current smoker	30,450	22.8	16.3	14.1	12.6	13.2	
 Missing data, %	2,993	1.2	1.3	1.5	1.6	2.1	
Pack-years of smoking		14.1 (16.8)	10.9 (15.4)	9.4 (14.3)	8.3 (13.6)	7.8 (13.4)	
Physical activity (moderate or vigorous), %	
 <0.5 h/d	74,481	40.2	38.3	37.7	37.5	39.6	
 0.5 to <1 h/d	40,567	21.6	21.6	21.1	21.1	19.9	
 1 to <1.5 h/d	29,928	15.0	15.9	15.9	15.7	15.2	
 1.5 to <2 h/d	8,728	4.2	4.8	4.8	4.5	4.4	
 ≥2 h/d	35,176	16.9	17.7	18.9	19.3	18.5	
 Missing data, %	3,771	2.0	1.7	1.7	1.9	2.4	
Education, %	
 <College education	99,682	52.9	51.1	51.1	51.0	52.5	
 ≥College education	90,749	46.0	47.8	47.8	47.7	46.2	
 Missing data, %	2,220	1.1	1.0	1.1	1.2	1.4	
History of polyps of intestines, %	10,498	5.8	5.5	5.8	5.7	4.5	
 Missing data, %	1	0.0	0.003	0.0	0.0	0.0	
History of diabetes, %	22,629	16.7	13.2	11.2	9.7	7.9	
 Missing data, %	1	0.0	0.003	0.0	0.0	0.0	
Family history of CRC, %	15,293	8.0	8.0	7.9	8.1	7.6	
 Missing data, %	25,206	12.6	12.6	12.6	13.3	14.4	
Postmenopausal status3, %	90,662	79.7	83.6	86.1	87.7	88.6	
 Missing data, %	1,022	1.2	0.9	0.9	0.9	1.0	
Current use of exogenous female hormones, %	29,269	10.8	14.1	16.1	17.3	17.7	
 Missing data, %	3,322	1.4	1.4	1.6	1.9	2.2	
NSAID use, %	99,038	48.2	50.7	51.7	53.0	53.3	
 Missing data, %	4,728	2.4	2.3	2.3	2.4	3.0	
Food and nutrient intake	
 Total energy, kcal/d		2248 (1110)	2207 (1061)	2181 (1053)	2130 (1031)	2132 (1079)	
 Protein, % energy/d		15.5 (3.2)	15.4 (2.6)	14.9 (2.4)	14.4 (2.4)	13.1 (2.6)	
 Fat, % energy/d		32.8 (8.0)	32.5 (6.4)	31.2 (6.0)	29.2 (5.8)	25.3 (6.1)	
 SFAs, % energy/d		9.5 (2.8)	9.7 (2.5)	9.4 (2.4)	9.0 (2.5)	7.9 (2.5)	
 MUFAs, % energy/d		12.1 (3.1)	11.9 (2.5)	11.3 (2.4)	10.5 (2.3)	9.0 (2.4)	
 PUFAs, % energy/d		7.7 (2.1)	7.6 (1.7)	7.2 (1.6)	6.7 (1.5)	5.7 (1.5)	
 n–3 PUFAs, % energy/d		0.8 (0.2)	0.8 (0.2)	0.8 (0.2)	0.7 (0.2)	0.6 (0.2)	
 n–6 PUFAs, % energy/d		6.8 (1.9)	6.7 (1.6)	6.4 (1.4)	5.9 (1.4)	5.0 (1.4)	
 Carbohydrate, % energy/d		46.3 (9.5)	49.3 (7.5)	51.8 (6.8)	54.8 (6.4)	60.6 (7.0)	
 Starch, % energy/d		28.9 (10.1)	26.6 (7.4)	25.1 (6.3)	23.5 (5.7)	20.3 (5.4)	
 Dietary fiber, g/1000 kcal		9.4 (3.4)	11.0 (3.3)	12.0 (3.6)	12.9 (4.0)	13.7 (5.3)	
 Calcium, mg/1000 kcal		397.8 (270.6)	476.8 (291.7)	525.1 (314.3)	571.0 (351.2)	584.9 (385.7)	
 Magnesium, mg/1000 kcal		143.1 (29.1)	151.9 (30.2)	157.7 (32.2)	162.7 (34.2)	161.7 (40.8)	
 Vitamin D, μg/1000 kcal		137.3 (188.1)	162.4 (192.7)	178.7 (199.9)	195.6 (213.7)	197.6 (224.7)	
 Vitamin B6, mg/1000 kcal		1.5 (1.2)	1.7 (1.2)	1.8 (1.3)	1.9 (1.4)	1.9 (1.5)	
 Vitamin B12, μg/1000 kcal		5.7 (7.3)	6.4 (7.4)	6.8 (7.7)	7.2 (8.2)	7.1 (8.5)	
 Folate, μg/1000 kcal		379.9 (260.4)	435.8 (273.2)	470.9 (291.6)	499.3 (311.3)	493.2 (330.1)	
 Red meat, g/1000 kcal		24.5 (14.9)	21.2 (12.5)	18.4 (11.4)	15.6 (10.4)	12.3 (9.4)	
 Processed red meat, g/1000 kcal		10.3 (8.1)	8.8 (6.8)	7.6 (6.2)	6.5 (5.6)	5.1 (5.1)	
 Fish, g/1000 kcal		9.3 (9.1)	9.0 (8.4)	8.5 (7.8)	7.8 (7.5)	6.2 (6.7)	
 Vegetables, g/1000 kcal		151.1 (71.3)	161.9 (72.9)	165.6 (79.0)	166.4 (83.8)	156.7 (92.4)	
 Whole fruits, g/1000 kcal		52.0 (38.1)	90.0 (52.6)	121.0 (66.6)	158.9 (86.3)	226.1 (150.4)	
 Fruit juice, g/1000 kcal		13.2 (20.1)	26.0 (31.7)	37.7 (41.4)	51.6 (52.4)	75.4 (77.3)	
 Dairy products, g/1000 kcal		56.7 (47.4)	88.2 (65.4)	107.2 (77.7)	127.2 (94.8)	139.0 (118.8)	
 Confectionaries, g/1000 kcal		20.6 (16.0)	30.7 (21.4)	35.8 (25.8)	38.7 (30.0)	36.9 (33.6)	
 Sugar-sweetened beverages4, g/1000 kcal		55.4 (97.6)	85.1 (123.3)	108.1 (143.4)	133.4 (162.1)	239.6 (274.8)	
 Alternate Mediterranean diet score (energy-adjusted)		3.5 (1.5)	4.0 (1.7)	4.3 (1.7)	4.5 (1.7)	4.3 (1.7)	
 Healthy Eating Index-2015 (HEI-2015)		61.4 (8.9)	66.2 (9.2)	68.5 (9.6)	70.4 (10.2)	70.4 (11.7)	
Abbreviations: CRC, colorectal cancer; MUFA, monounsaturated fatty acid; NSAID, nonsteroidal anti-inflammatory drug; PUFA, polyunsaturated fatty acid; SD, standard deviation; SFA, saturated fatty acid.

All percentage values are percentages of columns.

1 “Total sugars” represents the sum of the consumption of the following saccharides: glucose, fructose, galactose, sucrose, maltose, and lactose.

2 Mean (SD) (all such values).

3 Women only.

4 Sugar-sweetened beverages include regular sodas, fruit drinks, and coffee and tea with added sugar.

The multivariable-adjusted CRC HR (95% CI) for total sugars, comparing the highest with the lowest quintiles, was 1.13 (95% CI: 1.01, 1.27; P for trend = 0.01) (Table 2). The corresponding HRs were 1.11 (95% CI: 0.99, 1.23; P for trend = 0.01) for total fructose, 1.18 (95% CI: 1.06, 1.32; P for trend < 0.001) for glucose, 1.13 (95% CI: 1.01, 1.26; P for trend = 0.01) for fructose, and 1.23 (95% CI: 1.12, 1.36; P for trend < 0.001) for maltose. However, we did not observe any clear association between intakes of added sugars (HR: 1.03; 95% CI: 0.93, 1.14; P for trend = 0.74), galactose (HR: 0.89; 95% CI: 0.80, 1.00; P for trend = 0.28), sucrose (HR: 1.04; 95% CI: 0.94, 1.16; P for trend = 0.71), lactose (HR: 0.93; 95% CI: 0.82, 1.05; P for trend = 0.39), whole fruits (HR: 1.06; 95% CI: 0.93, 1.21; P for trend = 0.07), fruit juice (HR: 1.01; 95% CI: 0.91, 1.11; P for trend = 0.74), SSBs (HR: 1.03; 95% CI: 0.93, 1.15; P for trend = 0.29), and confectionary (HR: 1.07; 95% CI: 0.96, 1.19; P for trend = 0.49) and CRC risk (Table 2). Additionally, fructose from whole fruits and fructose from other sources were examined in association with CRC risk. The results showed no significant association for fructose from whole fruits for the highest compared with the lowest quintiles (HR: 1.06; 95% CI: 0.93, 1.20; P for trend = 0.12) and an increased risk of CRC for fructose from other sources (HR: 1.12; 95% CI: 1.01, 1.23; P for trend = 0.055) (data not shown).TABLE 2 HRs and 95% CIs of colorectal cancer according to quintiles of sugar and food intake (n = 155,536)

TABLE 2	Quintiles of sugar intake1	P for linear trend	
Q1	Q2	Q3	Q4	Q5	
Total sugars2,3 (% energy/d)	10.3 (8.6, 11.6)	14.3 (13.5, 15.1)	17.5 (16.7, 18.3)	21.1 (20.1, 22.2)	27.2 (25.1, 30.8)		
 Number of cases	955	896	876	877	799		
 Person-years	610,345	612,493	605,086	596,685	579,741		
 Model 14	1.00 (reference)	0.89 (0.82, 0.98)	0.86 (0.78, 0.94)	0.85 (0.77, 0.93)	0.81 (0.74, 0.89)	<0.001	
 Model 25	1.00 (reference)	1.03 (0.94, 1.13)	1.05 (0.96, 1.16)	1.12 (1.01, 1.24)	1.13 (1.01, 1.27)	0.01	
Total fructose2,6 (% energy/d)	4.2 (3.5, 4.7)	6.0 (5.6, 6.4)	7.5 (7.1, 7.9)	9.2 (8.7, 9.8)	12.4 (11.3, 14.4)		
 Number of cases	961	848	893	872	829		
 Person-years	601,461	613,103	605,389	597,383	587,014		
 Model 14	1.00 (reference)	0.82 (0.75, 0.91)	0.85 (0.78, 0.93)	0.82 (0.75, 0.90)	0.82 (0.74, 0.90)	<0.001	
 Model 25	1.00 (reference)	0.94 (0.85, 1.03)	1.03 (0.94, 1.14)	1.06 (0.96, 1.17)	1.11 (0.99, 1.23)	0.01	
Added sugars2 (% energy/d)	3.7 (3.0, 4.3)	5.6 (5.2, 6.0)	7.3 (6.9, 7.7)	9.4 (8.8, 10.1)	13.6 (12.0, 16.9)		
 Number of cases	945	914	869	821	854		
 Person-years	580,797	606,779	611,871	609,695	595,207		
 Model 14	1.00 (reference)	0.94 (0.86, 1.03)	0.90 (0.82, 0.99)	0.86 (0.78, 0.95)	0.97 (0.88, 1.07)	0.51	
 Model 25	1.00 (reference)	1.01 (0.92, 1.10)	0.98 (0.89, 1.08)	0.94 (0.86, 1.04)	1.03 (0.93, 1.14)	0.74	
Glucose2 (% energy/d)	2.0 (1.7, 2.3)	2.9 (2.7, 3.1)	3.7 (3.5, 3.9)	4.8 (4.4, 5.1)	6.9 (6.1, 8.4)		
 Number of cases	882	850	937	897	837		
 Person-years	607,980	609,538	604,063	596,579	586,190		
 Model 14	1.00 (reference)	0.90 (0.82, 0.99)	0.96 (0.88, 1.06)	0.91 (0.83, 1.00)	0.90 (0.82, 1.00)	0.11	
 Model 25	1.00 (reference)	1.00 (0.91, 1.10)	1.14 (1.04, 1.26)	1.14 (1.03, 1.26)	1.18 (1.06, 1.32)	<0.001	
Fructose2 (% energy/d)	1.9 (1.5, 2.2)	3.0 (2.7, 3.2)	4.0 (3.7, 4.3)	5.3 (4.9, 5.7)	7.8 (6.9, 9.6)		
 Number of cases	921	859	904	901	818		
 Person-years	603,697	605,722	605,892	600,270	588,769		
 Model 14	1.00 (reference)	0.87 (0.80, 0.96)	0.89 (0.81, 0.98)	0.88 (0.80, 0.97)	0.85 (0.77, 0.93)	0.01	
 Model 25	1.00 (reference)	0.99 (0.90, 1.09)	1.08 (0.98, 1.19)	1.13 (1.02, 1.25)	1.13 (1.01, 1.26)	0.01	
Galactose2 (% energy/d)	0.02 (0.02, 0.03)	0.04 (0.03, 0.04)	0.06 (0.05, 0.06)	0.09 (0.08, 0.10)	0.20 (0.15, 0.33)		
 Number of cases	1103	897	883	777	743		
 Person-years	578,741	598,726	596,544	609,199	621,139		
 Model 14	1.00 (reference)	0.83 (0.76, 0.91)	0.83 (0.76, 0.91)	0.74 (0.67, 0.82)	0.72 (0.66, 0.80)	<0.001	
 Model 25	1.00 (reference)	0.88 (0.80, 0.97)	0.93 (0.84, 1.02)	0.86 (0.77, 0.95)	0.89 (0.80, 1.00)	0.28	
Sucrose2 (% energy/d)	4.1 (3.4, 4.7)	5.8 (5.5, 6.1)	7.0 (6.7, 7.4)	8.4 (8.0, 8.8)	10.6 (9.8, 11.8)		
 Number of cases	955	915	900	809	824		
 Person-years	599,514	609,062	612,140	603,375	580,259		
 Model 14	1.00 (reference)	0.90 (0.82, 0.99)	0.86 (0.79, 0.95)	0.78 (0.71, 0.85)	0.82 (0.74, 0.90)	<0.001	
 Model 25	1.00 (reference)	1.02 (0.93, 1.12)	1.03 (0.93, 1.13)	0.96 (0.87, 1.07)	1.04 (0.94, 1.16)	0.71	
Maltose2 (% energy/d)	0.19 (0.16, 0.20)	0.24 (0.23, 0.26)	0.29 (0.28, 0.30)	0.34 (0.32, 0.36)	0.43 (0.40, 0.49)		
 Number of cases	782	829	876	940	976		
 Person-years	613,570	619,288	608,425	597,296	565,772		
 Model 14	1.00 (reference)	1.04 (0.95, 1.15)	1.09 (0.99, 1.20)	1.15 (1.05, 1.27)	1.20 (1.09, 1.32)	<0.001	
 Model 25	1.00 (reference)	1.07 (0.97, 1.18)	1.13 (1.03, 1.25)	1.20 (1.09, 1.32)	1.23 (1.12, 1.36)	<0.001	
Lactose2 (% energy/d)	0.3 (0.2, 0.5)	0.8 (0.7, 1.0)	1.5 (1.3, 1.7)	2.3 (2.1, 2.6)	4.0 (3.4, 5.2)		
 Number of cases	1019	932	808	870	774		
 Person-years	619,864	614,587	598,011	592,525	579,362		
 Model 14	1.00 (reference)	0.95 (0.87, 1.04)	0.81 (0.73, 0.89)	0.86 (0.79, 0.95)	0.78 (0.71, 0.86)	<0.001	
 Model 25	1.00 (reference)	0.99 (0.90, 1.08)	0.88 (0.80, 0.98)	0.98 (0.88, 1.09)	0.93 (0.82, 1.05)	0.39	
Whole fruits2 (g/1000 kcal)	24.7 (14.2, 34.2)	62.1 (52.6, 71.7)	103.1 (92.1, 115.0)	158.1 (142.4, 176.8)	265.5 (227.1, 331.1)		
 Number of cases	949	878	824	915	837		
 Person-years	627,653	616,355	601,633	590,586	568,123		
 Model 14	1.00 (reference)	0.84 (0.77, 0.92)	0.75 (0.68, 0.82)	0.79 (0.72, 0.87)	0.72 (0.66, 0.80)	<0.001	
 Model 25	1.00 (reference)	0.95 (0.86, 1.05)	0.91 (0.82, 1.01)	1.04 (0.93, 1.17)	1.06 (0.93, 1.21)	0.07	
Fruit juice2 (g/1000 kcal)	0.1 (0.0, 0.3)	6.6 (4.6, 9.0)	19.9 (15.5, 25.6)	49.0 (40.1, 59.0)	108.3 (86.3, 145.5)		
 Number of cases	903	929	877	828	866		
 Person-years	573,433	610,005	604,668	606,112	610,132		
 Model 14	1.00 (reference)	1.00 (0.91, 1.09)	0.95 (0.87, 1.04)	0.85 (0.77, 0.93)	0.88 (0.80, 0.96)	<0.001	
 Model 25	1.00 (reference)	1.02 (0.93, 1.12)	1.01 (0.92, 1.11)	0.94 (0.85, 1.04)	1.01 (0.91, 1.11)	0.74	
Sugar-sweetened beverages2,7 (g/1000 kcal)	0.0 (0.0, 1.1)	12.4 (7.6, 18.5)	47.2 (34.9, 63.5)	136.4 (107.9, 170.0)	351.2 (268.5, 492.5)		
 Number of cases	925	889	850	861	878		
 Person-years	588,455	613,034	602,397	596,912	603,552		
 Model 14	1.00 (reference)	0.95 (0.86, 1.04)	0.92 (0.83, 1.01)	0.94 (0.86, 1.04)	1.04 (0.94, 1.14)	0.07	
 Model 25	1.00 (reference)	0.98 (0.89, 1.08)	0.97 (0.88, 1.07)	0.98 (0.88, 1.08)	1.03 (0.93, 1.15)	0.29	
Confectionary2 (g/1000 kcal)	6.1 (2.8, 8.8)	16.2 (13.8, 18.6)	26.2 (23.5, 29.1)	39.4 (35.6, 44.0)	66.2 (56.4, 83.2)		
 Number of cases	806	929	899	888	881		
 Person-years	569,096	610,822	622,004	614,400	588,027		
 Model 14	1.00 (reference)	1.06 (0.97, 1.17)	1.00 (0.91, 1.10)	1.01 (0.91, 1.11)	1.01 (0.92, 1.11)	0.82	
 Model 25	1.00 (reference)	1.10 (1.00, 1.21)	1.05 (0.95, 1.15)	1.06 (0.96, 1.17)	1.07 (0.96, 1.19)	0.49	
Abbreviations: CI, confidence interval; HR, hazard ratio; NSAID, nonsteroidal anti-inflammatory drug.

Analyses were performed excluding participants with missing data on any of the covariates.

1 Participants were categorized into quintiles of sugar or food intake by sex for the analyses of the corresponding sugar type and food.

2 Median and interquartile range of intake.

3 “Total sugars” represent the sum of the consumption of the following saccharides: glucose, fructose, galactose, sucrose, maltose, and lactose.

4 Adjusted for race/ethnicity, sex, and age at cohort entry.

5 Additionally adjusted for body mass index, alcohol consumption, smoking status, number of pack-years, physical activity, education, family history of colorectal cancer, history of polyps of intestines, history of diabetes, use of exogenous female hormones (women only), NSAID use, total energy intake, and intake of saturated fatty acids, n–3 polyunsaturated fatty acids, vitamin D, calcium, dietary fiber, and folate.

6 “Total fructose” represents the sum of the consumption of fructose and half of sucrose.

7 Sugar-sweetened beverages include regular sodas, fruit drinks, and coffee and tea with added sugar.

In the analysis by CRC subsites, intakes of total sugars, total fructose, glucose, fructose, and maltose were positively associated with colon cancer risk but not with rectal cancer risk (P for heterogeneity: 0.001, 0.001, 0.005, 0.01, and 0.28, respectively) (Table 3). Fruit juices were inversely associated (P = 0.02) with rectal cancer but not with colon cancer (P = 0.36, P for heterogeneity < 0.001).TABLE 3 HRs and 95% CIs of colon or rectal cancer according to quintiles of sugar and food intake (n = 155,536)

TABLE 3No. of cases	Colon cancer	Proximal colon cancer	Distal colon cancer	Rectal cancer	
3351	2074	1173	1014	
HR (95%CI)1	HR (95%CI)1	HR (95%CI)1	HR (95%CI)1	
Total sugars2,3	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.04 (0.93, 1.16)	1.02 (0.88, 1.17)	1.07 (0.89, 1.28)	1.07 (0.89, 1.28)	
 Q3	1.11 (0.99, 1.25)	1.10 (0.95, 1.27)	1.15 (0.95, 1.39)	0.94 (0.77, 1.15)	
 Q4	1.20 (1.06, 1.35)	1.18 (1.01, 1.37)	1.23 (1.01, 1.50)	0.93 (0.75, 1.15)	
 Q5	1.16 (1.01, 1.32)	1.16 (0.98, 1.37)	1.16 (0.93, 1.44)	1.11 (0.88, 1.39)	
 P for trend	0.01	0.03	0.10	0.67	
 P for heterogeneity (colon vs. rectal, proximal vs. distal)	0.001	0.002			
Total fructose2,4	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	0.94 (0.84, 1.05)	0.95 (0.82, 1.10)	0.93 (0.77, 1.11)	0.97 (0.81, 1.17)	
 Q3	1.09 (0.98, 1.22)	1.11 (0.96, 1.28)	1.08 (0.90, 1.30)	0.89 (0.73, 1.09)	
 Q4	1.12 (1.00, 1.26)	1.10 (0.95, 1.28)	1.22 (1.01, 1.47)	0.90 (0.73, 1.10)	
 Q5	1.15 (1.01, 1.30)	1.16 (0.99, 1.36)	1.14 (0.93, 1.41)	1.00 (0.80, 1.25)	
 P for trend	0.002	0.02	0.04	0.92	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.001	0.01			
Added sugars2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	0.99 (0.89, 1.10)	0.98 (0.85, 1.12)	1.04 (0.87, 1.24)	1.09 (0.90, 1.32)	
 Q3	0.97 (0.87, 1.08)	0.98 (0.86, 1.13)	0.95 (0.79, 1.14)	1.05 (0.86, 1.28)	
 Q4	0.95 (0.85, 1.06)	0.95 (0.82, 1.09)	0.95 (0.78, 1.15)	0.97 (0.79, 1.19)	
 Q5	1.03 (0.92, 1.16)	1.03 (0.89, 1.20)	1.04 (0.86, 1.27)	1.06 (0.86, 1.31)	
 P for trend	0.65	0.68	0.87	0.89	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.54	0.63			
Glucose2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.05 (0.94, 1.18)	1.10 (0.95, 1.27)	0.99 (0.82, 1.19)	0.92 (0.76, 1.11)	
 Q3	1.22 (1.09, 1.36)	1.17 (1.02, 1.36)	1.31 (1.09, 1.56)	0.98 (0.81, 1.19)	
 Q4	1.23 (1.10, 1.38)	1.22 (1.05, 1.42)	1.27 (1.05, 1.54)	0.92 (0.75, 1.13)	
 Q5	1.26 (1.11, 1.42)	1.26 (1.07, 1.48)	1.28 (1.04, 1.57)	1.02 (0.82, 1.27)	
 P for trend	<0.001	0.005	0.01	0.73	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.005	0.09			
Fructose2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.02 (0.91, 1.13)	1.01 (0.88, 1.17)	1.02 (0.85, 1.22)	0.96 (0.80, 1.17)	
 Q3	1.10 (0.98, 1.23)	1.13 (0.98, 1.31)	1.08 (0.89, 1.30)	1.04 (0.86, 1.26)	
 Q4	1.20 (1.07, 1.35)	1.17 (1.01, 1.36)	1.30 (1.07, 1.57)	0.94 (0.76, 1.16)	
 Q5	1.17 (1.03, 1.33)	1.16 (0.99, 1.37)	1.20 (0.97, 1.48)	1.04 (0.83, 1.30)	
 P for trend	0.003	0.04	0.02	0.78	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.01	0.0496			
Galactose2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	0.90 (0.81, 1.00)	0.86 (0.75, 0.98)	0.96 (0.81, 1.15)	0.82 (0.68, 0.99)	
 Q3	0.98 (0.88, 1.09)	0.88 (0.76, 1.01)	1.15 (0.96, 1.39)	0.76 (0.62, 0.94)	
 Q4	0.86 (0.77, 0.97)	0.79 (0.68, 0.92)	0.99 (0.81, 1.22)	0.84 (0.68, 1.05)	
 Q5	0.90 (0.79, 1.02)	0.81 (0.69, 0.95)	1.05 (0.84, 1.31)	0.93 (0.73, 1.17)	
 P for trend	0.19	0.07	0.85	0.64	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.048	0.09			
Sucrose2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.04 (0.93, 1.16)	1.05 (0.92, 1.22)	1.05 (0.88, 1.25)	1.03 (0.85, 1.24)	
 Q3	1.08 (0.96, 1.20)	1.12 (0.97, 1.29)	1.04 (0.86, 1.25)	0.93 (0.76, 1.13)	
 Q4	0.99 (0.88, 1.11)	1.00 (0.86, 1.16)	1.01 (0.83, 1.23)	0.94 (0.76, 1.16)	
 Q5	1.06 (0.94, 1.19)	1.10 (0.94, 1.28)	1.01 (0.82, 1.24)	1.03 (0.83, 1.28)	
 P for trend	0.61	0.44	0.97	0.99	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.04	0.002			
Maltose2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.06 (0.94, 1.18)	1.13 (0.97, 1.30)	0.90 (0.75, 1.09)	1.13 (0.92, 1.38)	
 Q3	1.12 (1.00, 1.25)	1.10 (0.95, 1.28)	1.11 (0.92, 1.33)	1.22 (1.00, 1.49)	
 Q4	1.21 (1.08, 1.35)	1.23 (1.07, 1.42)	1.14 (0.95, 1.37)	1.22 (1.00, 1.49)	
 Q5	1.27 (1.13, 1.42)	1.30 (1.13, 1.50)	1.18 (0.98, 1.42)	1.15 (0.94, 1.42)	
 P for trend	<0.001	<0.001	0.01	0.20	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.28	0.44			
Lactose2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.01 (0.90, 1.12)	0.96 (0.84, 1.10)	1.08 (0.91, 1.29)	0.92 (0.76, 1.11)	
 Q3	0.90 (0.80, 1.01)	0.86 (0.74, 0.99)	0.97 (0.80, 1.18)	0.85 (0.69, 1.04)	
 Q4	1.01 (0.90, 1.14)	0.98 (0.84, 1.14)	1.02 (0.83, 1.26)	0.93 (0.74, 1.16)	
 Q5	0.96 (0.83, 1.11)	0.94 (0.79, 1.12)	1.01 (0.79, 1.29)	0.85 (0.65, 1.11)	
 P for trend	0.72	0.82	0.81	0.41	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.002	0.02			
Whole fruits2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	0.89 (0.79, 1.00)	0.88 (0.76, 1.02)	0.84 (0.69, 1.01)	1.14 (0.95, 1.38)	
 Q3	0.93 (0.82, 1.05)	0.86 (0.74, 1.00)	1.03 (0.85, 1.25)	0.84 (0.67, 1.05)	
 Q4	1.05 (0.93, 1.20)	1.03 (0.88, 1.21)	1.06 (0.85, 1.31)	1.00 (0.79, 1.27)	
 Q5	1.04 (0.89, 1.21)	0.97 (0.80, 1.18)	1.14 (0.88, 1.47)	1.13 (0.85, 1.49)	
 P for trend	0.11	0.46	0.08	0.45	
 P for heterogeneity (colon vs. rectal or proximal or distal)	<0.001	<0.001			
Fruit juice 2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.04 (0.93, 1.15)	1.01 (0.88, 1.16)	1.10 (0.91, 1.31)	0.96 (0.80, 1.16)	
 Q3	1.03 (0.93, 1.15)	0.97 (0.84, 1.12)	1.17 (0.97, 1.40)	0.95 (0.78, 1.15)	
 Q4	0.95 (0.85, 1.06)	0.95 (0.83, 1.10)	0.94 (0.78, 1.15)	0.90 (0.74, 1.09)	
 Q5	1.08 (0.96, 1.20)	1.03 (0.89, 1.18)	1.24 (1.03, 1.49)	0.79 (0.64, 0.97)	
 P for trend	0.36	0.70	0.12	0.02	
 P for heterogeneity (colon vs. rectal or proximal or distal)	<0.001	0.58			
Sugar-sweetened beverages2,5					
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	0.97 (0.87, 1.08)	0.94 (0.82, 1.07)	1.02 (0.85, 1.23)	1.05 (0.86, 1.28)	
 Q3	0.99 (0.89, 1.11)	0.96 (0.84, 1.10)	1.03 (0.86, 1.25)	0.91 (0.74, 1.12)	
 Q4	0.98 (0.87, 1.09)	0.90 (0.78, 1.04)	1.11 (0.91, 1.34)	0.99 (0.80, 1.21)	
 Q5	1.05 (0.93, 1.18)	0.98 (0.84, 1.14)	1.15 (0.94, 1.41)	0.99 (0.80, 1.23)	
 P for trend	0.23	0.93	0.12	0.93	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.17	0.004			
Confectionary2	
 Q1	1.00 (reference)	1.00 (reference)	1.00 (reference)	1.00 (reference)	
 Q2	1.10 (0.98, 1.23)	1.10 (0.95, 1.27)	1.13 (0.94, 1.36)	1.10 (0.90, 1.34)	
 Q3	1.07 (0.95, 1.19)	1.07 (0.92, 1.23)	1.12 (0.93, 1.35)	0.96 (0.78, 1.18)	
 Q4	1.05 (0.93, 1.18)	1.08 (0.93, 1.25)	1.01 (0.83, 1.23)	1.12 (0.91, 1.38)	
 Q5	1.07 (0.95, 1.21)	1.07 (0.92, 1.25)	1.13 (0.92, 1.39)	1.09 (0.87, 1.36)	
 P for trend	0.60	0.62	0.55	0.44	
 P for heterogeneity (colon vs. rectal or proximal or distal)	0.73	0.97			
Abbreviations: CI, confidence interval; CRC, colorectal cancer; HR, hazard ratio; NSAID, nonsteroidal anti-inflammatory drug.

Analyses were performed excluding participants with missing data on any of the covariates. CRC cases with subsites other than the one of interest were censored at the time of the CRC diagnosis.

1 Adjusted for race/ethnicity, sex, age at cohort entry, body mass index, alcohol consumption, smoking status, number of pack-years, physical activity, education, family history of CRC, history of polyps of intestines, history of diabetes, use of exogenous female hormones (women only), NSAID use, total energy intake, and intake of saturated fatty acids, n–3 polyunsaturated fatty acids, vitamin D, calcium, dietary fiber, and folate.

2 Participants were categorized into quintiles of sugar or food intake by sex for the analyses of the corresponding sugar type and food.

3 “Total sugars” represent the sum of the consumption of the following saccharides: glucose, fructose, galactose, sucrose, maltose, and lactose.

4 “Total fructose” represents the sum of the consumption of fructose and half of sucrose.

5 Sugar-sweetened beverages include regular sodas, fruit drinks, and coffee and tea with added sugar.

When we stratified the analyses by relevant risk factors, the positive association of total sugar intake was suggested to be stronger among participants who were aged 45 to 54 y at cohort entry (HR: 1.41; 95% CI: 1.11, 1.79; P for trend = 0.001), compared to those aged 55 to 64 or 65 to 75 y at baseline (P for heterogeneity = 0.19) (Supplemental Table 4). The mean and SD of age at diagnosis were 62.7 (10.5) for those aged 45 to 54 y, 71.5 (10.2) for those aged 55 to 64 y, and 79.3 (8.7) for those aged 65 to 75 y. This pattern was especially suggested for maltose (P for heterogeneity by age at cohort entry = 0.06) (Supplemental Tables 5, 6, 7, and 8). When we additionally examined the stratified analysis for food sources of sugar, intake of confectionary (HR: 1.32, 95% CI: 1.04, 1.68; P for trend = 0.05; P for heterogeneity = 0.04) and SSBs (HR: 1.35; 95% CI: 1.07, 1.70; P for trend = 0.01; P for heterogeneity = 0.23) also appeared to be stronger in the group 45 to 54 y old at cohort entry (Supplemental Tables 9 and 10) but not for whole fruits and fruit juice (Supplemental Tables 11 and 12). Moreover, this trend of increased risk among younger participants was more apparent for colon cancer, whereas the association with confectionary intake was more apparent with rectal cancer.

For total sugars, the positive association was observed among participants who were followed for over 21 y (HR: 1.33; 95% CI: 0.93, 1.89; P for trend = 0.04; P for heterogeneity < 0.001), and this pattern was similar for total fructose, glucose, and fructose (Supplemental Tables 4, 5, 6, and 7). In contrast, for maltose, the association was stronger among participants who were followed for ≤10 y or 11 to 20 y (HR: 1.25; 95% CI: 1.07, 1.46; P for trend < 0.001; P for heterogeneity = 0.03) (Supplemental Table 8). There was no heterogeneity by personal history of diabetes (Supplemental Tables 4, 5, 6, 7, and 8), whereas the positive association of total fructose intake was suggested to be stronger among participants who do not drink alcohol (P for heterogeneity = 0.06) (Supplemental Table 5). For confectionary intake, the positive association was suggested to be stronger among those who do not drink alcohol or were overweight (≥25 kg/m2) (P for heterogeneity = 0.02 or 0.005, respectively) (Supplemental Table 9). The results also suggested a stronger positive association of SSB intake among those who were ever smokers (P for heterogeneity = 0.03) (Supplemental Table 10).

When we analyzed the association stratified by race and ethnicity, a positive association with CRC was suggested among Latinos for total sugars, total fructose, added sugars, glucose, fructose, maltose, whole fruits, and confectionaries, although heterogeneity was nonsignificant. For maltose, a positive association was also suggested among Native Hawaiians, and African Americans showed a positive association for confectionaries (Supplemental Table 13). No clear differences were observed between colon and rectal cancer in this stratified analysis (Supplemental Tables 14 and 15).

When we examined the association using the imputed missing data, the increased risks of CRC were similar or slightly higher for intake of total sugar, total fructose, glucose, fructose, maltose, whole fruits, and confectionaries. Similarly, when we excluded the participants who had a CRC diagnosis within 3 y after baseline, the associations with CRC for these sugars and foods remained or became slightly stronger.

Discussion

In this large population-based multiethnic cohort study, we observed an increased risk of CRC with higher intakes of total sugar, total fructose, glucose, fructose, and maltose. These associations appeared to be limited to the colon and younger participants (i.e., 45–54 y at baseline) and were more apparent in Latinos. Among younger participants, the increased risk of colon cancer may be related to intake of SSBs. No clear association of other sugars or other sugar food sources with CRC risk was observed.

The increased risk of CRC with total sugar intake found in the current study was in contrast to the previous literature that overall showed no association [7,8,13]. In 2 of these past studies [7,8] conducted in the United States, the total sugar intake was similar to that of the MEC study participants, whereas it was lower in a Japanese study [13]. The inconsistent results may reflect differences in populations, diet assessment quality, or adjustment for risk factors. The importance of risk factor adjustment was suggested in our study by the change in direction observed in the association with CRC between the minimally- and fully-adjusted models (Table 2). The increased risk of CRC was detected only after full adjustment for baseline risk factors. Indeed, in our study, participants in the highest quintile of total sugar intake were less likely to be current smokers, to be obese, or to report a history of diabetes (Table 1). These characteristics, which increase the risk of CRC, may act as confounders of the association between sugar intake and CRC.

Our finding of an association between CRC and total fructose and fructose intake is consistent with those of several previous studies [10,12]. However, other studies [8,9,11] failed to observe such an association. Differences of adjustment for confounders may also explain these discrepancies. SSB consumption, which is one of the main food sources of fructose, has been reported to be positively associated with proximal colon cancer risk in several past studies [31,32]. In our results, although SSB consumption did not show any association with CRC either overall or by anatomical subsite, we observed an increased risk of colon cancer among younger participants who were aged 45 to 54 y at baseline. This result is consistent with a previous cohort study in the United States that reported that SSB intake was associated with an increase in the incidence of proximal colon cancer [32]. Whole fruit intake showed no increased CRC risk overall, although an association was suggested among Latinos. However, the increased CRC risk for SSB or whole fruit intake was limited to tumors located the colon or a specific generation and race. In contrast, fructose intake from sources other than whole fruits showed an increased CRC risk among all participants. Therefore, our findings of positive associations with total fructose and individual fructose intakes might be caused by the increased risk not only with higher whole fruit intake and SSBs but also due to other components (e.g., sweet food including high-fructose sweeteners). However, according to 2 previous meta-analyses of food groups and CRC risk [18,33], there was a nonlinear association of fruit intake with the risk of CRC. An increased risk was observed in lower intake of fruit [18] and a decreased risk was detected in moderate intake [33], but higher intake did not show more reduction of risk [18,33]. Therefore, the impact of excessive intake of whole fruits on the risk of CRC may need to be carefully assessed regarding heterogeneity among racial and ethnic groups.

Given that intake of maltose was not highly correlated with total sugar intake, our finding of an association with CRC risk for maltose was unexpected. Moreover, Latinos showed a clear positive association while Japanese Americans did not, despite their intake levels of maltose being similar. In previous literature, a case-control study in Iran suggested a positive association between maltose intake and CRC risk [34], but no association was reported in a prospective cohort study in Japan [13]. When we investigated the food sources of maltose in our data, beer was a greater contributor in Latinos than in Japanese Americans. Indeed, when we analyzed the association of maltose intake with CRC stratified by beer consumption level, the subgroup who drank beer showed a higher HR for maltose (Q5 compared with Q1: 1.36; 95% CI: 1.09, 1.69) than the non–beer-drinkers (HR: 1.20; 95% CI: 1.08, 1.34), although the P for heterogeneity was nonsignificant (P = 0.38). However, the contribution of regular beer consumption to maltose intake was limited (in the highest quintiles of maltose intake: 6.2% of the maltose intake in Latinos and ∼5% in African Americans, Native Hawaiians, and Whites, and <3% in Japanese Americans; data not shown in the tables), and we adjusted for alcohol consumption as a potential confounder in the model. When we examined the association with CRC risk stratified by follow-up periods, total sugar intake showed a positive association in participants who were followed up the longest (>21 y), whereas maltose intake was associated with CRC for the shorter follow-up categories (≤10 and 11–20 y). These differences in findings suggest that the associations of CRC with total sugar and maltose intakes, if causal, may reflect different underlying mechanisms. However, confounding cannot be excluded, especially for the association with maltose, which was unexpected.

Latinos showed clear positive associations of glucose and fructose intake with CRC risk. However, when we also surveyed the food sources of glucose and fructose, there were no apparent differences among races and ethnicities. Although we could not clarify a reason for this discrepancy of the association with CRC risk, it might be affected by genetic backgrounds or other unmeasured confounding. Further investigations are needed in other Latino populations to examine whether the increased risk of CRC with higher sugar intake can be reproduced.

The association of glycemic load (GL) with CRC risk in this MEC study was previously examined [35]. In that study, increasing GL was unexpectedly associated with a decreased risk of CRC. The results also suggested that rice intake, which is the main food source of starch, was the major contributor to GL in the study population. In the present analysis, we focused on the impact of simple sugar intakes on CRC; therefore, we did not include GL as well as glycemic index. However, because the previous analysis was conducted with a much shorter follow-up (8 y), we further examined the association of GL using the current data set with an average follow-up of 19 y. The results showed no significant association with CRC for both GL and energy-adjusted GL (data not shown).

Excessive habitual sugar intake induces weight gain and excess insulin secretion. Postprandial uptake of blood glucose induces insulin secretion, and insulin stimulates an increase in insulin-like growth factor-I (IGF-I). Insulin and IGF-I facilitate cell division and growth and inhibit apoptosis [36,37]. Insulin secretion directly responds to glucose intake; glucose is derived from sucrose, maltose, and lactose, as well as absorbed as glucose as a monosaccharide. In contrast, fructose intake has a more complex pathway to hyperinsulinemia caused by decreased insulin sensitivity due to excessive de novo lipogenesis [14,15]. Our findings for total sugar, total fructose, glucose, fructose, and maltose are in line with these proposed mechanisms. Moreover, intake of high-fructose corn syrup may enhance colon tumor growth via activation of glycolysis and synthesis of fatty acids within the tumors [19]. Our findings of an increased risk of colon cancer with SSB intake among younger participants is consistent with this mechanism. However, the results for sucrose (a disaccharide of glucose and fructose) and added sugar intakes, which both showed no association with the risk of CRC, were inconsistent with the experimental results. With regard to establishing a health recommendation for sugar and sugary food intake to prevent CRC, it may be essential to focus on total intake including not only added sugars but also naturally occurring sugars.

The strengths of our study include a large sample size, long follow-up, and racial/ethnic diversity, allowing the comparison of the association with the risk of CRC among 5 racial/ethnic groups and by tumor anatomical subsite with sufficient power. Nevertheless, several limitations should be considered. Measurement error in dietary intake assessment may have caused attenuations of the associations due to misclassification on exposures. The calibration study of the QFFQ used in the MEC study showed moderate validity for carbohydrate intake estimates [21]. Additionally, only exposure at baseline was considered in this analysis. Further, we could not exclude confounding by unmeasured risk factors, although we carefully selected our covariates in an attempt to minimize confounding in our analyses.

In conclusion, this large population-based multiethnic cohort study showed that higher intake of total sugar, total fructose, glucose, fructose, and maltose were associated with an increased risk of CRC. There was the suggestion that these associations may be limited to the colon and younger participants, possibly reflecting an association with intake of SSBs. Among the 5 racial and ethnic groups, Latinos were suggested to be at particularly increased risk.

Author contributions

The authors’ responsibilities were as follows – RK, MI, ST, NS, LRW, LLM: designed research; RK, S-YP, LRW, CAH, LLM: conducted research; RK, S-YP: analyzed data; RK: drafted the manuscript with input from other coauthors; RK, S-YP, YO, MI, ST, NS, MI, CAH, LRW, LLM: were involved in the interpretation of the results; RK, LLM: had primary responsibility for final content; and all authors: read and approved the final manuscript.

Conflict of interest

The authors report no conflicts of interest.

Funding

This work was supported by grant U01 CA164973 from the United States National Institutes of Health and the Japan Society for the Promotion of Science (JSPS KAKENHI Grant Number 22K17382).

Data availability

Data described in the manuscript, code book, and analytic code will be made available upon request, pending application to and approval by the Multiethnic Cohort Research Committee (https://www.uhcancercenter.org/for-researchers/mec-data-sharing).

Appendix A Supplementary data

The following is the Supplementary data to this article:Multimedia component 1

Multimedia component 1

Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.tjnut.2024.05.016.
==== Refs
References

1 Sung H. Ferlay J. Siegel R.L. Laversanne M. Soerjomataram I. Jemal A. Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries CA Cancer J Clin 71 3 2021 May 209 249 10.3322/caac.2166 33538338
2 Te Morenga L. Mallard S. Mann J. Dietary sugars and body weight: systematic review and meta-analyses of randomised controlled trials and cohort studies BMJ 346 2012 e7492 10.1136/bmj.e7492
3 Stanhope K.L. Sugar consumption, metabolic disease and obesity: the state of the controversy Crit. Rev. Clin. Lab. Sci. 53 1 2016 52 67 10.3109/10408363.2015.1084990 26376619
4 Yoon Y.S. Keum N. Zhang X. Cho E. Giovannucci E.L. Hyperinsulinemia, insulin resistance and colorectal adenomas: a meta-analysis Metabolism 64 10 2015 1324 1333 10.1016/j.metabol.2015.06.013 26169471
5 Giovannucci E. Insulin, insulin-like growth factors and colon cancer: a review of the evidence J. Nutr. 131 Suppl 11 2001 3109S 3120S 10.1093/jn/131.11.3109S 11694656
6 World Cancer Research Fund, American Institute for Cancer Research Continuous Update Project Report: Diet, Nutrition, Physical Activity and Colorectal Cancer 2018
7 Kabat G.C. Shikany J.M. Beresford S.A. Caan B. Neuhouser M.L. Tinker L.F. Dietary carbohydrate, glycemic index, and glycemic load in relation to colorectal cancer risk in the Women’s Health Initiative Cancer Causes Control 19 10 2008 1291 1298 10.1007/s10552-008-9200-3 18618276
8 Tasevska N. Jiao L. Cross A.J. Kipnis V. Subar A.F. Hollenbeck A. Sugars in diet and risk of cancer in the NIH-AARP Diet and Health Study Int. J. Cancer 130 1 2012 159 169 10.1002/ijc.25990 21328345
9 McCarl M. Harnack L. Limburg P.J. Anderson K.E. Folsom A.R. Incidence of colorectal cancer in relation to glycemic index and load in a cohort of women Cancer Epidemiol. Biomarkers Prev. 15 5 2006 892 896 10.1158/1055-9965.EPI-05-0700 16702366
10 Michaud D.S. Fuchs C.S. Liu S. Willett W.C. Colditz G.A. Giovannucci E. Dietary glycemic load, carbohydrate, sugar, and colorectal cancer risk in men and women Cancer Epidemiol. Biomarkers Prev. 14 1 2005 138 147 10.1158/1055-9965.138.14.1 15668487
11 Makarem N. Bandera E.V. Lin Y. Jacques P.F. Hayes R.B. Parekh N. Consumption of sugars, sugary foods, and sugary beverages in relation to adiposity-related cancer risk in the Framingham Offspring Cohort (1991-2013) Cancer Prev. Res. (Phila). 11 6 2018 347 358 10.1158/1940-6207.CAPR-17-0218 29674390
12 Higginbotham S. Zhang Z.F. Lee I.M. Cook N.R. Giovannucci E. Buring J.E. Dietary glycemic load and risk of colorectal cancer in the Women’s Health Study J. Natl. Cancer Inst. 96 3 2004 229 233 10.1093/jnci/djh020 14759990
13 Kanehara R. Katagiri R. Goto A. Yamaji T. Sawada N. Iwasaki M. Sugar intake and colorectal cancer risk: a prospective Japanese cohort study Cancer Sci 114 6 2023 2584 2595 10.1111/cas.15766 36851860
14 Stanhope K.L. Schwarz J.M. Keim N.L. Griffen S.C. Bremer A.A. Graham J.L. Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/obese humans J. Clin. Invest. 119 5 2009 1322 1334 10.1172/JCI37385 19381015
15 Elliott S.S. Keim N.L. Stern J.S. Teff K. Havel P.J. Fructose, weight gain, and the insulin resistance syndrome Am. J. Clin. Nutr. 76 5 2002 911 922 10.1093/ajcn/76.5.911 12399260
16 Evans R.C. Fear S. Ashby D. Hackett A. Williams E. Van Der Vliet M. Diet and colorectal cancer: an investigation of the lectin/galactose hypothesis Gastroenterology 122 7 2002 1784 1792 10.1053/gast.2002.33659 12055585
17 Wijnands M.V. Appel M.J. Hollanders V.M. Woutersen R.A. A comparison of the effects of dietary cellulose and fermentable galacto-oligosaccharide, in a rat model of colorectal carcinogenesis: fermentable fibre confers greater protection than non-fermentable fibre in both high and low fat backgrounds Carcinogenesis 20 4 1999 651 656 10.1093/carcin/20.4.651 10223195
18 Vieira A.R. Abar L. Chan D.S.M. Vingeliene S. Polemiti E. Stevens C. Foods and beverages and colorectal cancer risk: a systematic review and meta-analysis of cohort studies, an update of the evidence of the WCRF-AICR Continuous Update Project Ann. Oncol. 28 8 2017 1788 1802 10.1093/annonc/mdx171 28407090
19 Goncalves M.D. Lu C. Tutnauer J. Hartman T.E. Hwang S.K. Murphy C.J. High-fructose corn syrup enhances intestinal tumor growth in mice Science 363 6433 2019 1345 1349 10.1126/science.aat8515 30898933
20 Kolonel L.N. Henderson B.E. Hankin J.H. Nomura A.M. Wilkens L.R. Pike M.C. A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics Am. J. Epidemiol. 151 4 2000 346 357 10.1093/oxfordjournals.aje.a010213 10695593
21 Stram D.O. Hankin J.H. Wilkens L.R. Pike M.C. Monroe K.R. Park S. Calibration of the dietary questionnaire for a multiethnic cohort in Hawaii and Los Angeles Am. J. Epidemiol. 151 4 2000 358 370 10.1093/oxfordjournals.aje.a010214 10695594
22 Wilkens L.R. Kang M. Park S.Y. Sawyer A. Boushey C.J. Au D.L.M. Validation and comparison of nutrient intakes from two dietary questionnaires developed for the Multiethnic Cohort Study J. Acad. Nutr. Diet. 123 4 2023 655 663.e1 10.1016/j.jand.2022.09.016 36191897
23 US Department of Agriculture USDA Nutrient Database for Standard Reference: Release 10 1993 Nutrient Data Laboratory, Agricultural Research Service Riverdale, MD
24 US Department of Agriculture Nutrient Database for Individual Food Intake Surveys and Data Sets Used to Create It: Release 7. Report from the National Technical Information Service 1994 National Technical Information Service Springfield, VA
25 Kagawa R. Standard Tables of Food Composition in Japan 1995 Women’s University Tokyo
26 Pennington J. Bowes & Church’s Food Values of Portions Commonly Used 17th ed. 1998 Lippincott Philadelphia, PA
27 Chavez M.M. Hernandez M. Roldan J.A. Tablas de uso pactico del: valor nutritivo de los alimentos de mayor consume en Mexico 1992 Instituto Nacional de la Nutricion Salvador Zubiran Mexico City
28 Pollock N.K. Bundy V. Kanto W. Davis C.L. Bernard P.J. Zhu H. Greater fructose consumption is associated with cardiometabolic risk markers and visceral adiposity in adolescents J. Nutr. 142 2 2012 251 257 10.3945/jn.111.150219 22190023
29 Willett W. Nutritional Epidemiology 3rd ed. 2012 Oxford University Press New York
30 Lunn M. McNeil D. Applying Cox regression to competing risks Biometrics 51 2 1995 524 532 10.2307/2532940 7662841
31 Leung C.Y. Abe S.K. Sawada N. Ishihara J. Takachi R. Yamaji T. Sugary drink consumption and subsequent colorectal cancer risk: the Japan public health center-based prospective cohort study Cancer Epidemiol. Biomarkers Prev. 30 4 2021 782 788 10.1158/1055-9965.EPI-20-1364 33563648
32 Yuan C. Joh H.K. Wang Q.L. Zhang Y. Smith-Warner S.A. Wang M. Sugar-sweetened beverage and sugar consumption and colorectal cancer incidence and mortality according to anatomic subsite Am. J. Clin. Nutr. 115 6 2022 1481 1489 10.1093/ajcn/nqac040 35470384
33 Schwingshackl L. Schwedhelm C. Hoffmann G. Knüppel S. Laure Preterre A. Iqbal K. Food groups and risk of colorectal cancer Int. J. Cancer 142 9 2018 1748 1758 10.1002/ijc.31198 29210053
34 Jonoush M. Fathi S. Hassanpour Ardekanizadeh N. Khalatbari Mohseni G. Majidi N. Keshavarz S.A. The association between different types of dietary carbohydrates and colorectal cancer: a case-control study Front. Nutr. 9 2022 898337 10.3389/fnut.2022.898337
35 Howarth N.C. Murphy S.P. Wilkens L.R. Henderson B.E. Kolonel L.N. The association of glycemic load and carbohydrate intake with colorectal cancer risk in the Multiethnic Cohort Study Am. J. Clin. Nutr. 88 4 2008 1074 1082 10.1093/ajcn/88.4.1074 18842796
36 Kaaks R. Lukanova A. Energy balance and cancer: the role of insulin and insulin-like growth factor-I Proc. Nutr. Soc. 60 1 2001 91 106 10.1079/pns200070 11310428
37 Pollak M. Insulin and insulin-like growth factor signalling in neoplasia Nat. Rev. Cancer 8 12 2008 915 928 10.1038/nrc2536 19029956
