Dietary Sugar and Sweetened Beverage Intake Increases Inflammatory Bowel Disease Risk: A Systematic Review and Meta‐Analysis

ABSTRACT

Background and Aims

The impact of dietary sugar intake on the risk of developing inflammatory bowel disease is unclear, with inconsistent findings across studies. The aim of this systematic review and meta‐analysis was to clarify how sugar consumption contributes to the risk of developing inflammatory bowel disease (IBD) using the most recently available data.

Methods

A library informationist retrieved relevant articles from PubMed, EMBASE, CINAHL, Cochrane Central, Web of Science, and Scopus. Two independent reviewers screened the abstracts and full texts, yielding 45 studies for inclusion. Meta‐analyses estimated odd ratios using random effect models.

Results

11 prospective and 34 retrospective studies reported data on sugar intake and IBD risk. Pooled analysis showed that added sugar intake was associated with increased risk of Crohn's disease (OR 1.66; 95% Cl 1.21–2.29; n = 523,730; 14 studies) and ulcerative colitis (OR 1.59; 95% CI 1.25–2.02; n = 787,228; 18 studies). Similarly, soda/sweetened beverage intake was associated with increased risk of Crohn's disease (OR 1.58; 95% CI 1.18–2.12; n = 328,716; 12 studies) and ulcerative colitis (OR 1.72; 95% CI 1.23–2.391; n = 328,642; 13 studies).

Conclusions

Sugar and soda/sweetened beverage intake were associated with an increased risk of developing both Crohn's disease and ulcerative colitis. Although additional prospective investigation is warranted, current data suggest that reduction of sugar consumption might help reduce the risk of inflammatory bowel disease.

Abbreviations

BMI

Body Mass Index

CD

Crohn's disease

CI

Confidence Interval

HR

Hazard Ratio

IBD

Inflammatory Bowel Disease

IV

Inverse Variance

NOS

Newcastle Ottawa Scale

OR

Odds Ratio

RR

Relative Risk

SB

Sweetened Beverages

SE

Standard Error

SES

Socio‐Economic Status

SSB

Sugar‐Sweetened Beverages

UC

Ulcerative colitis

1. Introduction

Inflammatory bowel disease (IBD) is a chronic, immune‐mediated disorder of the gastrointestinal tract that is often characterized by abdominal pain, diarrhea, urgency, and blood in the stool. While IBD typically arises during early adulthood, its onset has become more prevalent during childhood in recent years [1]. Moreover, IBD occurrence has historically been higher in Western populations, but it has now become a global condition with incidence rising rapidly in the Eastern Hemisphere [2]. The effect of both Crohn's disease (CD) and ulcerative colitis (UC) on the quality of life in patients is significant, contributing to lost productivity, symptoms like gastrointestinal bleeding, and feelings of isolation in social settings [3]. Given the increasing incidence of, social consequences associated with, and burden on the medical system caused by IBD, it is crucial to better understand risk factors associated with IBD to better inform prevention and treatment of this illness.

Many environmental and genetic factors have been implicated in the pathogenesis of IBD, such as smoking, gut microbial dysbiosis, nutrition, and stress [4]. Dietary sugars, in particular, are linked to increasing inflammation in the body by indirectly promoting T‐cell activation [5]. Sugars are also responsible for increasing the abundance of Proteobacteria in the gut, which can lead to inflammation by introducing lipopolysaccharides in the small intestine [6]. As a result, their role has been studied in the development or exacerbation of rheumatoid arthritis, psoriasis, and other inflammatory conditions. A 2021 systematic review and meta‐analysis by Khademi et al. [7] reported on the association between sugar intake and risk of developing IBD, concluding that sugar intake, but not sugar‐sweetened beverage (SSB) or total carbohydrate consumption, is correlated with IBD. However, several studies, including two large prospective cohort studies, have been published in recent years that may contradict the findings of that review [8, 9]. Therefore, the aim of this study was to provide an updated systematic review and meta‐analysis including recent high‐quality studies to further elucidate the role of dietary sugar and sweetened beverage consumption as a risk factor for developing IBD.

2. Materials and Methods

Studies to be included in this review were identified from the electronic databases PubMed, EMBASE, CINAHL, Cochrane Central, Web of Science, and Scopus. A library informationist developed the search query with appropriate keywords including “dietary carbohydrates,” “candy,” “carbonated beverages,” “sweetening agents,” and “sweetened beverages” (Supporting Information S1: Material 1). Studies were included if they met the following criteria: prospective cohort or retrospective case‐control/cohort study design, investigation of sugar intake, and assessment of development of IBD (CD and/or UC). Sugar intake encompasses direct assessment of refined sugar intake as well as consumption of sugar‐containing foods, sweets, pastries, confectioneries, chocolates, candies, cakes, cookies, desserts, and honey. Sweetened beverages (SB) encompassed SSBs (including sodas, noncarbonated drinks, coffees, teas), artificially sweetened beverages, and/or juice. Studies were excluded if they included patients with baseline IBD (diagnosis of CD, UC, or indeterminate colitis prior to the study enrollment period for prospective studies or prior to the time of evaluated diet exposure for retrospective studies) or utilized animal models or in vitro data. Given our specific interest in sugar intake prior to the onset of IBD, studies assessing disease activity related to sugar intake were not included. Studies published in languages other than English were included.

The title and abstract of articles identified on the initial screen were assessed for inclusion by two independent reviewers. Discrepancies between reviewers were resolved by a third independent reviewer. Several articles deemed critical for inclusion through a separate directed search were referenced against those marked for inclusion through the blind screen to ensure that the search criteria were sufficiently expansive. Two independent reviewers screened the full texts for inclusion. Data were extracted from the included studies by two independent reviewers. The strength of evidence in each article was also graded by two independent reviewers using the Newcastle‐Ottawa Scale (NOS) [10].

Studies with qualitative data were included in the narrative review, while those reporting adequate numeric data to calculate the risk of IBD based on quantity or frequency of sugar consumed were included in the meta‐analyses. “Sweets” encompassed a broad range of sugar‐containing substances, including cakes, candy, chocolates, confectionaries, cookies, honey, ice cream, jam, pastries, and unspecified sweets, which were grouped together in this review given the inconsistent quantity and type of sugar in each type of sweet. Studies only reporting relative sugar intake, incomplete quantification of sugar added to beverages, sugar not separated from carbohydrate intake, sugar‐enriched dietary patterns, and those without clear comparators were excluded from meta‐analyses. The meta‐analyses estimated the odds ratio (OR) with 95% confidence intervals (CI) using random effect models. Analyses were grouped by type of sugar consumption (sugar vs. sweetened beverages), IBD subtype (CD or UC), and study design (retrospective or prospective). None of the studies solely evaluating juices or artificial sweeteners were included in the meta‐analysis, given the differing impacts of natural and artificial sugar on the gut; these studies were only included in the meta‐analysis if they also analyzed soft drink intake. Heterogeneity was qualitatively and quantitatively assessed for each outcome. The chi‐square test (p < 0.10 considered statistically significant) and I² statistic were used for quantitative assessment of heterogeneity.

3. Results

There were 4349 relevant articles (3194 after removing duplicates) identified in the initial library search, which were condensed to 86 articles after title and abstract screening (Figure 1). Forty‐five articles were ultimately included in this systematic review after full‐text screening, encompassing 15,503 total participants in 34 retrospective studies and 1,255,635 total participants in 11 prospective studies (Tables 1 and 2) [8, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53]. Twenty‐seven studies had data that were amenable to meta‐analysis [8, 9, 12, 14, 15, 17, 20, 24, 25, 26, 27, 28, 30, 31, 32, 38, 39, 40, 41, 43, 44, 45, 46, 50, 52, 53]. While not all studies reported socioeconomic status, education level, or income distribution, most did control for these variables in their analysis. Most (25, 74%) retrospective studies utilized food frequency questionnaires to gather data on dietary intake, with some (6, 18%) also including structured interviews performed by a dietician and very few (2, 6%) prospectively measuring dietary intake as a proxy for pre‐illness diet. Of the studies included in the meta‐analysis, all prospective studies [8, 9, 12, 14, 15] calculated the quantity of sugar or SB intake except 1, which addressed dietary patterns significantly enriched for sugar and SB intake [17]. Nine retrospective studies [26, 27, 28, 32, 39, 40, 42, 45, 46] assessed frequency of sugar or SB intake, 2 [25, 45] assessed quantity, and 1 [41] assessed sugar or SB‐enriched dietary patterns.

FIGURE 1.

PRISMA flow diagram of studies screened and included in the meta‐analysis.

TABLE 1.

Prospective studies assessing sugar and/or sweetened beverages and IBD risk.

Author/Year	Country	Cohort	# IBD (CD + UC)	# CD	# UC	# Controls	Time between diagnosis and data collection	Method used to assess sugar exposure (FFQ = food frequency questionnaire)	How sugar intake was reported (dietary pattern vs. direct assessment of sugar, sweetened beverage, etc)	Amount of sugar consumption	Incidence	Relative risk (95% CI)
Ananthakrishnan 2015	United States	Nurses' Health Study II	173	70	103	39,338	Prior to diagnosis	FFQ	Dietary pattern	Foods were grouped into dietary patterns. “Western diet” was most enriched for sugar and sweets.	IBD: 173/763,229 person‐years CD: 70/763,229 person‐years 12 175,813 person‐years in highest quartile of Western diet UC: 103/763,229 person‐years 21/175,813 in highest quartile of Western diet	For highest quartile in western diet: CD: HR 0.51 (0.25, 1.25); p = 0.10 UC: HR 0.83 (0.46, 1.47); p = 0.6
Chan 2014	United Kingdom, Germany, Italy, Sweden, Denmark, France, The Netherland, Greece	EPIC‐IBD	354	110	244	400,972; each case matched with 4 controls (110 for UC, 976 for CD)	Prior to diagnosis (CD: Median 5.1 years, UC: median 4.8 years)	FFQ	G/day sugar intake, grouped into quintiles	CD: 99.0 g/day (36.3–260.2) versus controls 99.7 g/day (24.7–397.2) UC: 107.0 g/day (18.8–238.5) versus controls 103.4 g/day (24.4–312.5)	IBD: 354/401,326 CD: 110/401,326 participants UC: 244/401,326 participants	Highest versus lowest quintile sugar intake: CD adjusted OR: 0.76 (0.28, 2.08); p = 0.50 UC adjusted OR: 1.12 (0.57, 2.17); p = 0.71
ElMouzan 2017	Saudi Arabia	No data	52	38	14	39	≥ 3 months prior to onset of symptoms	Direct interview + modified FFQ	Gaseous drink consumption frequency (daily, twice weekly, weekly, or monthly or less frequently)	More than once weekly gaseous drink consumption: IBD: 45 (87%); controls: 28 (74%)	No data	Highest gaseous drink intake: Unadjusted OR: 2.50 (0.86, 7.30) Adjusted: 2.07 (0.62, 6.90)
Fu 2022	UK	UK Biobank	510	143	367	120,980	Prior to diagnosis; follow‐up performed an average of 10.2 years later	24‐h diet recall	Mean amount of sweetened beverage intake (fizzy drink and squash, artificially sweetened beverage referred to low‐calorie drinks, and natural juices including pure orange juice, grapefruit juice and other pure fruit or vegetable juice)	Overall: 119.8 g sugar/day (44.4) > 1 SSB/day: 148.3 g sugar/day (48.8)	IBD: 41/100,000 person‐years CD: 12/100,000 person‐years UC: 30/100,000 person‐years	IBD: HR 1.51 (1.11, 2.05); p = 0.009 CD: HR 2.05 (1.22, 3.46) UC: HR 1.31 (0.89, 1.92)
Hart 2008	UK, Germany, Denmark, Sweden, Italy	EPIC	139	0	139	260,547	Median time between recruitment and diagnosis of 3.8 years (range 1.7–11.3 years)	FFQ (country‐specific)	Sugars, % of total energy intake split into quartiles	Quartiles with cases/controls: 1 (38/134) 2 (33/138) 3 (29/143) 4 (37/134)	IBD: 139/260,686 participants UC: 139/260,686 participants	OR trend = 0.97 (0.77–1.24); p = 0.83
Khalili 2019	Sweden	Swedish Mammography Study (SMC) and the Cohort of Swedish Men (CoSM)	492	143	349	82,550	Prior to diagnosis	FFQ, semiquantitative	Usual consumption, during the past year, of a standard glass (200 mL) of sweetened beverages in the form of soft drinks per day or week	Organized in quartiles, depending on daily consumption (0, 0.1—0.4, 0.5—0.9, and 1 servings/day).	CD: 11/100,000 person‐years UC: 28/100,000 person‐years	CD male patients (age‐adjusted): HR 1.00 (0.89–1.12) CD female patients (age‐adjusted): HR 1.00 (0.76–1.31) UC male patients(age‐adjusted): HR 0.99 (0.98–1.01) UC female patients (age‐adjusted): HR 0.89 (0.73–1.10)
Narula 2021	Argentina, Bangladesh, Brazil, Canada, Chile, China, Colombia, India, Iran, Malaysia, Palestine, Pakistan, Philippines, Poland, South Africa, Saudi Arabia, Sweden, Tanzania, Turkey, United Arab Emirates, and Zimbabwe	No data	467	90	377	115,570	≥ 1 year prior to diagnosis	FFQ (country‐specific)	Food frequency converted to daily intake and multiplied by USDA serving size	Mean ultra‐processed food servings/day: Overall: 1.8 (3.9) IBD: 4.0 (6.8) Mean ultra‐processed food g/day: Overall: 94.2 (228.3) IBD: 145.4 (223.2) Mean soft drinks servings/week: Overall: 1.2 (2.8) IBD: 1.5 (3.4) Mean refined sweetened food g/day: Overall: 92 (216.3) IBD: 100.4 (180.6)	CD: Soft drink: 17/12,106 participants versus 49/83,067 participants Sweets: 23/23,109 participants versus 16/20,812 participants UC: Soft drink: 49/12,106 participants versus 212/83,067 participants Sweets: 100/23,109 participants versus 44/20,812 participants	Soft drink: IBD: HR 1.94 (1.42,2.66) CD: HR 2.36 (1.23,4.55) UC: HR 1.84 (1.29,2.64) Sweets: IBD: HR 2.58 (1.44,4.62) CD: HR 1.17 (0.35,3.84) UC: HR 3.08 (1.57,6.05)
Peters 2022	Netherlands	No data	321	97	224	125,124	Prior to diagnosis, maximum 14‐year follow‐up	FFQ	Direct assessment and dietary pattern	Controls: 74.7 ± 45.3 g sugar/day, CD: 74.4 ± 47.8 g sugar/day, UC: 74.0 ± 43.5 g sugar/day; p = 0.973	UC: 224/125,445 participants	For dietary pattern 1 (most enriched for sugar) CD: OR 1.00 (0.90–1.11); p = 0.981 UC: OR 1.00 (0.93–1.06); p = 0.941
Racine 2016	Denmark, France, Germany, Italy, The Netherlands, Sweden, and the United Kingdom	No data	373	117	256	1490	Prior to diagnosis	FFQ	Dietary patterns: UC: Sugar and soft drinks, sugar/soft drinks/vegetables and legumes CD: Sugar and soft drinks	No data	UC: 24 incident cases (20%) versus 15 incident cases (13%)	CD: HR 1.20 (0.56–2.56) UC: HR 1.31 (0.85–2.02)
Vasseur 2021	France	No data	75	27	48	105,757	Prior to diagnosis	3 self‐administered online 24‐h dietary records collected at baseline and then every year during follow‐up; previously validated in several studies, comparing it with interviews by dieticians and biomarkers of nutritional status	Cakes, cookies, pastries (g/day), fatty and/or sweet products (g/day), and sweetened beverages including pure fruit juice (g/day)	Mean cakes, cookies, and pastries g/day: 46.11 (44.46 SD) (control), 55.26 (47.86 SD) (IBD) Mean fatty and/or sweet products g/day: 18.36 (25.34 SD) (control), 15.64 (19.87 SD) (IBD) Mean sweetened beverages including pure fruit juice g/day: 99.03 (123.04 SD) (control), 100.06 (100.64 SD) (IBD)	IBD: 75 incident cases/105,832 participants CD: 27/105,832 participants UC: 48/105,832 participants	Western dietary pattern: RR 1.18 [0.82–1.69] p = 0.57

TABLE 2.

Retrospective studies assessing sugar and/or sweetened beverages and IBD risk.

Author/Year	Country	# IBD (CD + UC)	# CD	# UC	# Controls	Time between diagnosis and data collection (years)	Method used to assess sugar exposure (FFQ = food frequency questionnaire)	How sugar intake was reported (dietary pattern vs. direct assessment of sugar, sweetened beverage, etc)	Amount of sugar consumption	Relative risk (95% CI)
Akbari 2022	Iran	81	0	81	163	≤ 1 year after diagnosis	FFQ	Dietary pattern	Foods grouped into dietary patterns. “Traditional diet” was most enriched for sugar and sweets.	Traditional dietary pattern: Crude OR: 4.22 (3.36–21.79) I: 4.63 (3.64–22.16) II: 4.97 (3.96–23.54) III: 4.67 (3.55–23.36) p < 0.05
Almofarreh 2022	Saudi Arabia	171	0	171	400	At initial diagnostic evaluation	FFQ	Frequency of sweetened beverage intake	UC: 63.2% frequent carbonated beverage intake, 36.8% infrequent Controls: 24.5% frequent, 75.5% infrequent	Unadjusted OR: 5.28 (3.59, 7.77) Age‐and sex‐adjusted OR: 8.34 (5.33, 13.06) Age‐sex‐BMI‐smoking‐adjusted OR: 9.82 (6.12, 15.76) p < 0.05
Amre 2007	Canada	130	130	0	202	≤ 1 month after diagnosis	FFQ	Not directly reported, assumed as carbohydrates without fiber	No data	Crude OR: 1.64 (0.88–3.06) Adjusted OR: 1.12 (0.39–3.28), p = 0.671
BianchiPorro 1985	Italy	233	109	124	250	No data	No data	No data	No data	UC: OR 2.37 CD: OR 2.38
Bikbavova 2021	Russia	81	0	81	39	No data	FFQ	Amount of added sugar in coffee and tea, frequency of sweet treats, soda	UC: 2 (1–3) tsp sugar in tea/coffee Controls: 1 (1–2) tsp. P = 0.041	No data
EGRC 1995	Japan	101	0	101	143	≤ 3 years after diagnosis	FFQ	Frequency of confectionery (chocolate, candy, snacks, cakes, Japanese cakes, ice cream) and sweetened beverage intake (canned coffee and cola and other effervescent drinks, and non‐carbonic soft drinks)	High consumption confectioneries: 34 IBD versus 36 controls High consumption soft drinks: 22 IBD versus 28 controls	High consumption confectioneries: OR 1.0, p = 0.94 High consumption soft drinks: OR 1.1, p = 0.95
Farsi 2022	Iran	112	26	86	122	No data but recruited “new cases” of IBD and collected dietary intake data on the past year	FFQ with direct interview	Frequency and calculated gram equivalents of sweets, sugars, honey and jam, and soft drinks	No data	IBD: Soft drinks: OR 0.99 (0.99–1.00), p = 0.307 Sweets: OR 0.99 (0.97–1.00), p = 0.062 Sugars: OR 0.99 (0.97–1.01), p = 0.166 Honey/Jam: OR 1.05 (1.00–1.10), p = 0.032 CD: Sweets: RR 0.95 Sugars: RR 0.98 Soney and jam: RR 1.03 Soft drinks: RR 0.99 UC: Sweets: RR 0.99 Sugars: RR 0.99 Honey and jam: RR 1.06 Soft drinks: RR 0.99
Halfvarson 2006	Sweden, Denmark	227	102	125	227	After diagnosis (average 17 years)	FFQ	Daily, weekly, or less frequent consumption of sugar on porridge, breakfast cereals, coffee or tea, grouped intake of tea, different types of bread, cereals, soft drinks juice, and fast food	No data	Sugar on porridge: UC OR 2.4 (1.2–4.9) p = 0.008; CD OR 1.2 (0.6–2.6) p = 0.58 Sugar on breakfast cereal not significant (NS) Sugar in coffee NS
Hansen 2011	Denmark	267	123	144	267	< 3 months	FFQ	Daily, weekly, or rarer consumption of fruit, cereal, sugar, and coffee. “High sugar intake” = ≥ 2 of the following: Sugar in coffee, sugar in tea, daily intake of soft drinks, sugar on breakfast cereals, sugar on porridge	“High sugar intake” = at least two of the following: Sugar in coffee, sugar in tea, daily intake of soft drinks, sugar on breakfast cereals, sugar on porridge	High sugar intake (CD): OR 3.50 (1.73–7.07) p < 0.05 High sugar intake (UC): OR 1.68 (0.96–2.97) p > 0.05
Jakobsen 2013	Denmark	118	59	56	477	Average < 1 year after diagnosis	FFQ (first year of life and year prior to diagnosis)	Frequency of soft drink and candy intake	No data	Univariate analysis: > Once weekly candy consumption OR 1.6 (1.1–2.6) p = 0.002; ≥ 4x/week soft drink consumption OR 4.8 (2.1–11.0) p = 0.0002 Multivariate analysis: > Once weekly candy consumption OR 1.5 (1.0–2.4) p = 0.07; ≥ 4x/week soft drink consumption OR 2.5 (1.0–6.2) p = 0.05 Univariate CD analysis: > Once weekly candy consumption OR 1.8 (1.0–3.2) p = 0.04; ≥ 4x/week soft drink consumption OR 6.6 (2.6–17.0) p < 0.0001 Multivariate CD analysis: ≥ 4x/week soft drink consumption OR 2.9 (1.0–8.5) p < 0.0001 Univariate UC analysis: ≥ 4x/week soft drink consumption OR 3.3 (1.0–10.1) p = 0.05 Multivariate UC sugar data not available
Kasper 1979	Germany	35	35	0	70	Average 1 year, maximum 2 years	Dietician interview	g/day estimated by dieticians based on response to questions	CD: 156 ± 14 g/day versus controls: 91 ± 5 g/day (p < 0.001) CD max intake: 385 g/day versus controls: 164 g/day In 11 of the 35 patients with CD (31%), the mean daily sugar consumption was more than 200 g. 90% of the sugar was eaten in the form of sucrose, 56% consumed as sweets, pastries, and as sugar in coffee and tea, and the remaining 34% by fruit drinks and lemonade	No data
Katschinski 1993	Germany	82	0	82	111	< 2 years (14%), 2–5 years (43%), 6–10 years (33%), > 10 years (10%)	FFQ	Sugar added to coffee and tea in g/day	None: 41 UC versus 58 controls Low (≤ 20 g): 26 UC versus 27 controls High (> 20 g): 15 UC versus 17 controls	RR 1.6 (0.87–2.9) adjusted for covariates
Kono 1994	Japan	101	0	101	143	≤ 3 years after diagnosis	FFQ with unanswered questions followed up by interview	Consumption frequency (0 = none or hardly, 1 = once or twice a week, 2 = 3–5x/week, 3 = almost daily) of sugar containing foods (confectioneries (chocolate, candy, snacks, cakes, Japanese cakes, and ice cream) and soft drinks (canned coffee, cola, and other carbonated and noncarbonated soft drinks), including amount of sugar added to coffee or tea. Frequency values were summed for individual items in each food group and glasses per day were summed for soft drinks. Subjects were classified into tertiles based on consumption scores	Soft drinks (IBD vs. controls): Low 48 versus 65, intermediate 30 versus 47, high 22 versus 28; Confectionaries: Low 34 versus 47, intermediated 32 versus 48, high 34 versus 46	Soda: UC: RR 1.1 (0.5–2.2), p = 0.95 Confectionery: UC: RR 1.0 (0.5–2.0), p = 0.94
Lautenschlager 2023	Sweden	1078	610	468	365	CD: Average 18 years UC: Average 16 years	FFQ	Frequency of consumption of sugary foods before age 18 (sugary drinks, eating artificial sugar, eating natural sugar) rated as more than other children, in average like other children, less than other children, or missing data.	More sugary drink intake than other children: 24 CD, 21 UC, 13 controls, p value of CD versus UC = 0.716, p value of IBD versus controls = 0.906 More artificial sugar than other children: 33 CD, 25 UC, 18 controls, p value of CD versus UC = 0.765, p value of IBD versus controls = 0.278 More natural sugar than other children: 54 CD, 45 UC, 36 controls, p value of CD versus UC = 0.720, p value of IBD versus controls = 0.427	No data
Martini 1976	Germany	63	63	0	63	Average 4.5 years after diagnosis	FFQ with estimation of grams per week; 15 subjects interviewed to verify consistency with FFQ	g/week of sweets, pastries, juice, lemonade	Sweets: CD: 742 g/week, controls: 285 g/week (p < 0.001) Pastries: CD: 1380 g/week, controls: 563 g/week (p < 0.001) Juices/Lemonade: CD: 2465 mL/week, controls: 2945 mL/week (p not significant, value not reported) Refined sugar intake: CD: 177 g/day, controls 74 g/day	No data
Matsui 1990	Japan	50	50	0	50	No data	FFQ	Sugar in g/day	CD 39.9 g/day sugar versus controls 27.5 g/day sugar, p < 0.05	No data
Mayberry 1981	United Kingdom	32	32	0	32	≤ 1 year after diagnosis	FFQ	Added sugar in g/day and g/week	Median daily addition of sugar to drinks (g): CD 48 (0–240) versus controls 24 (0–144), p < 0.03 Median weekly addition of sugar to drinks and cereals (g): CD: 336 (0–1752) versus controls 168 (0–1008), p < 0.007 Daily consumption of sugar‐containing foods (scored 0–6): CD 3 (1–6) versus controls: 1 (0–6), p < 0.007	No data
Mi 2022	China	50	10	40	50	≤ 1 year after diagnosis	FFQ	Consumption frequency categories for sweets and cakes and for carbonated drinks: Not at all or occasionally, 1–2 times per week, ≥ 3 times per week	Sweets and cakes: Not at all or occasionally: 14/50 IBD versus 20/50) controls 1–2x/week: 20/50 IBD versus 9/50 controls ≥ 3x/week: 16/50 IBD versus 11/50 controls, p = 0.366 Carbonated drinks: Not at all or occasionally: 27/50 IBD versus 24/50 controls 1–2x/week: 12/50 IBD versus 13/50 controls ≥ 3x/week: 11/50 IBD versus 13/50 controls, p = 0.862	No data
Miller 1976	Germany	34	34	0	34	No data	Questionnaire followed by a personal interview 14 days later to check and supplement the written information	Sugar in g/day	Average 55 g of sugar/day in controls versus 115 g of sugar/day in CD Before disease onset: 59% CD versus 9% controls consumed > 100 g sugar/day; p < 0.01	No data
Ng 2015	9 regions in Asia‐Pacific (China, Hong Kong, Indonesia, Sri Lanka, Macau, Malaysia, Singapore, Thailand and Australia)	442	186	256	940	0 years (data collected at diagnosis)	FFQ	Juice more than twice a week Soda more than twice a week	No data	Juice ≥ 2x/week: CD adjusted (Asia and Australia): OR 0.443 (0.249–0.788) p = 0.006 UC adjusted (Asia and Australia): OR 1.063 (0.573–1.947) p = 0.845 Soft drinks ≥ 2x/week: CD adjusted (Asia and Australia): OR 0.759 (0.386–1.491) p = 0.423 UC adjusted (Asia and Australia): OR 1.553 (0.831–2.899) p = 0.167
Niewiadomski 2016	Australia	132	81	51	104	Most patients completed the questionnaire within 6 months of diagnosis, except for those diagnosed in 2007/2008 (n = 21) who did so later	FFQ	High sugar diet defined as ≥ 2 of the following: Sugar in coffee, sugar in tea, daily intake of soft drinks, sugar on breakfast cereal/porridge	24/81 CD versus 17/73 controls; p = 0.350 13/51 UC versus 17/73 controls; p = 0.949	No data
Octoratou 2012	Greece	58	58	0	38	2–4 weeks for 28 participants, 2–11 years for 30 participants	No data	The consumption of each food item was measured by the use of an interval scale from 1 to 9, with category “1” corresponding to consumption “few times a year or at al,” and category “9” corresponding to consumption “7+ times are day”	Consumption scale for sugars and sweeteners: 6.00 new pts versus 5.47 old pts versus 4.16 healthy controls (p = 0.002)	OR 1.561 (1.023–2.382), p = 0.039
Persson 1992	Sweden	297	152	145	305	< 4 years after diagnosis	FFQ	Intake of carbohydrates, monosaccharides, disaccharides, and sucrose Frequency of consuming soft drinks	Monosaccharides: CD 62/152 (highest intake) versus 45/152 (lowest intake), UC 41/145 (highest intake) versus 52/145 (lowest intake), UC controls 103/305 (highest intake) versus 102/305 (lowest intake) Disaccharides: CD 57/152 (highest intake) versus 56/152 (lowest intake), UC 61/145 (highest intake) versus 46/145 (lowest intake), UC controls 88/305 (highest intake) versus 116/305 (lowest intake) Sucrose: CD 59/152 (highest intake) versus 47/152 (lowest intake), UC 39/145 (highest intake) versus 50/145 (lowest intake), UC controls 72/305 (highest intake) versus 138/305 (lowest intake)	Daily soft drinks: 2.8 (1.6–4.9) High sucrose after adjustment: 2.5 (1.3–4.8)
Preda 2020	Romania and Belgium	129	89	40	56	No data	FFQ	Categories of daily sugar‐sweetened beverage intake: < 1 L/day versus > 1 L/day Categories of sweets consumption: Rare, occasional, or frequent (daily)	No data	Daily sweets/sweetened drinks > 1 L/day: OR 3.36 (1.6–7), p < 0.001
Preda 2023	Romania	129	89	40	64	≥ 6 months, average 6 years CD versus 8 years UC	Interview and supervised FFQ	Categories of daily sugar‐sweetened beverage intake: < 1 L/day versus > 1 L/day Categories of ice cream consumption: Rare, occasional, or frequent (daily)	> 1 L/day sweetened drinks: 34/89 CD vs.11/64 controls, p = 0.022 19/40 UC vs.11/64 controls, p = 0.005 Daily ice cream: 15/89 CD versus 4/64 controls, p = 0.175 8/40 UC versus 4/64 controls, p = 0.144	No data
Reif 1997	Israel	87	33	54	144	All ≤ 1 year from onset of symptoms, most ≤ 3–4 months from onset of symptoms	Trained dietician interview with quantitative FFQ	Daily consumption categories of sugars: Low: < 107 g/day, medium: 107–158 g/day, or high: > 158 g/day Daily consumption categories of soft drinks: Low: < 133 mL/day, medium: 133–400 mL/day, or high: > 400 mL/day	No data	Medium total sugar intake: OR 1.8 (IBD vs. clinic controls, p = 0.04), OR 1.13 (IBD vs. population controls, p = 0.67), OR 1.99 (UC vs. controls, p = 0.05), OR 1.17 (CD vs. controls, p = 0.63) High total sugar intake: OR 3.2 (IBD vs. clinic controls, p = 0.04), OR 1.27 (IBD vs. population controls, p = 0.67), OR 3.98 (UC vs. controls, p = 0.05), OR 1.37 (CD vs. controls, p = 0.63) Medium soft drink intake: OR 1.99 (IBD vs. clinic controls, p = 0.01), OR 1.28 (IBD vs. population controls, p = 0.31), OR 1.84 (UC vs. controls, p = 0.02), 1.2 (CD vs. controls, p = 0.58) High soft drink intake: OR 3.99 (IBD vs. clinic controls, p = 0.01), OR 1.65 (IBD vs. population controls, p = 0.31), OR 3.39 (UC vs. controls, p = 0.02), OR 1.44 (CD vs. controls, p = 0.58)
Russel 1998	Netherlands	688	290	398	616	< 5 years after symptom onset	FFQ	Daily number of cups of coffee (with or without sugar) and tea (with or without sugar); categories of weekly consumption of cola drinks, orange juice, citrus fruits, chocolate, confectionary and chewing gum: none or one, between two and five times, and more than five times weekly	Soda: 128/290 CD versus 186/616 controls, 129/398 UC versus 186/616 controls Chocolate: 115/290 CD versus 198/616 controls, 157/398 UC versus 186/616 controls	Before first complaints: Cola drink intake (CD): OR 2.2 (1.5–3.1) Tea with sugar: OR 1.3 (0.9–1.8) Cola drink intake (UC): OR 1.6 (1.1–2.3) At time of questionnaire: Cola drink intake (CD): OR 1.4 (1.0–2.0) Chewing gum intake (CD): OR 1.3 (0.9–1.9) Cola drink intake (UC): OR 1.2 (0.8–1.6)
Sakamoto 2005	Japan	234	126	108	211	≤ 5 years	FFQ with interview follow‐up for unanswered questions	Frequency of “sugars and sweeteners” and “confectionaries” separated into quartiles Average daily intake of carbohydrates separated into quartiles	Highest quartile sugars and sweeteners: 47/53 CD versus 23/52 controls, 28/53 UC versus 33/52 controls Highest quartile confectioneries: 47/53 CD versus 20/52 controls, 43/53 UC vs.12/52 controls	Sugars and sweeteners: CD: OR 2.12 (1.08–4.17) UC: OR 0.71 (0.35–1.45) Confectioneries: CD: OR 2.83 (1.38–5.83) UC: OR 2.86 (1.24–6.57)
Thornton 1979	England	30	30	0	30	≤ 3 months	Dietician interview using dietary history method	Refined sugar in g/day (defined as any fiber‐depleted sugar, including brown sugar)	Total refined sugar median intake: 122 g/day IBD versus 65 g/day controls; p < 0.002 Refined sugar median intake in foods: 49 g/day IBD versus 25 g/day controls; p < 0.02 Refined sugar median intake in drinks: 69 g/day IBD versus 30 g/day controls; p < 0.01	No data
Thornton 1980	England	30	0	30	30	≤ 3 months	Dietician interview using dietary history method	Refined sugar in g/day	Refined sugar mean intake: 96 ± 9 g/day IBD versus 97 ± 9 g/day controls	No data
Tragnone 1995	Italy	104	51	53	208	≤ 6 months	Dietician interview using recall questionnaire	Refined sugar in g/day	Total IBD: 100.2 ± 42.9 versus Total controls 81.4 ± 24.2 CD: 98 ± 36 versus CD controls: 78.9 ± 27.1 UC: 101.43 ± 37.08 versus UC controls: 83.5 ± 30.6	> 100 g refined sugar intake per day: CD: OR 3.5 (1.5–8.1) p < 0.01 UC: OR 3.7 (1.6–8.6) p < 0.01
Trakman 2022	Australia, Hong Kong, and mainland China	274	274	0	257	For retrospective data: average 8 years, range 0–41 years. Also included 3‐day prospective dietary intake	FFQ matched against prospective 3‐day intake and verified by a dietician	Total artificial sweetener intake in mg/day	Higher in cases versus controls	No data
Vind 2008	Denmark	347	156	191	755	No data	FFQ	Sugar consumption, defined as daily use of either two or more teaspoons in each cup of tea or coffee, additional sugar on cereals, or intake of several deciliters of soft drinks daily	No data	Sugar consumption was significantly higher in CD patients than in UC patients (p = 0.0001)
Wang 2013	China	1308	0	1308	1308	No data	FFQ	Consumption frequency categories: None or rare, light (1–2 times a week), and heavy (3–6 times a week or every day for a period of at least 2 months)	Light sugar: 64.8% UC versus 53.1% controls, p < 0.0001 Heavy sugar: 23% UC versus 15.2% controls, p < 0.0001	OR 3.162 (2.480–4.032) for light sugar consumption; OR 3.390 (2.921–5.288) for heavy sugar consumption, p < 0.001
Yuan 2015	China	103	0	103	103	No data; questionnaire asked about “daily diet before the onset of disease”	FFQ with follow‐up by two trained investigators	Consumption frequency categories: Never, occasionally, often	Often sweets: 26/103 IBD versus 17/103 controls, p = 0.041	No data

3.1. Dietary Sugar/Sweeteners/Sweets

3.1.1. Sugars

Twenty retrospective [21, 22, 25, 26, 27, 29, 32, 34, 35, 37, 40, 41, 44, 46, 47, 48, 49, 51, 52, 53] and three prospective studies [12, 14, 16] assessed the specific impact of sugar intake on IBD development. None of the prospective studies found a significant association between sugar intake and the development of IBD.

Seven retrospective studies reported a significant association between sugar and IBD development [22, 27, 29, 34, 35, 51, 52]. Study design was most often lacking in comparability of those with IBD to controls and ascertainment of exposure to sugar. In contrast, four retrospective studies found no association between sugar intake and the development of IBD [21, 25, 32, 48]. In the Swiss IBD cohort, Lautenschlager et al. [32] found no significant effect of childhood sugar intake on the risk of developing CD or UC (p = 0.427). In Iran, Farsi et al. found that patients with IBD did not have a more frequent intake of sugars.

In China, Wang et al. [52] found that those with UC had a 3‐fold increased likelihood of consuming both light and heavy sugar. Sakamoto et al. [46] found that those with CD were more likely to have a higher frequency of sugars and sweeteners in their diets, whereas those with UC were not.

In the Middle East, Reif et al. [44] found a significantly higher likelihood of high total sugar intake in all IBD patients compared with clinic controls. High sucrose intake was associated with IBD when compared with clinic controls, population controls, and UC patients against their individual controls.

In Europe, Octoratou et al. [40] found significantly higher sweetener and sugar consumption in patients with CD, while Halfvarson et al. [26] found that patients with UC more frequently added sugar to porridge compared to controls. Tragnone et al. [49] found a higher refined sugar intake in people with IBD and an over 3‐fold relative risk for consuming over 100 g of sugar/day in people with IBD. Persson et al. [41]. found no significant difference in monosaccharide intake in patients with UC and CD. After adjustment for fiber intake, they found an association between CD and high sucrose intake. Thornton et al. [47] found about 60 g higher daily refined sugar intake in those with IBD (p < 0.002). However, another study by the authors used a similar methodology in patients with UC and found no difference in refined sugar intake between those with UC and controls [48]. Miller et al. [37] similarly found 60 g higher sugar intake in those with CD compared to controls, and a great proportion of those with CD consumed over 100 g/day of sugar prior to disease onset (p < 0.01).

3.1.2. Sweets

Two prospective [8, 18] and 11 retrospective [25, 28, 31, 36, 40, 42, 43, 45, 46] studies reported on sweets intake rather than specific quantity of sugar intake.

Narula et al. [8] assessed daily refined sweetened food intake in 116,037 participants across multiple continents and noted that high refined sweetened food intake was associated with the risk of IBD. Vasseur et al. [18] found a trend for higher intake of cakes, cookies, and pastries (p = 0.05) in incident IBD cases.

Of the retrospective studies, seven (64%) reported a statistically significant association between sweet intake and IBD incidence. However, Preda et al. [42] reported significance only when grouping sweets and sweetened beverage intake, and later found no significant association with sweets in their 2023 analysis [43], suggesting that significance in the 2020 analysis was driven by sweetened beverage intake. Overall, 5/11 (45%) of studies reported a significant association between sweets intake and IBD incidence. Those with significant findings more often reported a positive association with refined sweet intake (candy, chocolate, pastries, confectionaries, etc.) and a negative association with honey intake, as detailed below.

Of retrospective studies that concluded the presence of an association, Octoratou et al. [40] determined that there was significantly higher sweetener and sugar consumption in patients with CD but significantly lower honey intake, thus implicating refined sugar, rather than sweets in general, as being associated with IBD. Russell et al. [45] found significantly higher chocolate intake in the pre‐illness diet of patients with both types of IBD. Martini and Brandes [33] found that patients with CD consumed 2.55 times more sweets and pastries than controls.

Mi et al. [36] most recently found no significant association between IBD and frequency of sweets and cake intake in 10 participants with CD and 40 participants with UC in China (p = 0.366). However, their study was limited by the small sample size. Kono et al. and the Epidemiology Group of the Research Committee of Inflammatory Bowel Disease in Japan found no significant difference in high sugar consumption between those with UC and controls [24, 31].

3.1.3. Artificial Sweeteners

Of articles measuring artificial sweetener intake, only Russel et al. [45] specified solid food intake in the form of chewing gum. These authors note significantly more frequent chewing gum intake on retrospective analysis of 290 patients with CD from the Netherlands (OR 1.5 [1.1–2.2]) [45]. Other studies that did not specify the form in which artificial sugars were ingested were included in the sweetened beverages section given the likely predominance of artificial sugar intake in “diet” and “zero‐calorie” beverages in the modern diet.

3.1.4. Meta‐Analysis

Of the studies assessing sugar and/or sweet intake included in the meta‐analysis, 2 prospectively [8, 12] and 12 retrospectively [25, 26, 27, 28, 32, 39, 40, 41, 43, 44, 45, 46] assessed sugar intake in patients with CD (Figure 2) and 3 prospectively [8, 12, 14] and 15 retrospectively [24, 25, 26, 27, 30, 31, 32, 39, 41, 43, 44, 45, 46, 52, 53] assessed that in patients with UC (Figure 3). Studies containing grouped IBD analysis are included in Supporting Information S1: Figure S1; the direction of trends were similar to stratified analyses.

FIGURE 2.

Meta‐analysis of sweets intake and relative odds of Crohn's disease. Abbreviations: CI, confidence interval; IV, inverse variance; SE, standard error.

FIGURE 3.

Meta‐analysis of sweets intake and relative odds of ulcerative colitis. Abbreviations: CI, confidence interval; IV, inverse variance; SE, standard error.

High sweets intake was associated with an overall increased likelihood of CD (OR 1.66 [1.21–2.29], p = 0.002), although the results remained significant for retrospective studies (OR 1.77 [1.25–2.50], p = 0.001) rather than prospective studies (OR 0.91 [0.42–1.96], p = 0.81). There was significant heterogeneity among retrospective studies (p < 0.00001, I² = 86%) but not among prospective studies (p = 0.59, I² = 0%) assessing patients with CD. Both prospective studies were considered moderate‐to‐high quality evidence, while 10 retrospective studies were considered moderate‐to‐high quality and 2 were considered fair quality evidence (Supporting Information S1: Table S1).

Similarly, high sweets intake was associated with an overall increased likelihood of UC (OR 1.59 [1.25–2.02], p = 0.0002) only in retrospective studies (OR 1.61 [1.24–2.09], p = 0.0004) rather than prospective studies (OR 1.50 [0.73–3.08], p = 0.28). There was significant heterogeneity among retrospective studies (p < 0.00001, I² = 80%) and prospective studies (p = 0.04, I² = 70%). All three prospective studies were considered moderate‐to‐high quality evidence, while 10 retrospective studies were considered moderate‐to‐high quality and 5 were considered fair quality evidence (Supporting Information S1: Table S1).

3.2. Sweetened Beverages

3.2.1. Soda/Soft Drinks/Carbonated Beverages

Five prospective [8, 9, 13, 15, 17] and 11 retrospective studies [20, 24, 25, 27, 28, 31, 36, 41, 44, 45, 51] reported on sugar intake in the form of sodas, soft drinks, or carbonated beverages.

Three [8, 9, 13] out of the five [8, 9, 13, 15, 17] prospective studies reported that high SB intake was associated with a significantly increased likelihood of patients developing IBD. El Mouzan et al. [13] assessed the frequency of soft drink intake at least 3 months before the onset of symptoms. Participants with IBD had a higher frequency of more than once weekly soft drink intake than controls. Fu et al. [9] determined that participants who had more than one SB per day were more likely to develop IBD and CD but not UC an average of 10.2 years later. Narula et al. [8] also measured soft drink intake and found a significantly higher risk of both UC and CD with high soft drink intake. However, low CD incidence could underpower the findings of this study.

Two prospective studies found no link between soft drink intake and the risk of developing IBD. Khalili et al. [15] assessed 83,042 Swedish participants' usual consumption of soft drinks and found 11 incident CD cases (p = 0.49) and 28 incident UC cases (0.97) and no link between the consumption of soft drinks and development of CD or UC. Similarly, Racine et al. [17] found no significant increase in IBD incidence among 117 incident CD cases and 256 UC cases.

Of the retrospective studies, seven studies [20, 27, 28, 41, 44, 45, 51] concluded a link between soft drink consumption and the risk of developing IBD. Notably, Hansen et al. [27] assessed sugar intake in multiple categories (sugar in coffee, sugar in tea, soft drinks, sugar on breakfast cereals and porridge) and found that patients with CD had higher sugar intake than controls but patients with UC did not. Jakobsen et al. [28] specifically assessed the frequency of soft drink intake and found that on average, both patients with CD and UC had a higher frequency of soft drink consumption. Vind et al. [51] concluded that sugar consumption was significantly higher in patients with CD than in patients with UC (p = 0.0001), supporting the findings of Russel et al. Since the amount of sugar and other summary statistics on the different forms of sugar were not reported, Vind et al. was not included in the meta‐analysis.

Four studies [24, 25, 31, 36] concluded no significant association between soft drink consumption and IBD, although each study was limited by small and poorly generalizable samples assessed using low quality methodology. Halfvarson et al. [26] and Ng et al. [38] both found no difference in frequency of soft drink intake for either UC or CD, but found a significant difference for other categories of added sugar. These inconsistent findings across added sugar and the risk of IBD suggest that recall bias plays a larger role in these retrospective studies, or that unreliable survey methods were utilized.

3.2.2. Sugar Added to Tea or Coffee

Three studies [23, 27, 51] found that sugar added to tea or coffee is linked to IBD, but each paper utilized different means of measuring sugar intake and studied different patient populations. Bikbavova et al. [23] found that UC patients put a statistically greater (p = 0.041) number of teaspoons of sugar (2.0 (1–3 tsp)) in their tea/coffee than controls (1.0 (1–2 tsp)). Hansen et al. [27] and Vind et al. [51] reported data on high sugar intake as a general category rather than analyzing the impact of specific forms of sugar consumption. The remaining four articles found no significant impact of sugar in tea or coffee on the risk of developing IBD [26, 30, 39, 45].

3.2.3. Juice/Artificially Sweetened Beverages

Two prospective studies [8, 18] and six retrospective [26, 32, 33, 38, 45, 50] studies analyzed juice and artificially sweetened beverages and their relationship with the risk of IBD.

Fu et al. and Vasseur et al. both evaluated artificially sweetened beverages and natural/pure fruit juices, and neither found a significant link between these beverages and incident rates of CD or UC [8, 18].

In retrospective studies, Trakman et al. [50] studied only artificial sweeteners and found that patients with CD consumed significantly more artificial sweeteners (average ∼1 mg/day more) versus controls (p < 0.001). Ng et al. [38] studied juice intake and found that more than once weekly juice intake was significantly associated with increased risk of CD in unadjusted analysis but was associated with decreased risk in analysis when adjusted for sex, age, and country income. They concluded that there was no significant difference in unadjusted or adjusted analysis in participants with UC.

The remaining retrospective articles focusing on juice or artificial sweeteners and IBD found no significant difference between IBD and control patients [26, 33, 45]. In the most recent published paper, Lautenschlager et al. [32] found no significant connection between sugar intake before the age of 18 and risk of developing CD or UC.

3.2.4. Meta‐Analysis

While some studies categorized juices and artificially sweetened beverages as SSBs, most studies included these as a distinct category. As such, the term “sweetened beverages” (SBs) is used here to encompass all sub‐categories (soda, soft drinks, carbonated beverages, sugar added to tea or coffee, juice, and artificial sweeteners) that will be discussed. Of the studies evaluating SB intake included in our meta‐analysis, 4 prospective studies and 8 retrospective studies assessed SB intake in patients with CD (Figure 4). The same 4 prospective studies, along with 9 retrospective studies, assessed SB intake in patients with UC (Figure 5), and studies that grouped IBD are included in Supporting Information S1: Figure S2.

FIGURE 4.

Meta‐analysis of sweetened beverages intake and relative odds of Crohn's disease. Abbreviations: CI, confidence interval; IV, inverse variance; SE, standard error.

FIGURE 5.

Meta‐analysis of sweetened beverages intake and relative odds of ulcerative colitis. Abbreviations: CI, confidence interval; IV, inverse variance; SE, standard error.

In the meta‐analysis, high SB intake was positively associated with the likelihood of developing CD. All 4 prospective studies received a moderately high rating on the Newcastle‐Ottawa Scale and 6 retrospective studies received a moderate‐to‐high rating, while 2 received a fair rating (Supporting Information S1: Table S1). High SB intake was also positively associated with the likelihood of developing UC (OR 1.72 [1.23–2.39], p = 0.001). This result was consistent in both prospective studies (OR 1.37 [1.11–1.69], p = 0.003) and retrospective studies (OR 1.96 [1.13–3.40], p = 0.02). Heterogeneity was low among the prospective studies for UC (I² = 24%, p = 0.27) and high among retrospective studies (I² = 93%, p < 0.00001). The prospective studies used to assess soft drink intake and UC were the same as those used for CD, and all 4 received a moderate‐to‐high rating (Supporting Information S1: Table S1).

4. Discussion

Dietary sugar intake has become increasingly implicated in inflammatory conditions, making it a topic of great interest for patients, clinicians, and scientists. This systematic review accumulates data from 45 studies—including recently published population‐based cohort studies [8, 9]—to provide an updated assessment on how dietary sugars and sweetened beverages impact the development of IBD. Our systematic review presents evidence that both dietary sugar and SB intake are associated with increased risk of developing CD and UC.

Prior to this study, a systematic review from 2021 by Khademi et al. [7] concluded that higher dietary sugar intake was linked with increased risk of IBD, but that SBs and carbohydrate consumption had no effect. However, these findings did not include two large prospective studies and several retrospective studies that had not yet been published. The inclusion of Narula and colleagues' multicenter international prospective trial [8] and Fu and colleagues' analysis of the UK biobank [9] adds two large cohorts assessed with rigorous methodology for sugar quantification. Several other studies were identified in this review that were not included in the prior analysis, and the larger sample size of this review addresses a concern raised by Khademi and colleagues and may explain the difference in findings. Additionally, this review separated sugar intake from total carbohydrate intake, eliminating the confounding impact of potentially beneficial dietary fiber and other complex carbohydrates that were included in the previous review.

In this review, sweets intake was associated with a significantly increased risk of developing CD and UC. The relatively high quality of evidence—despite recall bias—and large sample sizes attest to the reliability of these results, and previous reviews similarly support a positive link between sweets intake and inflammation [54, 55, 56]. However, statistical significance was only found in retrospective studies. Lack of significant association in prospective studies suggests that the overall significance of the trend may have been driven by recall bias and suggests the need for more research into the link between sugar intake and IBD. Given the increasing incidence of IBD in non‐European (particularly East and South Asian) countries [57] and concurrent increases in sugar intake as these countries adopt more Westernized diets, large‐scale prospective trials in these countries may help elucidate this relationship.

Soda/sweetened beverage intake was also positively correlated with the risk of developing CD and UC. Both prospective and retrospective studies exhibited statistical significance in their pooled findings, which strengthens the conclusion. The available research presents a fairly high quality of evidence with large sample sizes and rigorous methodology. However, sugar intake added to tea or coffee was not included in the meta‐analysis because although all studies were rated moderately high on the Newcastle‐Ottawa Scale, they presented with high levels of heterogeneity in their methods for collecting sugar intake data and participant populations. More research on this category of sugar intake, with standardized data collection methods, would be beneficial to fully ascertain the connection between sweetened beverage intake and IBD.

Mechanistically, experimental models show that dietary simple sugars can shift gut microbiome populations toward those that exacerbate colitis [58], deplete short chain fatty acids essential for maintaining gut barrier integrity [59], and promote pro‐inflammatory cytokine production and macrophage and T cell activation [60]. These effects have been shown to be dose‐dependent in experimental models [61]. Moreover, high sugar intake in humans with IBD is associated with higher healthcare utilization [62] and all‐cause mortality [63], supporting its pathophysiologic role in promoting intestinal inflammation.

Although healthy gut microbiota profiles vary greatly between individuals, imbalance of the gut microbial system—a potential result of sweetened beverages—has been implicated in increased susceptibility to intestinal inflammation in mice. The repercussions of changes in gut microbiota can be extreme, potentially contributing to the development of IBD [64]. However, certain prebiotics, such as glucomannan and beta‐glucans, show promise in alleviating colitis symptoms in rodents and should be further investigated in humans [65].

In contrast, many studies have referenced the potentially protective role of fiber in ameliorating the inflammatory risk of sweet intake. Studies that measured fiber intake often only found significant associations after removing its effect from analysis [18, 21, 28, 41]. While complex carbohydrate foods naturally containing fiber, such as whole grains, legumes, and fruits, are known to have a multitude of health benefits and are considered protective for patients with both IBD subtypes [66], some authors postulate that fiber intake may even prevent the deleterious effects of refined or artificial sugar intake [21, 46]. Given inconclusive guidance on the role of fiber, we attempted to extricate our results from this confounder by only including sugar, rather than total carbohydrate or carbohydrate with fiber intake. However, fiber's protective potential raises intriguing possibilities and warrants further study.

Different types of sweeteners may have varied health implications. For example, natural sweeteners such as honey, maple syrup, and agave possess anti‐inflammatory, antiviral, and antineoplastic properties [67, 68, 69]. On the other hand, artificial sweetener consumption has been linked to changes in the gut microbiome and intestinal epithelial immune reactivity [70]. Free fructose, present in large amounts in popular soft drinks, increases inflammation by promoting macrophage infiltration into adipocytes and pro‐inflammatory cytokines and results in further inflammation [71, 72]. Both the positive and negative effects of different sweeteners may impact IBD risk and outcomes, and future research differentiating between sweetener subtypes and their relationship with IBD risk will better guide patients and public health policy.

This systematic review presents several strengths, including a stringent procedure for performing the screening, evidence grading, and data extraction of each publication. Our study encompasses a broader range of papers examining sugar and IBD than previous reviews, allowing for a more comprehensive understanding of the risks associated with dietary sugar intake. However, this review also has several limitations. Firstly, our meta‐analyses revealed differences in pooled findings between retrospective and prospective studies. These differences are further highlighted by substantial statistical heterogeneity (I²) among retrospective studies but not prospective studies. Some possible explanations could include drawbacks of temporally distant food recall, differing timeframes for food recall, varying methods to obtain dietary data, and different populations. Additionally, the majority of retrospective studies assessed frequency of sugar or SB intake, rather than quantity as measured by the majority of prospective studies, and did not standardize thresholds for low versus high sugar or SB intake, potentially confounding comparisons. The lack of quantification in sugar intake makes it challenging to provide a quantitative recommendation on the maximum amount of sugar that is considered safe. Moreover, individuals who drink SBs are also more likely to have a less healthy diet overall, consume more processed foods and diets higher in saturated fats, and struggle with food insecurity [73]. These confounders of diet could be linked with IBD development and be responsible for the observed relationship seen between SB intake and IBD. Secondly, although the cited studies would be considered “high quality” evidence based on NOS criteria, these criteria do not consider sample size or inconsistencies between studies. There were also global challenges in these dietary studies (e.g., recall bias, unmeasured confounders, methodologic heterogeneity) that would warrant downgrading the overall quality of evidence as moderate or low. Thirdly, although our review contained studies conducted in a wide range of countries and regions (51% European, 23% Asian, 10% Australian, 7% Middle Eastern, 6% North American, and 4% South American), Central America was not represented. These limitations highlight the need for future research specifically controlling for these factors and representing all patients with IBD.

Considering the prevalence of dietary sugars and the popularity of SBs across the world, it is vital that healthcare professionals and patients alike understand the risks associated with regular sugar consumption. Given the substantial clinical and methodological heterogeneity across studies, we were unable to define an acceptable threshold of sugar intake to minimize the risk of IBD using the data from our meta‐analysis. However, the findings support ongoing recommendations to moderate or reduce sugar consumption in general, particularly among populations at high risk for developing IBD (e.g., relatives of patients with IBD). Future research should focus on how reducing sugar intake affects disease outcomes and symptoms in IBD patients. As we work to better understand the impact of sugar—and other common environmental factors—on IBD, we can improve our treatment and management of this emerging chronic illness.

Author Contributions

Study design: Isha Maniyar, Arpita Jajoo, Gala Godoy Brewer, Preetha Iyengar, Andrew Nguyen, Alyssa Parian and Berkeley N. Limketkai. Data acquisition: Isha Maniyar, Arpita Jajoo, Gala Godoy Brewer, Preetha Iyengar, Andrew Nguyen, Alyssa Parian and Berkeley N. Limketkai. Data interpretation: Isha Maniyar, Arpita Jajoo, Alyssa Parian and Berkeley N. Limketkai. Article drafting: Isha Maniyar, Arpita Jajoo and Preetha Iyengar. Article revision for important intellectual content: Isha Maniyar, Arpita Jajoo, Gala Godoy Brewer, Preetha Iyengar, Andrew Nguyen, Alyssa Parian and Berkeley N. Limketkai. Final approval by Isha Maniyar, Arpita Jajoo, Gala Godoy Brewer, Preetha Iyengar, Andrew Nguyen, Alyssa Parian and Berkeley N. Limketkai. Data analysis: Berkeley N. Limketkai.

Disclosure

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information S1

Data Availability Statement

The data‐underlying this article are publicly available via their respective journals.