Skip to main content
Download PDF
Download ePub

Estimating effects of whole grain consumption on type 2 diabetes, colorectal cancer and cardiovascular disease: a burden of proof study

Nutrition Journal volume 23, Article number: 49 (2024) Cite this article

Abstract

Background

Previous studies on whole grain consumption had inconsistent findings and lacked quantitative assessments of evidence quality. Therefore, we aimed to summarize updated findings using the Burden of Proof analysis (BPRF) to investigate the relationship of whole grain consumption on type 2 diabetes (T2D), colorectal cancer (CRC), stroke, and ischemic heart disease (IHD).

Methods

We conducted a literature search in the Medline and Web of Science up to June 12, 2023, to identify related cohort studies and systematic reviews. The mean RR (relative risk) curve and uncertainty intervals (UIs), BPRF function, risk-outcome score (ROS), and the theoretical minimum risk exposure level (TMREL) were estimated to evaluate the level of four risk-outcome pairs.

Results

In total, 27 prospective cohorts were included in our analysis. Consuming whole grain at the range of TMREL (118.5–148.1 g per day) was associated with lower risks: T2D (declined by 37.3%, 95% UI: 5.8 to 59.5), CRC (declined by 17.3%, 6.5 to 27.7), stroke (declined by 21.8%, 7.3 to 35.1), and IHD (declined by 36.9%, 7.1 to 58.0). For all outcomes except stroke, we observed a non-linear, monotonic decrease as whole grain consumption increased; For stroke, it followed a J-shaped curve (the greatest decline in the risk of stroke at consuming 100 g whole grain for a day). The relationships between whole grain consumption and four diseases are all two-star pairs (ROS: 0.087, 0.068, 0.062, 0.095 for T2D, CRC, stroke, and IHD, respectively).

Conclusion

Consuming 100 g of whole grains per day offers broad protective benefits. However, exceeding this threshold may diminish the protective effects against stroke. Our findings endorse replacing refined grains with whole grains as the main source of daily carbohydrates.

Registry and registry number for systematic reviews or meta-analyses

We have registered our research in PROSPERO, and the identifier of our meta-analyses is CRD42023447345.

Peer Review reports

Introduction

Whole grains have been widely endorsed as a superior substitute for primary energy and carbohydrate sources in daily dietary guidelines because of their high dietary fiber content and numerous bioactive compounds [1]. The Global Burden of Disease Study 2019 (GBD 2019) has reported that lower intake of whole grain accounted for 1,844,836 (95% uncertainty interval [UI]: 2,338,609–921,291) deaths and 42.5 million (53.2–17.5) disability-adjusted life years (DALYs) [2]. The large estimated burden demonstrated the importance of fully appreciating the relationship between whole grain consumption and potentially related health outcomes and of further improving the strength of evidence supporting the understanding of those relationships.

Increasing evidence has found that a high intake of whole grains is related to a reduction in the risk of type 2 diabetes (T2D), colorectal cancer (CRC), ischemic heart disease (IHD), and stroke [2, 3]. However, regarding CRC, T2D and IHD, previous studies, including dose-response meta-analysis or cohort studies, exhibit variations in their consumption ranges. This complicates the comparability and consolidation of evidence [3,4,5,6]. Besides, in relation to stroke, recent meta-analyses have presented inconsistent findings [7, 8]. Although there is an increasing body of evidence supporting the positive impact of consuming whole grains on health, the challenge lies in accurately estimating RR associated with varying levels of consumption. This limitation hinders the ability of decision-makers to fully comprehend the strength of the connection between consuming whole grains and various health outcomes.

Burden of proof risk function (BPRF) is a new meta-analysis method that can quantitatively estimate the level of risk closest to the null hypothesis [9]. Hitherto, most of meta-regression studies applied given fixed knots to fit the spline models or forced a log-linear assumption to simplify statistical analysis. However, such a method may limit their ability to capture the effects of whole grain consumption on health outcomes, as the relationship between increasing whole grain intake and its impact on health might not be straightforward: it could lead to slight decreases in positive effects, or it could even become harmful if the consumption of whole grains goes beyond a certain point [3, 10]. Unlike existing methods, BPRF relaxed the conventional assumption of a log-linear shape in risk functions, and instead applied a data-driven approach to determine the relationship of risk-outcome pairs using a quadratic spline. Thus, BPRF can help to identify the ‘true’ shape of the risk function [11]. In addition, existing methods, such as Grading of Recommendations, Assessment, Development and Evaluations (GRADE) or NutriGRADE, are commonly applied to assess the quality of the underlying evidence [12]. However, such methods are unable to extend to quantify variation in true effect size caused by bias from covariates or other limitations of the evidence [11]. Nevertheless, BPRF can synthesize available evidence in algorithm to calculate uncertainty inclusive of between-study heterogeneity.

To precisely quantify the health effects of whole grain consumption, a meta-regression analysis was conducted on the evidence from prospective cohort studies. This study focused specifically on four health outcomes (T2D, CRC, IHD, and stroke) linked to whole grain consumption, as reported by the GBD study [13].

Methods

Our protocol has been registered in International Prospective Register of Systematic Reviews (PROSPERO, identifier: CRD42023447345). We followed a standard framework of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline to report our results [14].

Search strategy and selection criteria

We searched data published in English in the MEDLINE and Web of Science for systematic review, and cohort studies from 1 January, 2000 to 12 June, 2023, using standard search strings (Supplementary Table 1). A reference list of included publications was also manually screened to identify additional cohort studies. Titles and abstracts were screened by two reviewers (H Liu and J Wang), with discrepancies being reconciled through consulting a third author (Y Li).

Only prospective observational studies (for both incidence and mortality) published in English were included. Studies should report a relative risk ratio (RR), odds ratio (OR) or hazard ratio (HR) of the associations between whole grain consumption and at least one of the four outcomes. Additionally, they should specify the amount of whole grain consumption in both the reference group and the alternate group for comparison.

Retrospective studies, conference abstracts, ecological studies, case reports, case-series, letters to the editor, conference proceedings, umbrella reviews, systematic reviews or meta-analyses as well as studies conducted in animals, children, or adolescents were excluded. Besides, we excluded studies that failed to report whole grain consumption without grams or servings equivalent, such as studies that used aggregated “diet scores” as a measure of consumption, and those that only reported specific subtypes of grains were also excluded. And studies reporting outcomes outside the scope of interest, such as all-cause mortality, or lacking specificity such as cardiovascular disease or diabetes mellitus, have been excluded.

Data extraction

For each study, we collected the information of the eligible studies including the first author’s name, location, population characteristics (age, sex, race, and sample size), follow-up period, exposure definition, exposure assessment method, outcome definition, outcome ascertainment method, and covariates used in the study. Data were extracted by one author (H Liu) and checked by another author (J Wang) for accuracy. Besides, we also collected data on the range of exposure, sample size, person-years, number of events and risk estimate (RRs, HRs or ORs) and its corresponding uncertainty to conduct BPRF analysis. The uniform extraction procedures are shown in Supplementary Table 2.

We used a framework of BPRF methodology developed by Zheng et al. to assess the risk of bias in included studies [11, 15, 16]. For each included study, we extracted information concerning aspects of study design that could potentially bias the reported effect size and coded this information into study-level covariates [11]. These study-level covariates are followed as: follow-up time (≤ 10 months and > 10 months), exposure definitions, outcome definitions, effect size measures (HRs, RRs or ORs), the endpoint of outcome events (incidence or mortality), frequency of exposure measurements (single or repeat), outcome ascertainment methods (administrative records or self-reports), and the level of adjustment for relevant confounders (creating cascading dummy variables standing for the number of confounders adjusted in risk regression model from selected studies, and the minimum threshold for confounder adjustment for age and sex) [11]. These covariates would be further adjusted in our BPRF analysis if they significantly biased our estimated risk functions.

In addition to these covariates, we selected four common study characteristics that are highly relevant and likely to introduce bias, in order to evaluate the study quality [9, 17, 18]. These characteristics include the representativeness of the study population (whether it represents the general population or specific sub-groups such as high-risk populations), outcome confirmation, exposure mesurement and assessment, and control for confounding factors [11]. The quality score for each selected study was calculated by summing the scores across these four domains.

Statistical methods

The estimates for our primary indicators of this work are mean RRs across a range of exposures, BRPFs, ROSs and star ratings for each risk-outcome pair. And the exposure unit was standardized to grams of consumption per day before synthesis. For each study that reported means or quantiles consumption rather than ranges of whole grain consumption, midpoint of defined quantile as the cutoff for intake intervals was used [10, 19,20,21,22,23,24,25,26,27,28,29]. When the quantile dose range didn’t have a specific endpoint, and mean and standard deviation weren’t available, we assumed a consumption level of 0 g per day as the lowest amount [22, 24, 25, 28, 30,31,32,33]. For the upper limit of consumption, we used the range from the closest quartile or tertile within the cohort. In addition, we used 30 g to evaluate one serving of whole grain consumption if the value of a serving was not stated [34].

Estimating the shape of relationships between whole grain consumption and four health outcomes

We firstly modeled the mean log-RR (a measure of effect size) curve with MR-BRT (a Bayesian meta-regression tool) developed at the Institute for Health Metrics and Evaluation (IHME) [13, 35], and followed a uniform analysis procedure to select model specifications for all dietary risks, which is described by Zheng et al. [15]. For protective risk factors with hypothesis of monotonic deceresing, the final models were run applying quadratic splines with two internal knots and a linearity prior on the right tail [11, 15]. However, for the J-shaped risk curve, we employed quadratic splines with three internal knots and a linearity prior on both the right and left tails, without a monotonic prior [11]. Besides, to avoid the influence of extreme data and reduce publication bias, we trimmed 10% of data for each outcome as outliers [11].

Following the GRADE approach, we created binary covariates based on the the extracted information about specific study characteristics to identify potential sources of systematic bias within our included datasets. A step-wise Lasso approach were applied to assess the significance of these bias covariates at a threshold of 0.05. If the bias covariates were found to be significant, they were selected for adjustment in the final log-RR model.

To evaluate and adjust between-study heterogeneity, we quantified common sources of bias across the selected covariates that were likely to cause bias. And we calculated 95% UIs for each mean risk curve both with between-study heterogeneity incorporated (a ‘conservative’ UI) and without between-study heterogeneity incorporated (a ‘conventional’ UI) based on the selected biased covariates. Only the UIs that include between-study heterogeneity are presented in our main results unless specified.

Based on the aforementioned models, we then adjusted the selected bias covariates to decrease the variation in model residuals arising from differences in study quality and analysis. However, we primarily applied empirical evidence to choose bias covariate related to potential publication or reporting biases, which may ignore some confounders. Thus, Egger’s regression was conducted to detect publication bias, which estimated correlation between the study residuals and standard deviation of the corresponding data points. Funnel plots of the residuals of the risk function and standard deviations were generated to inspect reporting bias visually. And P value was used to assess the statistical significance of a risk for publication and/or reporting bias.

Estimating the TMREL/minimum risk exposure level

To draw robust conclusions about health benefits of whole grain consumption, we calculated the theoretical minimum risk exposure level (TMREL) of all potential outcomes linked to consuming whole grains. TMREL aligns with real-world consumption patterns supported by the included data, enabling an estimation of the average risk associated with whole grain intake. For protective risk factors, the lower bound of TMREL is defined as the 85th percentile of the lower limit within the highest consumption range across all studies. whereas the upper bound of TMREL is determined as the 85th percentile of the midpoint within the highest consumption range across all studies [15].

Estimating BPRF value, risk-outcome score (ROS) and star rating

Using the mean RR curves that incorporated between-study heterogeneity into uncertainty estimate, we estimated the BPRF from a conservative risk function. The BPRF was defined as the 5th (for harmful) or 95th percentile risk curve that is closest to the null. Afterwards, we calculated the ROS, which was equivalent to the mean log-BPRF averaged value over the 15th and 85th percentiles of the distribution of whole grain consumption. This value can give conservative interpretations regarding the association between whole grain consumption and four health outcomes [15, 36]. Then, the ROSs of risk-outcome pairs were converted into a comparison across risk-outcome pairs and a star rating (from one to five) was assigned based on the quantitative assessment of the association, where a one-star rating indicating a non-significant relationship based on the conservative interpretation, two-star through five-star ratings implying a decrease in risk with average exposure (compared to no exposure). And the ranges of ROS in 0–0.1398 stands for two-star pairs, > 0.1398–0.4055 for three-star pairs, > 0.4055–0.6152 for four-star pairs and greater than 0.6152 for five-star pairs for protective risks [11].

Sensitivity analyses

To strengthen our estimates on the association between whole grain intake and four health outcomes and reduce the impact of outliers, we used trimming analysis with the Least Trimmed Squares (LTS) method. This method automatically identifies and removes outliers within the model’s likelihood. In our study, we trimmed the top 10% of data points that deviated the most from the expected dose-response curve as part of sensitivity analysis [11, 35].

Dose-response analysis on whole grain consumption and four health outcomes was conducted by applying MR-BRT tool which included several Python packages (limetr 0.0.5, mrtool 0.0.1, IPOPT 1.2.0). And we executed BPRF analysis in Visual Studio Code with extensions of R version 4.2.1 and Python 3.9.0.

Results

Study identification

A total of 3118 articles were found using search strings, and of those we identified 28 population-based prospective cohort studies, presenting a total of 184 estimates of effect sizes for associations between whole grain consumption and the four included health outcomes [10, 25, 26, 28, 29, 32, 33]. Details of the literature search are shown in Supplementary Fig. 1. Eleven studies were from the United States, including data mainly from the HPFS, NHS, NHSII, the ATBC cohort, and the Cancer Prevention Study‑II Nutrition Cohort [10, 20, 21, 27,28,29, 32, 38, 39, 41, 42]. Fifteen of 28 studies were from the European population [19, 22,23,24,25,26, 30, 33, 25, 43,44,45,46,47]. Besides, one study (the PURE cohort) collected information covering 21 countries (including the regions of North America and Europe, South America, Africa, the Middle East, South Asia, South East Asia, and China) [48] and one study reported the role of whole grain consumption on IHD in Chinese population [40]. Detailed information about the included cohorts is displayed in Supplementary Table 3.

Characteristics of included studies

Of 28 included publications in the BPRF analysis, a total of eight studies investigated the association between whole grains and T2D [10, 28, 29, 32, 33, 37, 46, 47], seven for CRC [19,20,21, 30, 42,43,44], six for both IHD and stroke [23, 26, 39,40,41, 48], three for IHD [25, 38, 45], three for stroke [24, 27, 49], and one for IHD, stroke and CRC [22]. The median follow-up time of all included studies was 13.5 years (range: 6–25.8 years).

All included publications used dietary records or recalls, or food frequency questionnaires to collect data regarding whole grain intake. In total, seventeen publications used baseline data of whole grain intake in their analysis (single measurement) [19, 23,24,25,26, 30, 33, 39, 25, 43,44,45, 48], whereas ten considered the average whole grain intake throughout the follow-up (i.e., based on multiple measurements) as the main exposure [10, 20,21,22, 27, 37, 38, 40,41,42]. Three studies took self-report records to assess outcomes [28, 29, 32], and 25 studies used administrative medical records [10, 19, 21, 22, 25,26,27, 30, 33, 37, 41, 43, 44, 49]. Three studies used mortality as the endpoint [26, 39, 43], and the rest studies considered incidence as the endpoint. 8 studies reported effect sizes with RRs [19, 20, 24, 26, 30, 32, 33, 49], seventeen studies reported HRs [22, 23, 25, 27, 37,38,39,40,41, 48], one study reported ORs [46], and one study reported incidence rate ratios (IRRs) [44]. The detailed information is presented in Supplementary Tables 4–7.

Estimation of the shape of whole grains with T2D, CRC, IHD and stroke

Using BPRF methodology, our analyses revealed a correlation between higher whole grain intake and a reduced risk across all the outcomes considered. Figures 1, 2, 3 and 4 depict the BPRF curves for each risk-outcome pair, while Table 1 presents the results of the dose-response analysis.

Fig. 1
figure 1

BPRF analysis on the association between whole grain consumption and T2D. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

Full size image
Table 1 The relative risk at different levels of exposure, whole grains (g/day) and IHD, T2D, CRC, and stroke
Full size table
Fig. 2
figure 2

BPRF analysis on the association between whole grain consumption and CRC. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

Full size image
Fig. 3
figure 3

BPRF analysis on the association between whole grain consumption and IHD. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

Full size image
Fig. 4
figure 4

BPRF analysis on the association between whole grain consumption and stroke. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

Full size image

Specifically, our analysis revealed that the associations between whole grain consumption and the risk of T2D, CRC and IHD all exhibited non-linear, monotonically decreasing trends (Figs. 1, 2 and 3). In regard to T2D (Fig. 1a and b), the sharpest decline in risk was noted at daily consumption of 50 g, with a reduction of 34.3% (95% UI including between-study heterogeneity: 5.3 to 55.7), compared to no whole grain consumption (at 0 g per day). Nonetheless, the reduction in risk tapered off to a mere 1.2% (0.2–1.6) when comparing a consumption level of 90 g per day to 50 g per day. With respect to CRC (Fig. 2a and b), the largest reduction in CRC risk was identified when comparing the risk between an intake of 0 g per day and of 80 g per day, showcasing a noteworthy decline of 17.3%. we observed only marginal additional reductions in risk when consumption is beyond 80 g per day. As for IHD, the steepest decline of 32.1% (95% UI inclusive of between-study heterogeneity of 6.0 to 51.8) in IHD risk was observed when comparing risk between an intake of 0 g per day and of 30 g per day, with more modest marginal declines in IHD risk when consumption levels greater than 30 g per day (Fig. 3a and b).

Different from the aforementioned results, a J-shape association was found between whole grain consumption and the risk of stroke (Fig. 4a and b). The greatest reduction in stroke risk, observed at an intake of 100 g per day, was 24.6% (95% UI including between heterogeneity: 8.8 to 38.8). The mean risk of stroke at 60 g per day was 14.1% (7.0 to 21.3 including between-study heterogeneity) higher than at 100 g per day. And it was 1.5% (0.3 to 2.0) higher at 120 g per day compared to 100 g per day.

Additionally, the BPRF estimated ROSs for IHD, T2D, CRC and stroke of 0.095, 0.087, 0.068 and 0.062, respectively, which were applied to explore the average health benefits across the universe of whole grain consumption. Such estimates indicated that the consuming whole grains, on average, was related to a 9.9% decreased risk of IHD, a 9.1.% lower risk of T2D, a 7.0% lower risk of CRC and a 6.4% lower risk of stroke compared to a 0 g of whole grain intake. The star ratings of the four risk-outcome pairs all correspond to a two-star rating. After adjusting for between-study heterogeneity, the relationships still achieved statistical significance.

TMREL level of whole grain consumption

Based on observed exposure levels reported in the included studies, a TMREL of 118.5 g to 148.1 g per day, corresponding to approximately 4–5 servings per day, was estimated (detailed input information presented in Supplementary Table 8). Compared to TMREL (118.5–148.1 g), consuming no whole grains was associated with a 37.3% (5.8 to 59.5, inclusive of between-study heterogeneity) greater mean risk of T2D, a 17.3% (95% UI inclusive of between-study heterogeneity of 6.5 to 27.7) greater risk of CRC, a 36.9% (95% UI inclusive of between-study heterogeneity of 7.1 to 58.0) greater mean risk of IHD, and a 21.8% greater mean risk of stroke (95% UI inclusive of between-study heterogeneity of 7.3 to 35.1).

Sensitivity analysis and publication bias

The sensitivity analyses showed that trimming had significant effects on the ROS and reporting bias of the association between whole grain consumption and T2D and IHD. Without trimming, the ROS of T2D is -0.136, and a significant publication bias was detected using Egger’s regression (P = 0.023, as shown in Supplementary Fig. 2c). With respect to IHD, the results without trimming have reported a ROS of -0.273 and statistically significant evidence of small-study bias (P for Egger’s regression = 0.028, Supplementary Fig. 3c). However, both of the two health outcomes were found two significant study-level bias covariates (T2D: age of the population and outcome ascertainment methods; IHD: exposure measurement and outcome ascertainment methods). after removing outliers (T2D: 4 [31, 50], IHD: 5 [24, 50]) and adjusted for the selected bias covariates, no evidence of publication bias was observed (Figs. 1c and 3c). On the other hand, for CRC and stroke, trimming had a minor impact on the results. In both our analyses, with and without trimming, no significant evidence of publication bias was detected and no bias covariates were identified (Figs. 2c and 4c; and Supplementary Fig. 4c, 5c).

Discussion

In this analysis, we applied a BPRF framework, which takes into account between-study heterogeneity, to quantify the association between whole grain consumption and four health outcomes. Our results suggested that increasing the intake of whole grains was significantly related to a reduction in the risk of CRC, T2D, IHD and stroke. When comparing TMREL (118.5–148.1 g per day) with a daily intake of 0 g of whole grains, the risk reductions for four diseases (T2D, CRC, IHD, and stroke) were 37.2%, 27.3%, 26.9%, and 21.8%, respectively. For all outcomes except stroke, we observed that mean risk exhibited a non-linear, monotonic decrease as whole grain consumption increased. However, the relationship between whole grains and stroke is like a J-shaped as the risk increased with exposure levels above or below a global minimum. Based on a conservative interpretation of available data (the averaged BPRF value), we found a slight decline in the risk of stroke, CRC, T2D, and IHD compared to no whole grain intake (by at least 6.4%, 7.0%, 9.1%, and 9.9%, respectively). The converted grade ratings of our evidence were all two-star ratings.

The protective role of whole grain consumption on CRC risk is well-documented. The World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) has provided strong evidence on the protective role of whole grain consumption at 90 g/day on the risk of CRC (RR: 0.83; 95% CI: 0.79 to 0.89) [51]. Similarly, Schwingshackl et al. also reported an RR of 0.84 (95% CI: 0.78 to 0.90) of whole grain consumption at 90 g per day, and such evidence was rated as moderate according to the NutriGrade recommendation [5]. In consistent with previous studies, we found that consuming 90 g whole grains per day was also related with a similar reduction in the risk of CRC (RR: 0.83; 95% UI: 0.73 to 0.95), although the range of confidence interval was relatively wide (due to the consideration of between-study heterogeneity). Furthermore, as computing the mean across the universe of studies is appropriate to estimate the relationship between risk and outcome [52], we calculated the averaged BPRF value of 0.932, which is corresponding to a decline in the mean risk of CRC by 7.0%. Additionally, the star rating considering between-study heterogeneity of the whole grains-CRC pair is two-star. This estimate implied that the strength of the association evidence was relatively weak, in contrast to the WCRF/ARIC assessment, which categorized the evidence as “convincing” [11]. In fact, the extent of biases in nutritional epidemiological studies, including substantial residual confounding and selective reporting, can significantly impact the accuracy of health risks estimates related to the studied nutrients. Furthermore, the observational findings from prospective cohort studies exhibit considerable variation across different research endeavors [53]. Therefore, it is important to consider the strength of the association and employ a quantitative approach to assess consistency (that is, between-study heterogeneity) when evaluating evidence. Additionally, it is advisable to adopt a more conservative interpretation [54]. Our risk assessment indicated that increasing whole grain consumption can slightly reduce the risk of CRC, after correcting for biases due to factors such as study design, the representativeness of the study population, control for confounding, and so on.

With respect to IHD, the evidence stemming from previous meta-analyses has displayed a lack of consistency. For instance, Hu H et al. only found a linear association with 3 knots percentiles (25th, 50th, and 75th) selected [8]. However, Bechthold A, et al. provided evidence of a non-linear dose-response association (Pnon-linearity<0.001) for IHD using three fixed knots at 10%, 50%, and 90% through the total distribution of the reported intake [7]. The disparities in their findings could be potentially due to variations in the selection of different knot placements along the estimated risk function curve, which might influence on the resulting accuracy of a spline approximation of a curve [55]. On the other hand, BPRF analysis, according to the given degree and number of knots, automatically sampled a set of knot placements for a feasible knot distribution, evaluated each resulting model by computing its fit and curvature, and then aggregated the final model as a weighted combination of the ensemble to mitigate the effect of spline parameter selection results and draw a robust conclusion. With this methodology, we found a non-linear, monotonic decline association between whole grain consumption and IHD.

In the case of stroke, previous meta-analyses generated mixed results. Bechthold et al. observed no association between whole grain intake and the risk of stroke in the non-linear dose-response analysis [7]. Conversely, Aune et al. observed a protective role of whole grain consumption on stroke risk, but this role was only significant in their non-linear dose-response analysis, and the risk curve exhibited a J-shaped pattern [6]. The difference between these studies might be partially attributable to different included studies [6, 7]. Our analysis, including the results of newly published studies (the PURE study, China Kadoorie Biobank study and UK Biobank study) and applying BPRF methodology (free of log-linear hypothesis), found a J-shaped relationship between whole grain consumption and stroke, and we observed the greatest reduction in stroke risk observed at an intake of 100 g per day (RR: 0.75, 95%UI: 0.62 to 0.92). Unlike our analysis, both of the aforementioned studies assumed the association between whole grains and stroke to be log-linear [6, 7], which might be inappropriate. A log-linear association implies that a fixed increment of health roles of whole grain consumption (for example, 30 g/day) remains constant across all levels of intake; however, an increase in consumption from 0 to 120 g/day would not have the same impact as an increase from 240 to 360 g/day, especially considering that excessive consumption may cause health issues such as overweight [56].

Our analyses support the need for stronger efforts and policies to encourage increased whole grain consumption as a means to reduce the risk of chronic diseases. Whole grains are well-known for their abundance of dietary fiber and nutrients. However, they can also be a notable source of food-borne contaminants. Nonetheless, current evidence suggests that increasing whole grain consumption could improve public health [57]. We estimated a TMREL of 118.1–148.5 g per day as the high consumption levels of whole grain intake in the real world, and such estimates are in line with the recommended intake of whole grains promoted by the GBD and the World Health Organization (WHO) [58], which is at least 125 g per day [59]. To address both individual and environmental health, the Lancet EAT Commission recommends a primarily plant-based diet, including 232 g of whole grains per day to reduce the carbon footprint of animal-based foods [60]. Nevertheless, our analysis solely took the individual-level health benefits into consideration, and the potential environmental benefits of increased whole grain consumption were not evaluated. Based on our analyses, particularly the notable protective roles observed with daily consumption of 100 g of whole grains against the risk of stroke, it seems that incorporating a minimum of three servings of whole grains per day has the potential to lower the risk of chronic diseases.

Our study employed BPRF methodology to estimate the association between whole grain consumption and four health outcomes. Compared to traditional meta-analysis methods, this method could quantify between-study heterogeneity, and infer flexible risk functions. It does so without imposing a log-linear hypothesis, which may exaggerate risks at higher exposure levels and overlook crucial details at lower exposure levels. With this methodology, we have found that the risk curves for whole grain consumption and IHD, CRC and T2D displayed decreasing marginal returns, indicating that as whole grain intake increases, the incremental health benefits of whole grains decrease. In addition, quantifications of between-study heterogeneity and corrections for biases due to study design in the methods can contribute to a conservative interpretation and a better understanding of the protective role of whole grain consumption in real-world settings. Thirdly, by estimating RRs associated with consuming whole grains at the TMREL (in correspondence to high real-world consumption levels), we were able to provide sufficient evidence to justify more robust efforts and policies promoting increased whole grain consumption to reduce chronic disease risk, especially with regard to CRC, T2D, IHD and stroke. In general, our analysis results indicate that improving whole grain consumption is beneficial toward enhancing public health.

Although the methodological framework addressed by Zheng et al. overcame many of the limitations in existing meta-analysis approaches, this study still has several limitations. Firstly, all studies included in our analysis were observational, and we were unable to definitively assess causality. Besides, this study mainly focused on total or whole grain consumption, and the impacts of different specific subtypes of whole grains on health outcomes may vary. For example, previous reviews have indicated that increasing whole-grain breakfast cereals, other than whole-grain bread, may decrease the risk of stroke [6]; furthermore, oats or oatmeal are linked to lower all-cause mortality but show no impact on T2D and CVD incidence [6, 61]. Thus, further prospective cohort studies and randomized clinical trials focusing on different subtypes of whole grains and their associations with specific chronic diseases are required. Besides, most of the studies included were from the US and Europe, which limited the ability to make evidence-based recommendations, as dietary patterns can vary significantly between Asian and Western populations [62]. With respect to the Asian population, rather than whole grain consumption, most of studies investigated the role of refined grain consumption in the form of white rice and noodles [63], and further studies are needed to explore the association between whole grains and health outcomes on populations in Asia. Thirdly, the associations between whole grains and risks of different stroke types may be heterogeneous [24]. Unfortunately, we couldn’t investigate these associations separately due to a lack of reported data on stroke types in available studies.

Conclusion

In conclusion, the present study demonstrates that the consumption of whole grains plays a protective role in the risks of CRC, T2D, IHD and stroke, and the BPRF analysis, which did not rely on log-linear assumptions, revealed non-linear associations between whole grain intake and the four diseases of interests. The star ratings converted by ROSs for all four outcomes are all two stars, indicating that the associations between whole grain intake and CRC, T2D, IHD and stroke remain significant. The current body of evidence justifies the need for increased efforts and policies to promote higher whole grain consumption for the betterment of public health.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Cheng Z, Qiao D, Zhao S, Zhang B, Lin Q, Xie F. Whole grain rice: updated understanding of starch digestibility and the regulation of glucose and lipid metabolism. Compr Rev Food Sci Food Saf. 2022;21(4):3244–73.

    Article  CAS  PubMed  Google Scholar 

  2. Collaborators GBDRF. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of Disease Study 2019. Lancet. 2020;396(10258):1223–49.

    Article  Google Scholar 

  3. Reynolds A, Mann J, Cummings J, Winter N, Mete E, Te Morenga L. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. Lancet. 2019;393(10170):434–45.

    Article  CAS  PubMed  Google Scholar 

  4. Aune D, Norat T, Romundstad P, Vatten LJ. Whole grain and refined grain consumption and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies. Eur J Epidemiol. 2013;28(11):845–58.

    Article  CAS  PubMed  Google Scholar 

  5. Schwingshackl L, Schwedhelm C, Hoffmann G, Knuppel S, Laure Preterre A, Iqbal K, et al. Food groups and risk of colorectal cancer. Int J Cancer. 2018;142(9):1748–58.

    Article  CAS  PubMed  Google Scholar 

  6. Aune D, Keum N, Giovannucci E, Fadnes LT, Boffetta P, Greenwood DC, et al. Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2016;353:i2716.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Bechthold A, Boeing H, Schwedhelm C, Hoffmann G, Knuppel S, Iqbal K, et al. Food groups and risk of coronary heart disease, stroke and heart failure: a systematic review and dose-response meta-analysis of prospective studies. Crit Rev Food Sci Nutr. 2019;59(7):1071–90.

    Article  CAS  PubMed  Google Scholar 

  8. Hu H, Zhao Y, Feng Y, Yang X, Li Y, Wu Y, et al. Consumption of whole grains and refined grains and associated risk of cardiovascular disease events and all-cause mortality: a systematic review and dose-response meta-analysis of prospective cohort studies. Am J Clin Nutr. 2023;117(1):149–59.

    Article  PubMed  Google Scholar 

  9. Dai X, Gil GF, Reitsma MB, Ahmad NS, Anderson JA, Bisignano C, et al. Health effects associated with smoking: a Burden of Proof study. Nat Med. 2022;28(10):2045–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Hu Y, Ding M, Sampson L, Willett WC, Manson JE, Wang M, et al. Intake of whole grain foods and risk of type 2 diabetes: results from three prospective cohort studies. BMJ. 2020;370:m2206.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Zheng P, Afshin A, Biryukov S, Bisignano C, Brauer M, Bryazka D, et al. The Burden of Proof studies: assessing the evidence of risk. Nat Med. 2022;28(10):2038–44.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Schwingshackl L, Knuppel S, Schwedhelm C, Hoffmann G, Missbach B, Stelmach-Mardas M, et al. Perspective: NutriGrade: a Scoring System to assess and judge the Meta-evidence of Randomized controlled trials and Cohort studies in Nutrition Research. Adv Nutr. 2016;7(6):994–1004.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Collaborators GBDCRF. The global burden of cancer attributable to risk factors, 2010-19: a systematic analysis for the global burden of Disease Study 2019. Lancet. 2022;400(10352):563–91.

    Article  Google Scholar 

  14. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Stanaway JD, Afshin A, Ashbaugh C, Bisignano C, Brauer M, Ferrara G, et al. Health effects associated with vegetable consumption: a Burden of Proof study. Nat Med. 2022;28(10):2066–74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Lescinsky H, Afshin A, Ashbaugh C, Bisignano C, Brauer M, Ferrara G, et al. Health effects associated with consumption of unprocessed red meat: a Burden of Proof study. Nat Med. 2022;28(10):2075–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Siemieniuk R, Guyatt G. What is GRADE? BMJ Best Practice. 2021. https://bestpractice.bmj.com/info/us/toolkit/learn-ebm/what-is-grade/.

  18. The GRADE Working Group. GRADE handbook. 2013. https://gdt.gradepro.org/app/handbook/handbook.html.

  19. Pietinen P, Malila N, Virtanen M, Hartman TJ, Tangrea JA, Albanes D, et al. Diet and risk of colorectal cancer in a cohort of Finnish men. Cancer Causes Control. 1999;10(5):387–96.

    Article  CAS  PubMed  Google Scholar 

  20. He X, Wu K, Zhang X, Nishihara R, Cao Y, Fuchs CS, et al. Dietary intake of fiber, whole grains and risk of colorectal cancer: an updated analysis according to food sources, tumor location and molecular subtypes in two large US cohorts. Int J Cancer. 2019;145(11):3040–51.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Hullings AG, Sinha R, Liao LM, Freedman ND, Graubard BI, Loftfield E. Whole grain and dietary fiber intake and risk of colorectal cancer in the NIH-AARP Diet and Health Study cohort. Am J Clin Nutr. 2020;112(3):603–12.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Thompson AS, Tresserra-Rimbau A, Karavasiloglou N, Jennings A, Cantwell M, Hill C, et al. Association of Healthful Plant-based Diet Adherence with Risk of Mortality and Major Chronic diseases among adults in the UK. JAMA Netw Open. 2023;6(3):e234714.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Sonestedt E, Hellstrand S, Schulz CA, Wallstrom P, Drake I, Ericson U, et al. The association between carbohydrate-rich foods and risk of cardiovascular disease is not modified by genetic susceptibility to dyslipidemia as determined by 80 validated variants. PLoS ONE. 2015;10(4):e0126104.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Tektonidis TG, Akesson A, Gigante B, Wolk A, Larsson SC. A Mediterranean diet and risk of myocardial infarction, heart failure and stroke: a population-based cohort study. Atherosclerosis. 2015;243(1):93–8.

    Article  CAS  PubMed  Google Scholar 

  25. Helnaes A, Kyro C, Andersen I, Lacoppidan S, Overvad K, Christensen J, et al. Intake of whole grains is associated with lower risk of myocardial infarction: the Danish Diet, Cancer and Health Cohort. Am J Clin Nutr. 2016;103(4):999–1007.

    Article  CAS  PubMed  Google Scholar 

  26. Johnsen NF, Frederiksen K, Christensen J, Skeie G, Lund E, Landberg R, et al. Whole-grain products and whole-grain types are associated with lower all-cause and cause-specific mortality in the scandinavian HELGA cohort. Br J Nutr. 2015;114(4):608–23.

    Article  CAS  PubMed  Google Scholar 

  27. Juan J, Liu G, Willett WC, Hu FB, Rexrode KM, Sun Q. Whole grain consumption and risk of ischemic stroke: results from 2 prospective cohort studies. Stroke. 2017;48(12):3203–9.

    Article  PubMed  PubMed Central  Google Scholar 

  28. van Dam RM, Hu FB, Rosenberg L, Krishnan S, Palmer JR. Dietary calcium and magnesium, major food sources, and risk of type 2 diabetes in U.S. black women. Diabetes Care. 2006;29(10):2238–43.

    Article  PubMed  Google Scholar 

  29. Li J, Glenn AJ, Yang Q, Ding D, Zheng L, Bao W, et al. Dietary protein sources, mediating biomarkers, and incidence of type 2 diabetes: findings from the women’s Health Initiative and the UK Biobank. Diabetes Care. 2022;45(8):1742–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Larsson SC, Giovannucci E, Bergkvist L, Wolk A. Whole grain consumption and risk of colorectal cancer: a population-based cohort of 60,000 women. Br J Cancer. 2005;92(9):1803–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Wirstrom T, Hilding A, Gu HF, Ostenson CG, Bjorklund A. Consumption of whole grain reduces risk of deteriorating glucose tolerance, including progression to prediabetes. Am J Clin Nutr. 2013;97(1):179–87.

    Article  PubMed  Google Scholar 

  32. Meyer KA, Kushi LH, Jacobs DR Jr., Slavin J, Sellers TA, Folsom AR. Carbohydrates, dietary fiber, and incident type 2 diabetes in older women. Am J Clin Nutr. 2000;71(4):921–30.

    Article  CAS  PubMed  Google Scholar 

  33. Montonen J, Knekt P, Jarvinen R, Aromaa A, Reunanen A. Whole-grain and fiber intake and the incidence of type 2 diabetes. Am J Clin Nutr. 2003;77(3):622–9.

    Article  CAS  PubMed  Google Scholar 

  34. Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, et al. Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2011;343:d6617.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Zheng P, Barber R, Sorensen RJD, Murray CJL, Aravkin AY. Trimmed constrained mixed effects models: formulations and algorithms. J Comput Graphical Stat. 2021;30(3):544–56.

    Article  Google Scholar 

  36. Akbaraly TN, Sabia S, Shipley MJ, Batty GD, Kivimaki M. Adherence to healthy dietary guidelines and future depressive symptoms: evidence for sex differentials in the Whitehall II study. Am J Clin Nutr. 2013;97(2):419–27.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Maukonen M, Harald K, Kaartinen NE, Tapanainen H, Albanes D, Eriksson J, et al. Partial substitution of red or processed meat with plant-based foods and the risk of type 2 diabetes. Sci Rep. 2023;13(1):5874.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Hu Y, Willett WC, Manson JAE, Rosner B, Hu FB, Sun Q. Intake of whole grain foods and risk of coronary heart disease in US men and women. BMC Med. 2022;20(1):192.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Jacobs DR Jr., Andersen LF, Blomhoff R. Whole-grain consumption is associated with a reduced risk of noncardiovascular, noncancer death attributed to inflammatory diseases in the Iowa women’s Health Study. Am J Clin Nutr. 2007;85(6):1606–14.

    Article  CAS  PubMed  Google Scholar 

  40. Yang J, Du H, Guo Y, Bian Z, Yu C, Chen Y, et al. Coarse Grain Consumption and Risk of Cardiometabolic diseases: a prospective cohort study of Chinese adults. J Nutr. 2022;152(6):1476–86.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Steffen LM, Jacobs DR Jr., Stevens J, Shahar E, Carithers T, Folsom AR. Associations of whole-grain, refined-grain, and fruit and vegetable consumption with risks of all-cause mortality and incident coronary artery disease and ischemic stroke: the atherosclerosis risk in communities (ARIC) Study. Am J Clin Nutr. 2003;78(3):383–90.

    Article  CAS  PubMed  Google Scholar 

  42. Um CY, Campbell PT, Carter B, Wang Y, Gapstur SM, McCullough ML. Association between grains, gluten and the risk of colorectal cancer in the Cancer Prevention Study-II Nutrition Cohort. Eur J Nutr. 2020;59(4):1739–49.

    Article  CAS  PubMed  Google Scholar 

  43. Skeie G, Braaten T, Olsen A, Kyro C, Tjonneland A, Nilsson LM, et al. Whole grain intake and survival among scandinavian colorectal cancer patients. Nutr Cancer. 2014;66(1):6–13.

    Article  CAS  PubMed  Google Scholar 

  44. Egeberg R, Olsen A, Loft S, Christensen J, Johnsen NF, Overvad K, et al. Intake of wholegrain products and risk of colorectal cancers in the Diet, Cancer and Health cohort study. Br J Cancer. 2010;103(5):730–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Rautiainen S, Levitan EB, Orsini N, Åkesson A, Morgenstern R, Mittleman MA, et al. Total antioxidant capacity from diet and risk of myocardial infarction: a prospective cohort of women. Am J Med. 2012;125(10):974–80.

    Article  CAS  PubMed  Google Scholar 

  46. Wirström T, Hilding A, Gu HF, Östenson CG, Björklund A. Consumption of whole grain reduces risk of deteriorating glucose tolerance, including progression to prediabetes. Am J Clin Nutr. 2013;97(1):179–87.

    Article  PubMed  Google Scholar 

  47. Kyrø C, Tjønneland A, Overvad K, Olsen A, Landberg R. Higher whole-grain intake is Associated with Lower Risk of type 2 diabetes among Middle-aged men and women: the Danish Diet, Cancer, and Health Cohort. J Nutr. 2018;148(9):1434–44.

    Article  PubMed  Google Scholar 

  48. Swaminathan S, Dehghan M, Raj JM, Thomas T, Rangarajan S, Jenkins D, et al. Associations of cereal grains intake with cardiovascular disease and mortality across 21 countries in Prospective Urban and rural epidemiology study: prospective cohort study. BMJ. 2021;372:m4948.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Mizrahi A, Knekt P, Montonen J, Laaksonen MA, Heliovaara M, Jarvinen R. Plant foods and the risk of cerebrovascular diseases: a potential protection of fruit consumption. Br J Nutr. 2009;102(7):1075–83.

    Article  CAS  PubMed  Google Scholar 

  50. Kyro C, Tjonneland A, Overvad K, Olsen A, Landberg R. Higher whole-grain intake is Associated with Lower Risk of type 2 diabetes among Middle-aged men and women: the Danish Diet, Cancer, and Health Cohort. J Nutr. 2018;148(9):1434–44.

    Article  PubMed  Google Scholar 

  51. Vieira AR, Abar L, Chan DSM, Vingeliene S, Polemiti E, Stevens C, et al. 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. 2017;28(8):1788–802.

    Article  CAS  PubMed  Google Scholar 

  52. Murray CJ, Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S. Comparative quantification of health risks conceptual framework and methodological issues. Popul Health Metr. 2003;1(1):1.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Trepanowski JF, Ioannidis JPA, Perspective. Limiting dependence on Nonrandomized studies and improving randomized trials in Human Nutrition Research: why and how. Adv Nutr. 2018;9(4):367–77.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Ioannidis JPA. The challenge of reforming nutritional epidemiologic research. JAMA. 2018;320(10):969–70.

    Article  PubMed  Google Scholar 

  55. Yeh R, Nashed YSG, Peterka T, Tricoche X. Fast automatic knot Placement Method for Accurate B-spline curve fitting. Comput Aided Des. 2020;128.

  56. San-Cristobal R, Navas-Carretero S, Martinez-Gonzalez MA, Ordovas JM, Martinez JA. Contribution of macronutrients to obesity: implications for precision nutrition. Nat Rev Endocrinol. 2020;16(6):305–20.

    Article  PubMed  Google Scholar 

  57. Thielecke F, Nugent AP. Contaminants in Grain-A Major Risk for Whole Grain Safety? Nutrients. 2018;10(9).

  58. World Health Organization. Diet, nutrition and the prevention of chronic diseases: Report of a joint WHO/FAO expert consultation. 2018.

  59. Collaborators GBD. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the global burden of Disease Study 2017. Lancet. 2019;393(10184):1958–72.

    Article  Google Scholar 

  60. Willett W, Rockstrom J, Loken B, Springmann M, Lang T, Vermeulen S, et al. Food in the Anthropocene: the EAT-Lancet Commission on healthy diets from sustainable food systems. Lancet. 2019;393(10170):447–92.

    Article  PubMed  Google Scholar 

  61. Wehrli F, Taneri PE, Bano A, Bally L, Blekkenhorst LC, Bussler W et al. Oat intake and risk of type 2 diabetes, Cardiovascular Disease and all-cause mortality: a systematic review and Meta-analysis. Nutrients. 2021;13(8).

  62. Zhou BF, Stamler J, Dennis B, Moag-Stahlberg A, Okuda N, Robertson C, et al. Nutrient intakes of middle-aged men and women in China, Japan, United Kingdom, and United States in the late 1990s: the INTERMAP study. J Hum Hypertens. 2003;17(9):623–30.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Muthayya S, Sugimoto JD, Montgomery S, Maberly GF. An overview of global rice production, supply, trade, and consumption. Ann N Y Acad Sci. 2014;1324:7–14.

    Article  PubMed  Google Scholar 

Download references

Funding

This study was supported by the Key Discipline of Zhejiang Province in Public Health and Preventative Medicine (First Class, Category A), Hangzhou Medical College.

Author information

Author notes

  1. Jing Wang and Yingjun Li contributed equally to this work.

Authors and Affiliations

  1. Department of Epidemiology and Health Statistics, School of Public Health, Hangzhou Medical College, 481 Binwen Road, Hangzhou, 310053, China

    Houpu Liu, Jiahao Zhu, Rui Gao, Lilu Ding, Ye Yang, Wenxia Zhao, Jing Wang & Yingjun Li

  2. Department of Radiology, Key Laboratory of Cancer Prevention and Therapy, Tianjin Medical University Cancer Institute & Hospital, National Clinical Research Center for Cancer, Tianjin, China

    Xiaonan Cui

  3. Department of Epidemiology and Health Statistics, School of Public health, Tianjin Medical University, Tianjin, China

    Wenli Lu

Authors
  1. Houpu Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jiahao Zhu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Rui Gao
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Lilu Ding
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Ye Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Wenxia Zhao
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Xiaonan Cui
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Wenli Lu
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Jing Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Yingjun Li
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

All authors contributed to the study’s conception and design. Liu Houpu, Wang, Jing, and Li, Yingjun conceived the study, searched the literature and performed data extraction. Liu, Houpu and Zhu, Jiahao had the idea for the article, performed the main analysis and written the first manuscript. Gao Rui, Yang Ye and Zhao Wenxia conducted the sensitivity analysis and inspected the analysis results. Ding Lilu, Cui, Xiaonan, and Lu, Wenli provided statistical expertise. Wang, Jing and Li, Yingjun reviewed the original manuscript and critically revised the work. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Jing Wang or Yingjun Li.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Patient and public involvement

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, H., Zhu, J., Gao, R. et al. Estimating effects of whole grain consumption on type 2 diabetes, colorectal cancer and cardiovascular disease: a burden of proof study. Nutr J 23, 49 (2024). https://doi.org/10.1186/s12937-024-00957-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12937-024-00957-x

Keywords

Nutrition Journal

ISSN: 1475-2891

Contact us