-
1.
Fructose-containing food sources and blood pressure: A systematic review and meta-analysis of controlled feeding trials.
Liu, Q, Chiavaroli, L, Ayoub-Charette, S, Ahmed, A, Khan, TA, Au-Yeung, F, Lee, D, Cheung, A, Zurbau, A, Choo, VL, et al
PloS one. 2023;(8):e0264802
Abstract
Whether food source or energy mediates the effect of fructose-containing sugars on blood pressure (BP) is unclear. We conducted a systematic review and meta-analysis of the effect of different food sources of fructose-containing sugars at different levels of energy control on BP. We searched MEDLINE, Embase and the Cochrane Library through June 2021 for controlled trials ≥7-days. We prespecified 4 trial designs: substitution (energy matched substitution of sugars); addition (excess energy from sugars added); subtraction (excess energy from sugars subtracted); and ad libitum (energy from sugars freely replaced). Outcomes were systolic and diastolic BP. Independent reviewers extracted data. GRADE assessed the certainty of evidence. We included 93 reports (147 trial comparisons, N = 5,213) assessing 12 different food sources across 4 energy control levels in adults with and without hypertension or at risk for hypertension. Total fructose-containing sugars had no effect in substitution, subtraction, or ad libitum trials but decreased systolic and diastolic BP in addition trials (P<0.05). There was evidence of interaction/influence by food source: fruit and 100% fruit juice decreased and mixed sources (with sugar-sweetened beverages [SSBs]) increased BP in addition trials and the removal of SSBs (linear dose response gradient) and mixed sources (with SSBs) decreased BP in subtraction trials. The certainty of evidence was generally moderate. Food source and energy control appear to mediate the effect of fructose-containing sugars on BP. The evidence provides a good indication that fruit and 100% fruit juice at low doses (up to or less than the public health threshold of ~10% E) lead to small, but important reductions in BP, while the addition of excess energy of mixed sources (with SSBs) at high doses (up to 23%) leads to moderate increases and their removal or the removal of SSBs alone (up to ~20% E) leads to small, but important decreases in BP in adults with and without hypertension or at risk for hypertension. Trial registration: Clinicaltrials.gov: NCT02716870.
-
2.
Important Food Sources of Fructose-Containing Sugars and Non-Alcoholic Fatty Liver Disease: A Systematic Review and Meta-Analysis of Controlled Trials.
Lee, D, Chiavaroli, L, Ayoub-Charette, S, Khan, TA, Zurbau, A, Au-Yeung, F, Cheung, A, Liu, Q, Qi, X, Ahmed, A, et al
Nutrients. 2022;(14)
Abstract
Background: Fructose providing excess calories in the form of sugar sweetened beverages (SSBs) increases markers of non-alcoholic fatty liver disease (NAFLD). Whether this effect holds for other important food sources of fructose-containing sugars is unclear. To investigate the role of food source and energy, we conducted a systematic review and meta-analysis of controlled trials of the effect of fructose-containing sugars by food source at different levels of energy control on non-alcoholic fatty liver disease (NAFLD) markers. Methods and Findings: MEDLINE, Embase, and the Cochrane Library were searched through 7 January 2022 for controlled trials ≥7-days. Four trial designs were prespecified: substitution (energy-matched substitution of sugars for other macronutrients); addition (excess energy from sugars added to diets); subtraction (excess energy from sugars subtracted from diets); and ad libitum (energy from sugars freely replaced by other macronutrients). The primary outcome was intrahepatocellular lipid (IHCL). Secondary outcomes were alanine aminotransferase (ALT) and aspartate aminotransferase (AST). Independent reviewers extracted data and assessed risk of bias. The certainty of evidence was assessed using GRADE. We included 51 trials (75 trial comparisons, n = 2059) of 10 food sources (sugar-sweetened beverages (SSBs); sweetened dairy alternative; 100% fruit juice; fruit; dried fruit; mixed fruit sources; sweets and desserts; added nutritive sweetener; honey; and mixed sources (with SSBs)) in predominantly healthy mixed weight or overweight/obese younger adults. Total fructose-containing sugars increased IHCL (standardized mean difference = 1.72 [95% CI, 1.08 to 2.36], p < 0.001) in addition trials and decreased AST in subtraction trials with no effect on any outcome in substitution or ad libitum trials. There was evidence of influence by food source with SSBs increasing IHCL and ALT in addition trials and mixed sources (with SSBs) decreasing AST in subtraction trials. The certainty of evidence was high for the effect on IHCL and moderate for the effect on ALT for SSBs in addition trials, low for the effect on AST for the removal of energy from mixed sources (with SSBs) in subtraction trials, and generally low to moderate for all other comparisons. Conclusions: Energy control and food source appear to mediate the effect of fructose-containing sugars on NAFLD markers. The evidence provides a good indication that the addition of excess energy from SSBs leads to large increases in liver fat and small important increases in ALT while there is less of an indication that the removal of energy from mixed sources (with SSBs) leads to moderate reductions in AST. Varying uncertainty remains for the lack of effect of other important food sources of fructose-containing sugars at different levels of energy control.
-
3.
Effect of Important Food Sources of Fructose-Containing Sugars on Inflammatory Biomarkers: A Systematic Review and Meta-Analysis of Controlled Feeding Trials.
Qi, X, Chiavaroli, L, Lee, D, Ayoub-Charette, S, Khan, TA, Au-Yeung, F, Ahmed, A, Cheung, A, Liu, Q, Blanco Mejia, S, et al
Nutrients. 2022;(19)
Abstract
BACKGROUND Fructose-containing sugars as sugar-sweetened beverages (SSBs) may increase inflammatory biomarkers. Whether this effect is mediated by the food matrix at different levels of energy is unknown. To investigate the role of food source and energy, we conducted a systematic review and meta-analysis of controlled trials on the effect of different food sources of fructose-containing sugars on inflammatory markers at different levels of energy control. METHODS MEDLINE, Embase, and the Cochrane Library were searched through March 2022 for controlled feeding trials ≥ 7 days. Four trial designs were prespecified by energy control: substitution (energy matched replacement of sugars); addition (excess energy from sugars added to diets); subtraction (energy from sugars subtracted from diets); and ad libitum (energy from sugars freely replaced). The primary outcome was C-reactive protein (CRP). Secondary outcomes were tumour necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6). Independent reviewers extracted data and assessed risk of bias. GRADE assessed certainty of evidence. RESULTS We identified 64 controlled trials (91 trial comparisons, n = 4094) assessing 12 food sources (SSB; sweetened dairy; sweetened dairy alternative [soy]; 100% fruit juice; fruit; dried fruit; mixed fruit forms; sweetened cereal grains and bars; sweets and desserts; added nutritive [caloric] sweetener; mixed sources [with SSBs]; and mixed sources [without SSBs]) at 4 levels of energy control over a median 6-weeks in predominantly healthy mixed weight or overweight/obese adults. Total fructose-containing sugars decreased CRP in addition trials and had no effect in substitution, subtraction or ad libitum trials. No effect was observed on other outcomes at any level of energy control. There was evidence of interaction/influence by food source: substitution trials (sweetened dairy alternative (soy) and 100% fruit juice decreased, and mixed sources (with SSBs) increased CRP); and addition trials (fruit decreased CRP and TNF-α; sweets and desserts (dark chocolate) decreased IL-6). The certainty of evidence was moderate-to-low for the majority of analyses. CONCLUSIONS Food source appears to mediate the effect of fructose-containing sugars on inflammatory markers over the short-to-medium term. The evidence provides good indication that mixed sources that contain SSBs increase CRP, while most other food sources have no effect with some sources (fruit, 100% fruit juice, sweetened soy beverage or dark chocolate) showing decreases, which may be dependent on energy control. CLINICALTRIALS gov: (NCT02716870).
-
4.
Systematic review and meta-analysis examining the relationship between postprandial hypotension, cardiovascular events, and all-cause mortality.
Jenkins, DJA, Sahye-Pudaruth, S, Khodabandehlou, K, Liang, F, Kasmani, M, Wanyan, J, Wang, M, Selvaganesh, K, Paquette, M, Patel, D, et al
The American journal of clinical nutrition. 2022;(3):663-671
-
-
Free full text
-
Abstract
BACKGROUND Postprandial hypotension (PPH) has been reported to be associated with syncope, falls, adverse cardiovascular outcomes, and increased all-cause mortality. It has been reported to have an incidence as high as 30% in the elderly and persons with diabetes. We therefore performed a meta-analysis to determine the relation of PPH with cardiovascular disease (CVD) events and all-cause mortality. OBJECTIVES Our objective was to conduct a systematic review and meta-analysis of cohort and cross-sectional studies to determine the association of PPH with CVD and all-cause mortality. METHODS We searched the databases MEDLINE, EMBASE, and Cochrane library up to 13 April 2022 for prospective cohort and cross-sectional studies that examined the association of PPH with CVD outcomes and all-cause mortality. Data were analyzed using the generic inverse variance method with a random-effects model. Grading of Recommendations, Assessment, Development, and Evaluation approach assessed the certainty of evidence. RESULTS Seven studies that included 2389 participants met our inclusion criteria. PPH was associated with each outcome individually, including increased all-cause mortality, total CVD, CVD mortality, and stroke. CVD outcomes and all-cause mortality combined were also associated with PPH (RR: 1.52; 95% CI: 1.05, 2.18; P = 0.03; I2 = 77%). The certainty of evidence was graded as very low due to significant heterogeneity and the limited number of studies. CONCLUSIONS This assessment indicates an association of PPH with CVD and all-cause mortality. Further studies are required to improve CVD and mortality estimates, but the potential seriousness of CVD and all-cause mortality as outcomes of PPH justifies more screening, diagnosis, and research.
-
5.
White rice, brown rice and the risk of type 2 diabetes: a systematic review and meta-analysis.
Yu, J, Balaji, B, Tinajero, M, Jarvis, S, Khan, T, Vasudevan, S, Ranawana, V, Poobalan, A, Bhupathiraju, S, Sun, Q, et al
BMJ open. 2022;12(9):e065426
-
-
-
Free full text
Plain language summary
The prevalence of type 2 diabetes (T2D) has continued to increase worldwide, especially in low-income and middle-income countries. Since rice is a staple food in many cultures and is predominant in most Asian diets, it is hypothesised that improving diet quality by replacing white rice with brown rice could play an important role in the prevention of T2D. The aim of this study was to assess the relationship between rice intake and the risk of T2D. This study is a systematic review and meta-analysis of nineteen studies – 8 cohort studies and 11 randomised controlled trials. Results from the meta-analyses of the cohort studies show a positive association between intake of white rice and risk of T2D. The associations were stronger in women compared with men. Additionally, brown rice was inversely associated with risk of T2D however, the results are based on limited data. Furthermore, the randomised controlled trials showed that the between-group difference in high-density lipoprotein cholesterol was statistically significant in favour of the brown rice group. Authors conclude that replacing white rice with brown rice or other whole grains has the potential to be a low-cost and feasible lifestyle strategy to improve diet quality and help reduce T2D risk.
Abstract
OBJECTIVE Intake of white rice has been associated with elevated risk for type 2 diabetes (T2D), while studies on brown rice are conflicting. To inform dietary guidance, we synthesised the evidence on white rice and brown rice with T2D risk. DESIGN Systematic review and meta-analysis. DATA SOURCES PubMed, EMBASE and Cochrane databases were searched through November 2021. ELIGIBILITY CRITERIA Prospective cohort studies of white and brown rice intake on T2D risk (≥1 year), and randomised controlled trials (RCTs) comparing brown rice with white rice on cardiometabolic risk factors (≥2 weeks). DATA EXTRACTION AND SYNTHESIS Data were extracted by the primary reviewer and two additional reviewers. Meta-analyses were conducted using random-effects models and reporting followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Risk of bias was assessed using the Newcastle Ottawa Scale for prospective cohort studies and the Cochrane Risk of Bias Tool for RCTs. Strength of the meta-evidence was assessed using NutriGrade. RESULTS Nineteen articles were included: 8 cohort studies providing 18 estimates (white rice: 15 estimates, 25 956 cases, n=5 77 426; brown rice: 3 estimates, 10 507 cases, n=1 97 228) and 11 RCTs (n=1034). In cohort studies, white rice was associated with higher risk of T2D (pooled RR, 1.16; 95% CI: 1.02 to 1.32) comparing extreme categories. At intakes above ~300 g/day, a dose-response was observed (each 158 g/day serving was associated with 13% (11%-15%) higher risk of T2D). Intake of brown rice was associated with lower risk of T2D (pooled RR, 0.89; 95% CI: 0.81 to 0.97) comparing extreme categories. Each 50 g/day serving of brown rice was associated with 13% (6%-20%) lower risk of T2D. Cohort studies were considered to be of good or fair quality. RCTs showed an increase in high-density lipoprotein-cholesterol (0.06 mmol/L; 0.00 to 0.11 mmol/L) in the brown compared with white rice group. No other significant differences in risk factors were observed. The majority of RCTs were found to have some concern for risk of bias. Overall strength of the meta-evidence was moderate for cohort studies and moderate and low for RCTs. CONCLUSION Intake of white rice was associated with higher risk of T2D, while intake of brown rice was associated with lower risk. Findings from substitution trials on cardiometabolic risk factors were inconsistent. PROSPERO REGISTRATION NUMBER CRD42020158466.
-
6.
Nut consumption and type 2 diabetes risk: a systematic review and meta-analysis of observational studies.
Becerra-Tomás, N, Paz-Graniel, I, Hernández-Alonso, P, Jenkins, DJA, Kendall, CWC, Sievenpiper, JL, Salas-Salvadó, J
The American journal of clinical nutrition. 2021;(4):960-971
-
-
Free full text
-
Abstract
BACKGROUND Previous meta-analyses, with some methodological controversies, have assessed the relation between nut consumption and type 2 diabetes (T2D) risk and pointed to contradictory results, making desirable the performance of an updated meta-analysis. OBJECTIVES We aimed to systematically review and meta-analyze all the published studies investigating the relations of total nuts and different types of nuts-i.e., walnuts, peanuts, peanut butter, and total tree nuts-with the prevalence and incidence of T2D. METHODS A systematic search was conducted in the PubMed and Cochrane databases through 12 August, 2020. The inverse variance method with fixed-effect models was used to pool data across studies, expressed as risk ratios (RRs) or ORs and 95% CIs for prospective cohort and cross-sectional studies, respectively. The Cochran Q test and I2 statistics were used to test and quantify heterogeneity, respectively. Dose-response meta-analysis was also conducted. RESULTS Eight studies (5 prospective and 3 cross-sectional) were included in the quantitative synthesis. Meta-analyses of cross-sectional studies and prospective cohort studies, comparing the highest with the lowest categories, revealed a nonsignificant association between total nut consumption and T2D. Meta-analyses of prospective cohort studies showed an inverse association between peanut butter consumption and T2D incidence (RR: 0.87; 95% CI: 0.77, 0.98; I2 = 50.6%; Pheterogeneity = 0.16), whereas no association was observed between peanuts or tree nuts and T2D. There was no evidence of a linear dose-response or nonlinear dose-response gradient for total nut and peanut consumption in prospective cohort studies. The certainty of the evidence using NutriGrade was very low for all the exposures. CONCLUSIONS Current results do not demonstrate an association of total nut, peanut, or tree nut consumption with T2D. Peanut butter consumption may be inversely associated with this disease.This review protocol was registered at www.crd.york.ac.uk/prospero/ as CRD42020149756.
-
7.
Are fatty nuts a weighty concern? A systematic review and meta-analysis and dose-response meta-regression of prospective cohorts and randomized controlled trials.
Nishi, SK, Viguiliouk, E, Blanco Mejia, S, Kendall, CWC, Bazinet, RP, Hanley, AJ, Comelli, EM, Salas Salvadó, J, Jenkins, DJA, Sievenpiper, JL
Obesity reviews : an official journal of the International Association for the Study of Obesity. 2021;(11):e13330
-
-
Free full text
-
Abstract
Nuts are recommended for cardiovascular health, yet concerns remain that nuts may contribute to weight gain due to their high energy density. A systematic review and meta-analysis of prospective cohorts and randomized controlled trials (RCTs) was conducted to update the evidence, provide a dose-response analysis, and assess differences in nut type, comparator and more in subgroup analyses. MEDLINE, EMBASE, and Cochrane were searched, along with manual searches. Data from eligible studies were pooled using meta-analysis methods. Interstudy heterogeneity was assessed (Cochran Q statistic) and quantified (I2 statistic). Certainty of the evidence was assessed by Grading of Recommendations Assessment, Development, and Evaluation (GRADE). Six prospective cohort studies (7 unique cohorts, n = 569,910) and 86 RCTs (114 comparisons, n = 5873) met eligibility criteria. Nuts were associated with lower incidence of overweight/obesity (RR 0.93 [95% CI 0.88 to 0.98] P < 0.001, "moderate" certainty of evidence) in prospective cohorts. RCTs presented no adverse effect of nuts on body weight (MD 0.09 kg, [95% CI -0.09 to 0.27 kg] P < 0.001, "high" certainty of evidence). Meta-regression showed that higher nut intake was associated with reductions in body weight and body fat. Current evidence demonstrates the concern that nut consumption contributes to increased adiposity appears unwarranted.
-
8.
Supplemental Vitamins and Minerals for Cardiovascular Disease Prevention and Treatment: JACC Focus Seminar.
Jenkins, DJA, Spence, JD, Giovannucci, EL, Kim, YI, Josse, RG, Vieth, R, Sahye-Pudaruth, S, Paquette, M, Patel, D, Blanco Mejia, S, et al
Journal of the American College of Cardiology. 2021;(4):423-436
Abstract
This is an update of the previous 2018 systematic review and meta-analysis of vitamin and mineral supplementation on cardiovascular disease outcomes and all-cause mortality. New randomized controlled trials and meta-analyses were identified by searching the Cochrane library, Medline, and Embase, and data were analyzed using random effects models and classified by the Grading of Recommendations Assessment Development and Evaluation approach. This updated review shows similar findings to the previous report for preventive benefits from both folic acid and B vitamins for stroke and has been graded with moderate quality. No effect was seen for the commonly used multivitamins, vitamin D, calcium, and vitamin C, and an increased risk was seen with niacin (with statin) for all-cause mortality. Conclusive evidence for the benefit of supplements across different dietary backgrounds, when the nutrient is sufficient, has not been demonstrated.
-
9.
Effect of low glycaemic index or load dietary patterns on glycaemic control and cardiometabolic risk factors in diabetes: systematic review and meta-analysis of randomised controlled trials.
Chiavaroli, L, Lee, D, Ahmed, A, Cheung, A, Khan, TA, Blanco, S, Mejia, , Mirrahimi, A, Jenkins, DJA, Livesey, G, et al
BMJ (Clinical research ed.). 2021;:n1651
Abstract
OBJECTIVE To inform the update of the European Association for the Study of Diabetes clinical practice guidelines for nutrition therapy. DESIGN Systematic review and meta-analysis of randomised controlled trials. DATA SOURCES Medline, Embase, and the Cochrane Library searched up to 13 May 2021. ELIGIBILITY CRITERIA FOR SELECTING STUDIES Randomised controlled trials of three or more weeks investigating the effect of diets with low glycaemic index (GI)/glycaemic load (GL) in diabetes. OUTCOME AND MEASURES The primary outcome was glycated haemoglobin (HbA1c). Secondary outcomes included other markers of glycaemic control (fasting glucose, fasting insulin); blood lipids (low density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol (HDL-C), non-HDL-C, apo B, triglycerides); adiposity (body weight, BMI (body mass index), waist circumference), blood pressure (systolic blood pressure (SBP) and diastolic blood pressure (DBP)), and inflammation (C reactive protein (CRP)). DATA EXTRACTION AND SYNTHESIS Two independent reviewers extracted data and assessed risk of bias. Data were pooled by random effects models. GRADE (grading of recommendations assessment, development, and evaluation) was used to assess the certainty of evidence. RESULTS 29 trial comparisons were identified in 1617 participants with type 1 and 2 diabetes who were predominantly middle aged, overweight, or obese with moderately controlled type 2 diabetes treated by hyperglycaemia drugs or insulin. Low GI/GL dietary patterns reduced HbA1c in comparison with higher GI/GL control diets (mean difference −0.31% (95% confidence interval −0.42 to −0.19%), P<0.001; substantial heterogeneity, I2=75%, P<0.001). Reductions occurred also in fasting glucose, LDL-C, non-HDL-C, apo B, triglycerides, body weight, BMI, systolic blood pressure (dose-response), and CRP (P<0.05), but not blood insulin, HDL-C, waist circumference, or diastolic blood pressure. A positive dose-response gradient was seen for the difference in GL and HbA1c and for absolute dietary GI and SBP (P<0.05). The certainty of evidence was high for the reduction in HbA1c and moderate for most secondary outcomes, with downgrades due mainly to imprecision. CONCLUSIONS This synthesis suggests that low GI/GL dietary patterns result in small important improvements in established targets of glycaemic control, blood lipids, adiposity, blood pressure, and inflammation beyond concurrent treatment with hyperglycaemia drugs or insulin, predominantly in adults with moderately controlled type 1 and type 2 diabetes. The available evidence provides a good indication of the likely benefit in this population. STUDY REGISTRATION ClinicalTrials.gov NCT04045938.
-
10.
Selenium, antioxidants, cardiovascular disease, and all-cause mortality: a systematic review and meta-analysis of randomized controlled trials.
Jenkins, DJA, Kitts, D, Giovannucci, EL, Sahye-Pudaruth, S, Paquette, M, Blanco Mejia, S, Patel, D, Kavanagh, M, Tsirakis, T, Kendall, CWC, et al
The American journal of clinical nutrition. 2020;112(6):1642-1652
-
-
-
Free full text
-
Plain language summary
Oxidative damage is a shared characteristic in chronic diseases such as cardiovascular disease (CVD), diabetes, cancer and ageing. Antioxidants mitigate the impact of oxidants and have been widely investigated in ageing and disease. However, the evidence for supplementary antioxidants has been mixed and some authorities have advised against the use of certain single nutrients for the prevention of CVD or cancer. This systematic review and meta-analysis focused on selenium due to its vital role in the antioxidant system and associations of low selenium blood levels with increased risk of CVD, cancers and death. The study included 43 randomised controlled trials (RCTs) evaluating the effect of supplemental selenium and antioxidants with or without selenium and their impact on CVD risk, cancer and all-cause mortality. Overall supplemental selenium or antioxidants alone did not seem to be associated with CVD outcomes, cancer, CVD and cancer mortality, or all-cause mortality. On close examination, a decreased risk was seen for CVD mortality when antioxidants were combined with selenium, whilst antioxidant mixtures without selenium demonstrated an increased risk in all-cause mortality. The findings did not seem to be influenced by dietary selenium intake. The authors suggested that inclusion of selenium as part of an antioxidant mix could be key for an antioxidant associated risk reduction. However, in the absence of further long term studies, a balanced antioxidant-rich diet was advocated as the safest approach. In clinical practice, where antioxidant support beyond diet is warranted, supplemental antioxidant use should be concurrent with adequate selenium supplementation, with dose benefits of 50-200mcg observed.
Abstract
BACKGROUND Antioxidants have been promoted for cardiovascular disease (CVD) risk reduction and for the prevention of cancer. Our preliminary analysis suggested that only when selenium was present were antioxidant mixtures associated with reduced all-cause mortality. OBJECTIVE We conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) to determine the effect of selenium supplementation alone and of antioxidant mixtures with or without selenium on the risk of CVD, cancer, and mortality. METHODS We identified studies using the Cochrane Library, Medline, and Embase for potential CVD outcomes, cancer, and all-cause mortality following selenium supplementation alone or after antioxidant supplement mixtures with and without selenium up to June 5, 2020. RCTs of ≥24 wk were included and data were analyzed using random-effects models and classified by the Grading of Recommendations, Assessment, Development, and Evaluation approach. RESULTS The meta-analysis identified 9423 studies, of which 43 were used in the final analysis. Overall, no association of selenium alone or antioxidants was seen with CVD and all-cause mortality. However, a decreased risk with antioxidant mixtures was seen for CVD mortality when selenium was part of the mix (RR: 0.77; 95% CI: 0.62, 0.97; P = 0.02), with no association when selenium was absent. Similarly, when selenium was part of the antioxidant mixture, a decreased risk was seen for all-cause mortality (RR: 0.90; 95% CI: 0.82, 0.98; P = 0.02) as opposed to an increased risk when selenium was absent (RR: 1.09; 95% CI: 1.04, 1.13; P = 0.0002). CONCLUSION The addition of selenium should be considered for supplements containing antioxidant mixtures if they are to be associated with CVD and all-cause mortality risk reduction. This trial was registered at https://www.crd.york.ac.uk/PROSPERO/ as CRD42019138268.