Wholegrain cereals for coronary heart disease
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
The aim of this systematic review is to assess the effect of wholegrain foods or diets on total mortality from coronary heart disease (CHD), CHD events and total mortality, using all available randomised controlled trials and concurrent controlled trials and meta‐analytic techniques where appropriate.
Background
In western society, coronary heart disease (CHD) is the major cause of death. The prevalence of CHD is increasing worldwide (Murray 1997). In the UK, there were 238 deaths from CHD per 100,000 population in 1997 (OHE 1999). In England and Wales people aged over 65, one fifth of men and one eighth of women were treated for CHD in general practice in 1996. (Carter 1999). CHD costs the health care system in the UK about £1,750 million a year (BHF 2004). In addition, CHD costs the UK economy about £5,300 million because of days lost due to death, illness and informal care of people with the disease (BHF 2004).
A wholegrain contains the entire edible parts of a natural grain kernel. The structure of all whole grains is similar and includes the endosperm, germ and bran. Wholegrains are rich in a wide range of compounds with known health effects (Slavin 2003). In the grain‐refining process, most of the bran and some of the germ is removed, resulting in the loss of dietary fibre, vitamins, minerals, lignans, phyto‐oestrogens, phenolic compunds and phytic acid. The remaining starchy endosperm is ground to produce refined white flours.
Important grains in the western diet include wheat, rice, maize and oats. Wholemeal foods are made from wholegrains which have been milled to a finer texture rather than leaving them whole in the final product. Both wholegrain and wholemeal cereal foods are those grain foods which include the outer layers of the grain, including the bran and germ. Food Standards Australia and New Zealand (FSANZ) have proposed a definition of the term wholegrain as ' wholegrain is the intact, dehulled, ground, cracked or flaked grains where the components of the endosperm, germ and bran are present in substantially the same proportions as they exist in the intact grain' (FSANZ 2003). The proposal also relates to foods made from whole grains such as breads, breakfast cereals, pasta, biscuits, oats, rice and grain‐based snack foods. Research has shown that such processing of whole‐grains does not remove biologically important compounds (Slavin 2001). Nutritionally wholegrain and wholemeal foods are similar.
The US Food and Drug Administration (FDA) does not define wholegrain however since 1999 has permitted food manufacturers to make the following health claim on wholegrain food products 'diets rich in wholegrain foods and other plant foods and low in total fat, saturated fat and cholesterol may reduce the risk of heart disease and some cancers' (FDA 1999). For the purpose of this health claim, wholegrain foods must contain 51% or more of wholegrain ingredients (bran, germ and endosperm) by weight per reference amount.
A few epidemiological studies have specifically examined the association between intakes of wholegrain foods and risk of CHD. The Iowa Women's Health study (Jacobs 1998; Jacobs 1999) of 34,492 post menopausal women followed for 6 years found that a greater intake of wholegrain was associated with a reduced risk of death from CHD. The 10‐year Nurses' Health Study, a large prospective study of 75,521 women aged 38‐63, found that increased whole‐grain intake was associated with decreased risk of CHD (Liu 1999). The lower risk associated with higher wholegrain intake was not fully explained by its contribution to intakes of dietary fibre, folate, vitamin B6 and vitamin E. The Atherosclerosis Risk in Communities (ARIC) study found a beneficial relationship between wholegrain consumption and the risk of total mortality and the incidence of coronary artery disease but not with the risk of ischemic stroke (Steffen 2003). The study followed 15,792 people aged 45‐64 for 11 years.
Other studies have focused on the relationship between cereal fibre and CHD. The amount of cereal fibre is much greater in wholegrain foods than in refined grain products so the cereal fibre intake is likely to mirror the wholegrain intake. In a large prospective study of 43, 757 male health professionals, cereal fibre was associated with reduced risk for myocardial infarction, with a 29% decrease in risk for each 10g increase in cereal fibre (Rimm 1996). A significant inverse association between total dietary fibre intake and risk of CHD was found in another large study of 68,782 women who were followed for 10 years (Wolk 1999). There was a significant inverse relation with cereal fibre but not with vegetable or fruit fibre. Women in the highest quintile of cereal fibre had a 34% lower risk of CHD events compared with those in the lowest quintile.
Associations between wholegrain consumption and risk factors for CHD have also been reported. In the Framingham Offspring study, diets rich in wholegrains were inversely associated with total cholesterol, LDL cholesterol and body mass index (McKeown 2002). A prospective study of women aged 38 to 63 years found that substituting wholegrain products for refined grain products may decrease the risk of Type 2 diabetes mellitus (Liu 2000). Diabetes is a major risk factor for CHD. Reduced insulin sensitivity is another factor that contributes to the metabolic environment that predisposes to CHD (Reaven 1993). There is evidence of reduced insulin sensitivity at diagnosis in 60% of patients with CHD (Feranninni 1991). Whole‐grain consumption improves insulin sensitivity in overweight and obese adults (Pereira 2002).
Objectives
The aim of this systematic review is to assess the effect of wholegrain foods or diets on total mortality from coronary heart disease (CHD), CHD events and total mortality, using all available randomised controlled trials and concurrent controlled trials and meta‐analytic techniques where appropriate.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled studies (RCTs). Cross‐over studies and parallel studies will be included. Trials will only be included if outcome data can be collected (by communication with authors where necessary).
Types of participants
Non‐institutionalised adults of either sex with diagnosed CHD, or with at least one major risk factor for CHD, will be included. Major risk factors for CHD include overweight (body mass index >25), abdominal obesity, diabetes, family history of CHD, raised blood pressure, hypertension, abnormal lipid levels (high density lipoproteins (HDL) and low density lipoproteins (LDL) cholesterol, triglycerides or total cholesterol), impaired glucose tolerance, reduced insulin sensitivity, insulin resistance, hyperinsulinaemia, hyperglycaemia or abnormal clotting factors. Participants with myocardial infarction, coronary artery bypass graft, percutaneous transluminal coronary angioplasty, or who have angina pectoris or coronary artery disease defined by angiography will be included.
Children (age <16 years), pregnant women, those with congenital heart disease or people who are acutely ill will be excluded. Acutely ill people includes those diagnosed with cancer, undergoing heart or renal transplant, with HIV or AIDS, on haemodialysis or with any renal problem except in diabetics).
Participants on medication will be included with the exception of those on weight loss medication who will be excluded.
Types of interventions
Studies will be included which compare the effect of individual wholegrain foods, or diets high in wholegrain foods, with other diets or foods with lower levels or no wholegrains. Comparisons should be between diets with similar overall carbohydrate, fat, protein and energy levels. For the purpose of this review the term wholegrain will include foods based on milled wholegrains, such as wholemeal or oatmeal, where the components of the endosperm, bran and germ are not removed. Studies must have a minimum of four weeks intervention period (or follow‐up period following dietary advice).
Studies will not be included if there are multiple component interventions, or interventions which incorporate factors other than wholegrain foods or diets, unless the effect of wholegrain foods or diets can be separated out from the other factors. Studies on foods which are based only on individual components (eg bran, germ or other components) of the grain will not be included e.g. oat bran or wheatgerm. Studies that examine the effect of high fibre, dietary fibre or cereal fibre, but where the specific effect of wholegrain foods or diets cannot be distinguished, will not be included.
Types of outcome measures
Primary outcomes
(1) Total CHD mortality;
(2) Combined CHD events and morbidity (to include fatal and non fatal myocardial infarction, angina, unplanned coronary artery bypass graft or percutaneous transluminal coronary angioplasty);
(3) Changes in major risk factors for CHD including overweight, lipids (HDL, LDL cholesterol levels, triglycerides and total cholesterol), blood pressure, measures of diabetic control including changes in medication, glycosylated haemoglobin, glucose tolerance and control), insulin resistance, insulin sensitivity, clotting factors, hyperinsulinaemia, hyperglycaemia.
Secondary outcomes
(1) Measures of quality of life and attitudes to diets
Where reported, harms such as bloating, nausea, weight gain, difficulty in eating out, difficulty in preparing meals, excessive weight loss, intention to continue diet, will be noted.
Search methods for identification of studies
(1) Electronic databases will be searched including the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (1966 to present), EMBASE (1980 to present), CINAHL (1982 to present) and ProQuest Digital Dissertations. The searches will be limited to studies in humans. There are no language restrictions for either searching or trial inclusion;
(2) The reference lists of review articles and of all included studies will be searched in order to find other potentially eligible studies;
(3) Known experts in the field will be consulted to ensure completeness.
The following search strategy will be used to search MEDLINE:
1. wholegrain$.ab,ti.
2. wholemeal$.ab,ti.
3. wholewheat$.ab,ti.
4. (whole adj3 grain$).ab,ti.
5. (whole adj3 meal).ab,ti.
6. (whole adj3 wheat).ab,ti.
7. (whole adj3 food).ab,ti.
8. (wheat adj3 meal).ab,ti.
9. cereal$.ab,ti.
10. bread$.ab,ti.
11. wheat$.ab,ti.
12. oat$.ab,ti.
13. rye$.ab,ti.
14. barley$.ab,ti.
15. maize$.ab,ti.
16. corn.ab,ti.
17. cornmeal.ab,ti.
18. popcorn.ab,ti.
19. sorghum$.ab,ti.
20. (bulgar or bulghar).ab,ti.
21. couscous$.ab,ti.
22. grain$.ab,ti.
23. porridge.ab,ti.
24. rice$.ab,ti
25. millet$.ab,ti
26. exp CEREALS/
27. exp BREAD/
28. exp Dietary Fiber/
29. exp Coronary Disease/
30. exp Cardiovascular Diseases/
31. heart disease$.tw.
32. coronary disease$.tw.
33. chd.tw.
34. cardiovascular.tw.
35. angina.tw.
36. cvd.tw.
37. exp CHOLESTEROL/
38. exp Blood Pressure/
39. exp Obesity/
40. exp Insulin Resistance/
41. exp Diabetes Mellitus/
42. insulin resistance.ab,ti.
43. insulin sensitivity.ab,ti.
44. (glyc?emic adj3 control).ab,ti.
45. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26
46. 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42
47. 43 and 44
48. randomized controlled trial.pt.
49. controlled clinical trial.pt.
50. exp Randomized Controlled Trials/
51. random allocation.sh.
52. double blind method.sh.
53. single‐blind method.sh.
54. 46 or 47 or 48 or 49 or 50 or 51
55. (animal not human).sh.
56. 52 not 53
57. clinical trial.pt.
58. exp Clinical Trials/
59. (clin$ adj25 trial$).ab,ti.
60. ((singl$ or doubl$ or trebl$ or tripl$) adj (blind$ or mask$)).ab,ti.
61. placebos.sh.
62. placebo$.ab,ti.
63. random$.ab,ti.
64. research design.sh.
65. 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62
66. 63 not 53
67. 64 not 54
68. comparative study.sh.
69. exp Evaluation Studies/
70. follow up studies.sh.
71. prospective studies.sh.
72. (control$ or prospectiv$ or volunteer$).ab,ti.
73. 66 or 67 or 68 or 69 or 70
74. 71 not 53
75. 72 not (54 or 65)
76. 54 or 65 or 73
77. 45 and 74
Unless otherwise stated, search terms are free text terms: MeSH: Medical subject heading (MEDLINE medical index term); a dollar sign stands for 'any character(s)' .
The above MEDLINE search will be adapted for CINAHL, EMBASE and the Cochrane Central Register of Controlled Trials
Published systematic reviews in the area of wholegrain foods will also be sought as a source of RCT's.
Data collection and analysis
Trials selection
One reviewer will scan the titles and abstracts of each record retrieved from the search. Articles will only be rejected on initial screen if the reviewer can determine from the title and abstract that the article definitely does not meet the reviews inclusion criteria.
When a title/abstract cannot be rejected with certainty, the full text of the article will be obtained for further evaluation and the entire article assessed by two reviewers independently. An in/out form will be used to assess studies inclusion (or otherwise) into the review. If a trial is excluded at any time after this point, a record of both the article and exclusion reason will be kept.
Data extraction
Original reports of trial results will be abstracted by two reviewers independently. Any differences between reviewers' abstraction results will be resolved by discussion and, when necessary, in consultation with a third reviewer.
Data which will be abstracted will include the following:
1. General information: Published/unpublished, title, authors, source, country, year of publication, duplicate publications.
2. Trial characteristics: Design, duration, randomisation (and method), allocation concealment (and method), blinding (outcome assessors), check of blinding.
3. Intervention: Dietary information/foods provided, length of intervention, comparison interventions.
4. Participants: Sampling (random/convenience), exclusion criteria, total number and number in comparison groups, gender/age, diagnosis of CHD or risk factors, similarity of groups at baseline, withdrawals/losses to follow‐up, assessment of compliance, medications used, smoking status where provided
5. Outcomes: Outcomes as specified above including what was the main outcome assessed in the study, other events, length of follow‐up
6. Results: For outcomes and times of assessment
Quality assessment of trials
The quality of each trial will be assessed based mainly on the quality criteria specified in the Cochrane Reviewers Handbook (Section 4.1.6). In particular the following factors will be examined:
1. Method of randomisation ‐ each factor will be marked as 'done', 'not done' or 'unclear'.
Can the study be described as randomised (including use of words such as "random", "randomly" and "randomisation")?
Does the study describe the method of randomisation and was it an appropriate method? A method to determine the sequence of randomisation will be regarded as appropriate if it allows for each study participant to have the same chance of receiving each intervention and for the investigators to be unable to predict which treatment will be next. Methods of allocation using date of birth, date of admission, hospital numbers, or alternation will not be regarded as appropriate.
2. Concealment of allocation ‐ this will be scored A (adequate), B (unclear), C (inadequate), following criteria adopted from the Cochrane Reviewer's Handbook (Section 4.1.6) and Schulz et al (Schulz 1995).
A ‐ Adequate measures to conceal allocations such as central randomisation; serially numbered, opaque, sealed envelopes; or other descriptions with convincing concealment.
B ‐ Unclearly concealed trials, in which the authors either did not report allocation concealment at all, or reported an approach that did not fall into one of the categories in (A).
C ‐ Inadequately concealed trials, in which the method of allocation was not concealed, such as alteration methods or use of case record numbers.
3. Blinding ‐ each factor will be marked as 'done', 'not done' or 'unclear'
With lifestyle interventions, such as the topic for this review, it is difficult to blind participants and those providing dietary advice. However, it is possible to blind outcome assessors and this will be marked as 'done', 'not done' or 'unclear'.
4. Intention‐to‐treat analysis ‐ will be marked as 'done', 'not done' or 'unclear'
Whether an intention‐to‐treat analysis is possible on all patients from the published data (i.e. whether there were any exclusions from the trial after randomisation) and the number of patients who were lost to follow‐up. If there were no withdrawals this should be stated in the article. An intention‐to‐ treat analysis will be considered adequate if outcome data is analysed for all participants randomised.
Based on these criteria, studies will be broadly subdivided into the following three categories (see Cochrane Handbook, section 6.7.1):
A ‐ all quality criteria met: low risk of bias.
B ‐ one or more of the quality criteria only partly met: moderate risk of bias.
C ‐ one or more criteria not met: high risk of bias.
This classification will be used as the basis of a sensitivity analysis. Additionally, we will explore the influence of individual quality criteria in a sensitivity analysis.
Trials will be independently assessed by two reviewers. The level of inter‐rater agreement will be calculated using the kappa‐statistic. In cases of disagreement, the rest of the group will be consulted and a judgement will be made based on consensus. Studies will not be excluded on the basis of a low quality score.
Data analysis
Where available data is sufficiently similar and of sufficient quality, statistical analyses will be performed using the RevMan software. Heterogeneity between trial results will be tested for using a standard chi‐squared test. Tests of heterogeneity are used for examining whether the observed variation in study results is compatible with variation expected with chance alone. If heterogeneity is found, then meta‐analysis will not be carried out.
Primary measures of interest will be the effect of wholegrain foods or diets on 1. total mortality 2. cardiovascular events and morbidity 3. Risk factors for CHD. Where appropriate, post treatment values / end of study values in the outcomes will be combined across studies using relative risks or standardised differences in the means in fixed effects meta‐analysis. If trials are found where participants are randomised by cluster/groups the patient numbers will be reduced to an effective sample size as described by Hauck (Hauck 1991).
Sensitivity analysis will be used to take into account the influence of various factors e.g. (a) exclusion from the analysis of any unpublished studies, (b) study quality, (c) exclusion of particularly long‐term or large trials on effect size. Funnel plots will be used to assess for evidence of bias (Egger 1997).
Subgroup analyses will be performed, where feasible, to assess whether particular groups of subjects could obtain more benefit from intervention than other groups e.g. diabetics, those with diagnosed heart disease, those with risk factors for heart disease.
