|Home | About | Journals | Submit | Contact Us | Français|
Diet is an important cause of many chronic diseases.
Intensive advice to healthy people to reduce sodium intake reduces sodium intake, as measured by sodium excretion.
Reducing sodium intake reduces blood pressure, even in people without hypertension.
Advice to reduce saturated fat intake may reduce the saturated fat intake, and multiple advice components reduce saturated fat intake more.
Reducing saturated fat intake can reduce mortality in the longer term.
Complex combined interventions to lose weight (physical plus dietary plus behavioural) are effective in helping people lose weight. Simpler interventions are less effective.
Increasing fruit and vegetable intake may decrease the risk of cardiovascular disease.
Taking a high dose of antioxidant supplements (vitamin E and beta carotene) does not reduce mortality or cardiovascular events.
We don't know whether omega 3 oil supplementation or advice to increase omega 3 intake can reduce mortality.
We also don't know how effective a Mediterranean diet is at reducing cardiovascular events or deaths in the general population.
Diet is important in the cause of many chronic diseases. Individual change in behaviour has the potential to decrease the burden of chronic disease, particularly cardiovascular disease (CVD). This review focuses on the evidence that specific interventions to improve diet and increase weight loss lead to changed behaviour, and that these changes may prevent CVD. Primary prevention in this context is the long-term management of people at increased risk but with no evidence of CVD. Clinically overt ischaemic vascular disease includes acute myocardial infarction, angina, stroke, and peripheral vascular disease. Many adults have no symptoms or obvious signs of vascular disease, even though they have atheroma and are at increased risk of ischaemic vascular events because of one or more risk factors. In this review, we have taken primary prevention to apply to people who have not had clinically overt CVD, or people at low risk of ischaemic cardiovascular events. Prevention of cerebrovascular events is discussed in detail elsewhere in BMJ Clinical Evidence (see review on stroke prevention).
CVD was responsible for 39% of UK deaths in 2002. Half of these were from coronary heart disease (CHD), and a quarter from stroke. CVD is also a major cause of death before 75 years of age, causing 34% of early deaths in men and 25% of deaths before 75 years of age in women. CHD deaths rose dramatically in the UK during the 20th century, peaked in the 1970s, and have fallen since then. Numbers of people living with CVD are not falling, and the British Heart Foundation estimates that there are about 1.5 million men and 1.2 million women who have or have had a myocardial infarction or angina. Worldwide, it is estimated that 17 million people die of CVDs every year, and more than 60% of the global burden of CHD is found in resource-poor countries (10% of disability-adjusted life years [DALYs] lost in low- and middle-income countries and 18% in high-income countries). The USA has a similar burden of heart disease to the UK; in 2002, 18% of deaths in the USA were from heart disease, compared with 20% in the UK. The USA lost 8 DALYs per 1000 population to heart disease and a further 4 DALYs per 1000 population to stroke, and the UK lost 7 DALYs per 1000 population to heart disease and 4 DALYs per 1000 population to stroke. Afghanistan has the highest rate of DALYs lost to heart disease (36 DALYs per 1000 population), and France, Andorra, Monaco, Japan, Korea, Dominica, and Kiribati have the lowest (1-3 DALYs per 1000 population). Mongolia has the highest rate for stroke (25 DALYs per 1000 population lost) and Switzerland the lowest (2 DALYs per 1000 population lost).
Deaths from CHD are not evenly distributed across the population. They are more common in men than in women; 67% more common in men from Scotland and the north of England than the south of England; 58% more common in male manual workers; twice as common in female manual workers than female non-manual workers; and about 50% higher in South Asian people living in the UK than in the average UK population. In the UK there are 14% more CHD deaths in the winter months than in the rest of the year. CVD in the UK generally results from the slow build-up of atherosclerosis over many decades, with or without thrombosis. The long development time of atherosclerosis means that small changes in lifestyle may have profound effects on risk of CVD over decades. However, while there is strong evidence from epidemiological studies of the importance of lifestyle factors — such as smoking, physical activity, and diet — in the process of development of CVD, adjusting for confounding can be difficult, and the long timescales involved make proving the effectiveness of preventive interventions in trials difficult. In practice, risk factors — rather than disease outcomes — are often the only practical outcomes for intervention studies in low-risk people. Such risk factors include blood pressure, body mass index, serum lipids, and development of diabetes.
Improvements in diet and reduction in weight may lower the risk of cardiovascular disease by exerting favourable changes on CVD risk factors (obesity, high blood pressure, elevated serum lipids, diabetes).
To change lifestyle factors, cardiovascular risk factors, and risk of CVD and death in the general population (adults without existing serious risk factors such as hypertension, dyslipidaemia, or diabetes), with minimum adverse effects.
Lifestyle change; changes in risk factors such as serum lipids, weight, blood pressure, and glucose tolerance; cardiovascular events such as myocardial infarction, angina, stroke, and heart failure; deaths (total or from CVD); adverse effects.
BMJ Clinical Evidence search and appraisal August 2006. The following databases were used to identify studies for this review: Medline 1966 to August 2006, Embase 1980 to August 2006, and The Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Clinical Trials 2006, Issue 3. Additional searches were carried out using these websites: NHS Centre for Reviews and Dissemination (CRD) — for Database of Abstracts of Reviews of Effects (DARE) and Health Technology Assessment (HTA), Turning Research into Practice (TRIP), and National Institute for Health and Clinical Excellence (NICE). Abstracts of the studies retrieved were assessed independently by two information specialists using pre-determined criteria to identify relevant studies. Study design criteria for inclusion in this review were: published systematic reviews and RCTs in any language containing more than 20 individuals of whom more than 90% were followed up for a minimum of 12 months. We included open studies. In addition, we use a regular surveillance protocol to capture harms alerts from organisations such as the US Food and Drug Administration (FDA) and the UK Medicines and Healthcare products Regulatory Agency (MHRA), which are added to the reviews as required.
The information contained in this publication is intended for medical professionals. Categories presented in Clinical Evidence indicate a judgement about the strength of the evidence available to our contributors prior to publication and the relevant importance of benefit and harms. We rely on our contributors to confirm the accuracy of the information presented and to adhere to describe accepted practices. Readers should be aware that professionals in the field may have different opinions. Because of this and regular advances in medical research we strongly recommend that readers' independently verify specified treatments and drugs including manufacturers' guidance. Also, the categories do not indicate whether a particular treatment is generally appropriate or whether it is suitable for a particular individual. Ultimately it is the readers' responsibility to make their own professional judgements, so to appropriately advise and treat their patients.To the fullest extent permitted by law, BMJ Publishing Group Limited and its editors are not responsible for any losses, injury or damage caused to any person or property (including under contract, by negligence, products liability or otherwise) whether they be direct or indirect, special, incidental or consequential, resulting from the application of the information in this publication.
Primary prevention of CVD: physical activity
Two systematic reviews found that reducing sodium intake in people without hypertension reduced blood pressure compared with usual diet. One subsequent RCT of very tight control of dietary sodium in people with hypertension found that reducing sodium intake (as part of a low-fat and high-fruit-and-vegetable diet) reduced blood pressure.
The first systematic review (search date not reported, 3 RCTs, 2326 adults without hypertension) found that dietary advice to people without hypertension to reduce their sodium intake compared with continuing on their usual diet significantly reduced mean systolic and diastolic blood pressure at 6–12 months (mean fall in systolic blood pressure: 2.3 mm Hg, 95% CI 1.6 mm Hg to 3.1 mm Hg; mean fall in diastolic blood pressure: 1.2 mm Hg, 95% CI 0.6 mm Hg to 1.8 mm Hg). At 13–60 months, effects of low-sodium dietary advice compared with usual diet were smaller (mean fall in systolic blood pressure: 1.1 mm Hg, 95% CI 0.3 mm Hg to 1.9 mm Hg; mean fall in diastolic blood pressure: 0.5 mm Hg, 95% CI 0 mm Hg to 1.1 mm Hg). The advice was intensive, more than is usual in primary care. The second systematic review (search date 1996, 29 RCTs, 2 of which were in the first review, number of people not reported) compared less-intensive advice to reduce sodium in people with normal blood pressure at baseline versus no restriction on the length of follow up required. The review did not perform a meta-analysis, but it reported that a reduction in sodium intake of 100 mmol daily (a reduction of about 60%) would result in reductions of about 1.0 mm Hg systolic and 0.1 mm Hg diastolic blood pressure. We found one subsequent RCT (412 adults from the USA with systolic blood pressure of 120–159 mm Hg and diastolic blood pressure of 80–95 mm Hg at baseline, 41% had hypertension [140–159 mm Hg systolic or 90–95 mm Hg average in 3 screening visits], mean age 48 years, 56% women, 56% black, 40% non-Hispanic white). This RCT is reported here despite its short length of follow up and high number of people with hypertension as it is unique in providing very tight control of dietary sodium and other constituents by providing all food consumed for three 1-month periods, and therefore gives good estimates of actual blood pressure response to specific changes in sodium intake. The RCT compared a diet rich in fruit, vegetables, and low-fat dairy foods and low in saturated and trans fats (the DASH diet [Dietary Approaches to Stop Hypertension]) versus a typical US diet. Within each diet, both groups had 1 month periods of either high sodium (142 mmol sodium/day, which is close to the UK daily mean of 187 mmol in men and 139 mmol in women), intermediate (106 mmol/day), or low (65 mmol/day) sodium. For people without hypertension, those on the US diet had significant decreases in systolic and diastolic blood pressure when moving from high to intermediate and from intermediate to low sodium intakes and those on the DASH diet had smaller decreases in systolic and diastolic blood pressure when moving from high to intermediate or from intermediate to low sodium intakes (see table 1 ). Additionally, being on a DASH diet resulted in consistently lower blood pressure than being on the standard US diet; a dietary change to a diet high in fruit, vegetables, and low-fat dairy foods, low in saturated fat resulted in almost the same reduction in blood pressure as major salt reduction, but was likely to confer additional health benefits (e.g. on lipid levels). People on the low-sodium DASH diet had the lowest systolic blood pressure of all.
One systematic review found that intensive advice to reduce sodium intake decreased sodium intake (measured by sodium excretion) compared with usual diet.
We found one systematic review (search date 2000, 3 RCTs, 2326 adults without hypertension). The review found that advice to reduce sodium intake significantly reduced excreted sodium compared with usual diet at 13–60 months (WMD –34 mmol sodium/24 hours, 95% CI –19 mmol/24 hours to –50 mmol/24 hours). These RCTs were in healthy people, predominantly white, male, and from the USA, mean age 40 years, with blood pressure just below the definition of hypertension (mean diastolic blood pressure: 83–86 mm Hg; systolic blood pressure: 124–127 mm Hg). The advice in these three RCTs was intensive, and involved regular group and individual counselling for extended periods, as well as additional support materials such as recipes, lists of sodium contents of common food brands, shopping expeditions, etc.
Trials of dietary advice to reduce sodium in adults without hypertension found that those in the low-sodium arm had difficulties with their diets — including the diets being inconvenient, conflicting with schedules, needing extra time to plan, and being difficult to stick to when eating out. However, in one trial, psychological wellbeing scores increased in those on the low-sodium diet compared with those on their usual diets.
This type of intensive intervention is not realistic in primary care, but we found no systematic review or RCTs assessing the effect of public health interventions such as reducing the sodium content of processed foods, although this would seem a sensible way of making it easier for the whole population to consume less sodium. Under highly controlled conditions, halving dietary sodium has similar effects in reducing blood pressure as a diet high in fruit, vegetables, low-fat dairy foods, wholegrain products, nuts, fish, and poultry, and low in saturated fat and sugars, compared with a usual US diet — and the fruit-and-vegetable diet would be expected to have additional beneficial effects on lipids and other risk markers. There is evidence that a reduction of 20 mm Hg in systolic blood pressure is associated with halving of the risk of vascular mortality, so a reduction of 1 mm Hg systolic resulting from intensive dietary interventions to reduce sodium intake in the studies above would be associated with a reduction in risk of about 2.5%. This type of reduction is clinically important in a population but not in individuals. As the intensive intervention provided in the trials is relevant to individuals, but not to populations, the results are clinically disappointing. The more Mediterranean-pattern diet is likely to have similar-sized effects through its blood pressure-lowering effects, as well as additional benefits in promotion of healthy lipids, body weight, prevention of osteoporosis, provision of additional fruit and vegetables, etc, and so may have a greater effect on prevention of cardiovascular disease and other chronic diseases.
One systematic review found no significant difference in mortality between reducing dietary saturated fat intake compared with usual care at less than 2 years; however, it did find that reducing dietary saturated fat intake decreased mortality at 2 years or longer. A second systematic review found no significant difference in mortality between multiple interventions including reducing dietary saturated fat intake compared with usual care; however, the people in the RCTs were young, and follow-up was sometimes short.
We found two systematic reviews. The first systematic review (search date 1999, 27 RCTs, 18,196 adults with and without cardiovascular disease [CVD], 14 RCTs of people with low CVD risk, 6 RCTs of moderate risk, and 7 RCTs of high risk) compared low saturated fat advice or dietary support versus usual diet for at least 6 months. The review found no significant difference in total mortality or cardiovascular mortality between advice to reduce dietary fat compared with usual diet (total mortality: RR 0.98, 95% CI 0.86 to 1.12; cardiovascular mortality: RR 0.84, 95% CI 0.77 to 1.07). However, the review found that there were significantly fewer cardiovascular events with advice to reduce dietary fat (RR 0.84, 95% CI 0.72 to 0.99). The review found no significant difference in cardiovascular events at 2 years or less between low saturated fat advice or dietary support compared with usual diet, but found that cardiovascular events were significantly fewer at 2 years or more with advice to reduce fat (2 years or less: RR 0.96, 95% CI 0.75 to 1.23; 2 years or more: RR 0.76, 95% CI 0.65 to 0.90). Subgroup analysis by level of cardiovascular risk (low risk: general population; medium risk: those with risk factors; high risk: those with existing CVD) and type of support (dietary advice, dietary advice plus supplements, or diet provided) seemed to make little difference to the effect size, although those at high risk showed a significant effect as there were more events in this group (cardiovascular events for those at low cardiovascular risk: RR 0.82, 95% CI 0.56 to 1.20; high cardiovascular risk: RR 0.84, 95% CI 0.70 to 0.99). The second systematic review (search date 1995, 14 RCTs with 6 months' follow-up or longer, 1,206,000 person-years of observation in RCTs with risk-factor or clinical-event outcomes, and excluding those in secondary prevention, or in children or adults under 40 years of age) compared multiple-risk-factor interventions versus no intervention. The review found no significant difference in reduction in total or coronary heart disease mortality between multiple-risk-factor interventions and no intervention (total mortality: OR 0.97, 95% CI 0.92 to 1.02; coronary heart disease mortality: OR 0.96, 95% CI 0.88 to 1.04) over 903,000 person-years of observation, but the people were generally young (mean age 45.2–54.1 years) and follow-up was sometimes of short duration (mean follow up 1.5–11.8 years).
One systematic review found that multiple advice components to reduce saturated fat intake were associated with more reduction in saturated fat intake compared with simple dietary advice. A second systematic review found inconclusive evidence on the effects of advice to reduce saturated fat intake compared with no advice. Four subsequent RCTs found that advice to reduce saturated fat intake was associated with reduced saturated fat intake.
We found two systematic reviews and four subsequent RCTs. The first systematic review (search date 2001, 17 RCTs, more than 27,795 adults with and without risk factors for cardiovascular disease [CVD]) compared dietary advice to reduce total or saturated fat intake for at least 3 months versus usual diet. Of the 17 RCTs assessing the effect of dietary counselling on saturated fat intake, six RCTs found a self-reported large effect size in people's reduction in saturated fat intake (as either grams of saturated fat or percentage of calories as saturated fat) (a reduction of more than 3%), five RCTs found a medium effect (1.3–3.0% reduction), and six RCTs found a small effect (less than 1.3% reduction). Effect sizes were calculated using (in order of preference): net difference in change, difference at final follow-up, or relative change. In this analysis, RCTs that used more intensive interventions achieved greater effect sizes. Those in primary care settings tended to produce small- or medium-sized effects, mainly because they used lower-intensity interventions than those in special research clinics. Using more components (dietary assessment, family involvement, social support, group counselling, food interaction such as taste testing and cooking, goal setting, and advice appropriate to the patient group) was associated with greater effect sizes. The authors of the review reported that baseline level of CVD risk did not seem associated with effect size after stratification for intensity of intervention. The second systematic review (search date 1993) identified eight RCTs, 3 of which were also identified by the first review, assessing diets to reduce dietary fat in primary prevention of free-living adults over at least 3 months. It reported that, at 9–18 months, only four RCTs noted effects in reported dietary fat intake, two of which were trials of breast cancer prevention. The review did not perform a meta-analysis because of heterogeneity between RCTs. Two RCTs found a significant reduction in dietary fat intake with advice compared with no advice — however, these were the two trials of breast cancer prevention. RCTs included in the review that reported both dietary saturated fat intake and blood lipids suggested significant effects on dietary fat, but non-significant effects on lipids — suggesting that self-reported saturated fat intake may not be a good indicator of usual intake. The first subsequent RCT (235 healthy people aged 30–59 years with siblings who were diagnosed with coronary heart disease before the age of 60 years) found that nurse counselling significantly reduced saturated fat intake at 2 years compared with usual care (–4.9 g/day with counselling v +1.9 g/day with usual care; P = 0.0002). The second subsequent RCT (545 premenopausal women, mean age 47 years) found that a 5-year cognitive behavioural programme significantly reduced saturated fat as a percentage of total calories compared with no programme (9% with programme v 11% with no programme; P less than 0.05). The third subsequent RCT (616 healthy women aged 40–70 years) found a significant reduction in total fat as a percentage of energy in the intervention group compared with control (advice for breast self-examination) at 12 months (35% with counselling v 39% with control; P less than 0.001). The fourth subsequent RCT (143 people with an “elevated” risk for CVD, mean age 58 years) also found a significant reduction in energy from saturated fat 12 months after nutritional counselling by family physicians compared with no advice (–3% with counselling v –1% with no advice; P less than 0.05).
Once people in a trial have been told to reduce total or saturated fat, they have a tendency to underreport their intake to a greater extent than control groups. There is often inconsistency in effectiveness when trials report both changes in fat intake (which is self reported in a variety of ways) and serum lipids (which are likely to provide a better view of saturated fat intake over the past few weeks). Trials often report significant reductions in dietary fats without significant changes in total or low-density lipoprotein cholesterol.
Systematic reviews and RCTs found that combinations of interventions (physical, dietary, and behavioural) reduced body weight compared with usual care. Systematic reviews and RCTs of less complex interventions found that they were less effective at reducing body weight.
We found two systematic reviews and one subsequent RCT. The first systematic review (search date 2001, 26 RCTs, 3048 people with body mass index [BMI] of 28 kg/m2 or greater) included RCTs that compared weight-reducing diets versus usual diets, where weight was measured after at least 12 months. Thirteen RCTs of low-fat diets compared with usual diet found that low-fat diets resulted in greater weight loss at 12 months (WMD –5.3 kg, 95% CI –5.9 kg to –4.8 kg), 24 months (WMD –2.4 kg, 95% CI –3.6 kg to –1.2 kg), and 36 months (WMD –3.6 kg, 95% CI –4.5 kg to –2.6 kg), with no significant difference at 60 months (WMD –0.2 kg, 95% CI –2.0 kg to +1.6 kg). Comparisons of low-calorie or very-low-calorie diets versus usual diet were only found in subgroups of people with chronic illness. RCTs comparing low-calorie diets versus low-fat diets, very-low-calorie diets versus low-fat diets, and very-low-calorie diets versus low-calorie diets were found, but studies were small and no significant differences were found at 12 months or later. The second systematic review (search date 2003, 29 RCTs) included systematic reviews and RCTs of screening for obesity, weight-reducing, and weight-maintaining interventions. The review found that weight loss trials most likely to succeed were high intensity (where contact was more frequent than monthly). The subsequent RCT (160 people) compared four weight-loss diets (Atkins, Zone, Weight Watchers, and Ornish diets) and found no significant differences in weight loss at 1 year or study completion (at 1 year: –2.1 kg with Atkins v –3.2 kg with Zone v –3.0 kg with Weight Watchers v –3.3 kg with Ornish; P = 0.4 between groups; at study completion: –52% with Atkins v –65% with Zone v –65% with Weight Watchers v –50% with Ornish; P = 0.08 between groups).
We found one systematic review (search date 2003, 36 RCTs, 3495 overweight or obese people, BMI greater than 25 kg/m2 ) of psychological interventions for sustained weight loss in overweight or obese adults. Ten of these RCTs compared a behavioural intervention versus no treatment, and two reported outcomes later than 12 months, finding that those in the behavioural-intervention arm lost more weight on average than those in the control arm (–2.0 kg, 95% CI –2.7 kg to –1.3 kg). Comparing more-intensive versus less-intensive behavioural interventions (10 RCTs, 306 people) found significantly greater weight loss in those on more-intensive interventions in studies of up to 1 year (–2.3 kg, 95% CI –3.3 kg to –1.4 kg). One RCT assessed outcomes longer than 1 year and found no significant difference in weight loss. Cognitive behavioural therapy (CBT) added to diet plus exercise versus diet plus exercise alone (2 RCTs, 63 people) showed significantly more weight loss with CBT (–4.9 kg, 95% CI –7.3 kg to –2.4 kg). Direct comparison of CBT compared with behavioural interventions found that CBT was associated with significantly more weight loss (1 RCT, 24 people, difference of 5.7 kg; P less than 0.01). We found a second systematic review asking whether brief interventions using motivational interviewing were effective across several behavioural domains including diet and exercise (search date 1999, 29 RCTs of which 5 were of diet plus exercise), and which included RCTs that compared motivational interviewing versus a control group. Using unit-free effect sizes, they suggest that three of the five trials showed statistically significant benefit. One additional RCT of 92 overweight adults (mean BMI 33.1 kg/m2 ) compared behavioural counselling plus internet-based weight-loss programme versus internet-based weight-loss programme alone, and found greater weight loss in the behavioural counselling group at 12 months (–4.4 kg with behavioural counselling plus internet-based weight-loss programme v –2.0 kg with internet-based weight-loss programme alone; P = 0.04).
We found one systematic review (search date 2003), which suggested that successful interventions were more likely to include two or three strategies (combining behavioural, dietary, and physical activity intervention). One large RCT identified by the review (1191 men and women) comparing weight loss using dietary, physical, and social support versus usual care found that combined interventions reduced weight by 2.7 kg at 18 months and 2.0 kg at 36 months (no further data reported). We also found three subsequent RCTs. The first subsequent RCT (423 overweight and obese adults) identified by the review compared the effect of participation in the Weight Watchers commercial programme of group-based dietary, physical activity, and behavioural change with two 20-minute nutritionist interviews and self-help materials. It found that those randomised to the commercial programme had lost more weight at 2 years (–2.7 kg; P less than 0.001), and had greater decreases in waist circumference and BMI. The second subsequent RCT (44 obese sedentary postmenopausal women) added self-control skills training to a comprehensive programme of lifestyle, exercise, attitudes, relationships, and nutrition. The RCT found that self-control skills plus comprehensive programme significantly reduced weight compared with comprehensive programme (–6.5%; no further data reported). The third subsequent RCT (110 obese men with erectile dysfunction) found that detailed advice on diet and physical activity (compared with general advice to the control group) resulted in significant reductions in BMI in the intervention compared with the control group, after 2 years (no further data reported).
One systematic review compared weight-reducing diets plus exercise versus weight-reducing diets alone in adults with obesity (search date 2001). The review found that adding exercise to diet significantly increased mean weight loss (at 12 months: –1.95 kg, 95% CI –3.22 kg to –0.68 kg; at 36 months: –8.22 kg, 95% CI –15.27 kg to –1.16 kg).
One systematic review found that adding behavioural therapy to weight-reducing diet significantly improved weight loss at 12 months, but found no significant difference at later time points (4 RCTs; at 12 months: –7.7 kg, 95% CI –12.0 kg to –3.4 kg; at 36 months: +2.9 kg, 95% CI –2.8 kg to +8.6 kg; at 60 months: –1.9 kg, 95% CI –7.6 kg to +3.8 kg). However, in one cluster RCT, addition of behavioural therapy to weight-reducing diets plus exercise did not improve weight loss at 12 months compared with weight-reducing diets plus exercise, and addition of exercise and behavioural therapy to diet did not significantly improve weight loss at 12 months. A second systematic review found that adding behavioural therapy to diet plus exercise (compared with diet plus exercise alone) produced a heterogeneous result, but that five of the six studies favoured the addition of behavioural therapy (the sixth favoured diet plus exercise alone).
Adverse effects were addressed in few reviews or trials. One review found that adverse effects (such as illnesses and deaths) occurred in both arms, but, as such events were rare, patterns were not clear. One subsequent RCT reported that it did not identify any serious adverse effects of the diets.
The main methodological issue in weight-reducing trials is the high level of withdrawals, particularly in the longer trials, which can provide more useful information. People who completed trials tended to do better, and differential withdrawal rates can misrepresent the effectiveness of interventions. Potential adverse effects that should be assessed in trials and reviews include eating disorders such as anorexia or bulimia (which might be precipitated by a concentration on body image, food, and exercise) and rebound binge eating (which might lead to increased weight, but would tend to occur in those who have withdrawn from the study).
One systematic review found that weight-maintenance strategies can maintain weight loss; however, the review reported few details of how the conclusion was reached. RCTs found inconclusive evidence on the effects of different lifestyle interventions (physical, dietary, and behavioural) to maintain weight loss.
We found one systematic review (search date 2003, 29 RCTs, included systematic reviews and RCTs of screening for obesity, weight-reducing and weight-maintaining interventions). It found that maintenance strategies helped to maintain weight loss (mentioning 2 trials) but provided few details of how this conclusion was reached. We found six RCTs of weight-maintenance programmes of at least 1 year after a weight-loss programme. The first RCT (82 obese men, mean body mass index [BMI] 32.9 kg/m2, mean weight 106.0 kg, who had a very-low-energy diet for 2 months, mean body weight after the diet 91.7 kg) found no significant difference in mean body weight at 31 months between walking (45 minutes 3 times weekly for 6 months), resistance training (45 minutes 3 times weekly for 6 months), and no exercise (102.0 kg with walking v 99.9 kg with resistance training v 100.7 kg with no exercise; P = 0.8 between groups). The second RCT (82 obese premenopausal women, mean BMI 34.0 kg/m2, mean weight 92.0 kg, who had a very-low-energy diet for 3 months, mean weight loss after the diet 13.1 kg) found no significant difference in mean body weight between diet counselling plus a walking programme (to expend 4.2 MJ/week), diet counselling plus a more intensive walking programme (to expend 8.4 MJ/week), and diet counselling alone (83.9 kg with lower walking programme v 87.4 kg with higher walking programme v 89.7 kg with no walking; P = 0.07 between groups). The third RCT (80 obese women whose mean weight loss was 8.74 kg in an initial 20-week behavioural treatment programme, who then had either: problem-solving therapy, relapse prevention training, or no further treatment for 12 months) found that women who had subsequent problem-solving therapy lost significantly more body weight than women who had no further treatment (10.8 kg with problem-solving therapy v 4.1 kg with no treatment; P = 0.019). The RCT found no significant difference between any other comparison (P values not reported). The fourth RCT (128 African-American people, mean BMI 37.5 kg/m2 who had a 10-week healthy eating and lifestyle programme, mean weight loss 1.5 kg, mean BMI after the programme 37.0 kg/m2 ) found no significant difference at 8–20 months in change of body weight from baseline between people who had either further group counselling, staff-facilitated self help, or clinic visits only (–0.8 kg with group counselling v –1.3 kg with self help v –1.4 kg with clinic visits; P = 0.90 between treatments). The fifth RCT (255 people, mean BMI 31.8 kg/m2, mean body weight 89.4 kg, who had a 6-month weight control programme by interactive television, mean weight loss 7.8 kg) found no significant difference in overall weight loss at the end of a 12-month maintenance period between three maintenance strategies: frequent therapist contact, minimum therapist contact, and internet therapist support (5.1 kg with frequent therapist contact v 5.5 kg with minimum therapist contact v 7.6 kg with internet therapist support; P = 0.23). The sixth RCT (67 people who had a 6-month weight-loss programme, mean weight loss 8.8 kg, mean BMI after the programme 30.8 kg/m2, mean body weight 85.2 kg, mean weight and BMI before programme not reported) found that people completing a weight-focused maintenance programme gained significantly less weight pver 12 months than people on an exercise-focused maintenance programme (+0.8 kg with weight-focused maintenance programme v +4.4 kg with exercise-focused maintenance programme; P less than 0.01).
One systematic review and two RCTs found inconclusive evidence on the effects of lifestyle interventions to prevent weight gain.
We found one systematic review, one additional RCT, and one subsequent RCT. The systematic review (search date not reported, 5 studies [4 controlled and 1 uncontrolled], number of people not reported) assessed community interventions promoting physical activity for preventing weight gain. It found no significant effects of the intervention on weight gain. The additional RCT compared education (by newsletters) versus education plus incentives versus no education to prevent weight gain. It found no significant difference in a between-group comparison in weight gain at 3 years (+1.6 kg with education v +1.5 kg with education plus incentives v +1.8 kg with no education; P = 0.80). The subsequent RCT compared a dietary and physical advice programme with no programme (assessment only) (535 premenopausal women, aged 44–50 years). It found that weight gain was significantly greater with the advice programme than with no programme (–0.1 kg with advice v +2.4 kg with no advice; reported as significant; P value not reported). It also found that body mass index increase was significantly less with the advice programme than with no programme (+0.05 kg/m2 with advice v +0.96 kg/m2 with no advice; P less than 0.001).
The review and additional RCT gave no information on adverse effects. The subsequent RCT found no signs of increased stress, depressive symptoms, or eating restraint among people in the intervention group.
One systematic review and one subsequent RCT found inconclusive evidence on the effects of training health professionals for promoting weight loss.
We found one systematic review and two subsequent RCTs. The systematic review (search date 2000; 5 RCTs, 1 controlled before and after study, 2992 overweight or obese people) compared interventions to improve health professionals' management of weight versus no interventions. It did not perform a meta-analysis, as some of the studies were small and of low quality; however, the review authors suggested that reminder systems, brief training interventions, shared care, inpatient care, and dietician-led treatments may be effective in improving management of weight. The first subsequent RCT cluster compared obesity management training versus no training in 44 general practices. The RCT found no significant difference in weight change in obese people in the practices with training than with no training at 12 months (843 obese people; +1 kg, 95% CI –1.9 kg to +3.9 kg; P = 0.5). The second subsequent RCT compared dietician counselling, dietician counselling plus doctor counselling, and no counselling. It found that dietician counselling resulted in significantly more weight loss than no counselling (–6.6%, 95% CI –7.6% to –5.8%), and that dietician plus doctor counselling resulted in significantly more weight loss than no counselling (–7.3%, 95% CI –8.3% to –6.5%). There was no significant difference in weight loss between doctor plus dietician counselling and dietician only counselling (+0.7%, 95% CI –0.42% to +1.82%).
We found no systematic review or RCTs assessing the effects of diets and behavioural intervention in reducing cardiovascular risk in the general population.
We found no systematic review or RCTs.
We found no RCTs.
Two systematic reviews of cohort studies found that increasing fruit and vegetable intake decreased cardiovascular risk.
We found no systematic review of RCTs but found two systematic reviews of cohort studies. The first systematic review (search date 2002, 8 cohort studies assessing the effects of fruit and vegetable consumption on CVD, total number of people not reported) reported that seven of the eight cohort studies found that a higher intake of fruit and vegetables decreased CVD risk compared with low intake — suggesting protection with higher intakes (no further data reported; results presented graphically). The second systematic review (search date not reported, 11 cohort studies on the effect of fruit and vegetable intake on deaths from ischaemic heart disease) found that most markers of fruit and vegetable intake (including dietary carotenoids, vitamin C, fruit fibre, and vegetable fibre) decreased ischaemic heart disease. Total fruit intake and total vegetable intake also decreased ischaemic heart disease, but less so.
One earlier systematic review (search date 1995, ecological, case control and cohort studies assessing the relationship between fruit and vegetable intake and CVD; number of people not reported) found that six of 16 cohort studies suggested a statistically significant effect.
One systematic review found no significant difference in increase in fruit and vegetable intake between advice to increase fruit and vegetable intake and usual diet. One subsequent RCT found inconclusive evidence on counselling to increase fruit and vegetable intake. A second subsequent RCT found that computer-assisted intervention to increase fruit and vegetable intake in healthy women increased intake compared with a control of counselling on breast self-examination.
We found two systematic reviews and one subsequent RCT of dietary interventions to increase the quantity of fruit and vegetables eaten. The first systematic review (search date not reported, included 12 controlled trials [unclear if they were RCTs], number of people not reported) assessed the effect of behavioural interventions to increase fruit and vegetable intake compared with control or usual diet. It found no significant difference in the median “difference in deltas” of increase in quantity of fruit and vegetables eaten (+16.6, range –3.7 to +60.9). The second systematic review (search date 2001, 10 RCTs of counselling to increase fruit and vegetable intake compared with no counselling; 1 of the RCTs was included in the first systematic review; number of people not recorded) found that three RCTs produced small or no increases in fruit and vegetables (less than 0.3 servings/day), five RCTs found medium increases (0.3–0.8 servings/day), and two RCTs found large effects (1.4–3.2 servings/day; no further data reported). One subsequent RCT (616 healthy women) found that a brief computer-assisted intervention to reduce dietary fat and increase fruit and vegetable consumption compared with a control intervention (counselling on breast self-examination) resulted in significantly higher self-reported intake of fruit and vegetable servings with computer intervention at 12 months compared with control intervention (4.33 servings/day with computer intervention v 3.40 servings/day with control intervention; P less than 0.001).
Systematic reviews found that high-dose supplements of vitamin E and beta carotene did not reduce mortality or cardiovascular events.
We found four systematic reviews and one subsequent RCT on the effects of vitamin E on cardiovascular disease (CVD) and mortality. The first review (search date 2004, including 9 RCTs that tested vitamin E alone, and 10 that tested vitamin E in combination with other vitamins and minerals) compared vitamin E versus placebo or control in non-pregnant adults over at least 1 year. This review aimed to assess any dose–response relationship between vitamin E supplementation and total mortality. Some studies included those with, or at high risk of, CVD, but most did not. The review found that, overall, there was no effect of vitamin E on total mortality (RR 1.01, 95% CI 0.98 to 1.04). Dose–response analysis suggested that all-cause mortality increased with dose, and high-dose vitamin E (400 IU/day or more) significantly increased deaths (risk difference: 39/10,000 people, 95% CI 3/10,000 people to 74/10,000 people). However, RCTs using high doses tended to be small and to include people with chronic disease. The second review (search date 2002, 84 RCTs of vitamin E on CVD, risk factors, or death) found no evidence to recommend vitamin E alone or in combination for the primary prevention of CVD). The primary prevention RCTs were not pooled. One of two primary prevention RCTs showed a significant reduction in total mortality in a trial of vitamin E plus beta carotene plus selenium in rural China (mainly owing to reductions in cancer mortality). Neither (of two) primary prevention trials of vitamin E with other vitamins showed any effect of vitamin E on total, fatal, or non-fatal myocardial infarction. There is a suggestion of statistically (but not clinically) significant reductions in total and low-density lipoprotein cholesterol levels in those taking vitamin E in primary prevention studies, but units, confidence intervals, and P values were not reported. The third review (search date 2001, 8 RCTs, 104 512 people) analysed the effects of vitamin E, beta carotene, and ascorbic acid for primary prevention. The review found no significant in CVD at 3–12 years difference between vitamin E compared with no antioxidant supplement (4 RCTs, 48 346 people: OR 0.96, 95% CI 0.88 to 1.04). The fourth review (search date not reported, 7 RCTs, 106 615 people [including secondary prevention]) found no significant difference in cardiovascular events between high-dose vitamin E (30–800 mg/day) compared with no vitamin E (any event: OR 0.98, 95% CI 0.94 to 1.03; non-fatal myocardial infarction: OR 1.00, 95% CI 0.92 to 1.09; non-fatal stroke: OR 1.03, 95% CI 0.93 to 1.14; cardiovascular death: 1.00, 95% CI 0.94 to 1.05). One subsequent RCT (39,876 healthy women in the USA) compared vitamin E (600 IU on alternate days) versus placebo for 10 years. The RCT found no significant difference in myocardial infarction, stroke, or total mortality between vitamin E compared with placebo (myocardial infarction: RR 1.01, 95% CI 0.82 to 1.23; stroke: RR 0.98, 95% CI 0.82 to 1.17; total mortality: RR 1.04, 95% CI 0.93 to 1.16). However, the RCT did find significantly fewer cardiovascular deaths with vitamin E compared with placebo (RR 0.76, 95% CI 0.59 to 0.98).
One systematic review (search date 2001, 8 RCTs, 104,512 people) analysed the effects of vitamin E, beta carotene, and ascorbic acid for primary prevention. The review found no significant difference in CVD at 3–12 years between beta carotene and no antioxidant supplement (6 RCTs, 85 056 people: OR 1.02, 95% CI 0.96 to 1.08). A second systematic review (search date not reported, 8 RCTs, 138,113 people with and without CVD) found that beta carotene 15–50 mg daily significantly increased all-cause mortality at 1.4–12.0 years compared with placebo (OR 1.07, 95% CI 1.02 to 1.11). However, the data for both primary and secondary prevention RCTs were pooled, and no separate meta-analysis for primary prevention RCTs was reported.
One RCT compared a combination of antioxidants (ascorbic acid 120 mg plus vitamin E 30 mg plus beta carotene 6 mg plus selenium 100 ìg plus zinc 20 mg) versus placebo for 7.5 years. The RCT found no significant difference in CVD events or total mortality at 7.5 years (13,017 healthy middle-aged people in France; CVD: RR 0.97, 95% CI 0.77 to 1.20; total mortality: RR 0.77, 95% CI 0.57 to 1.00). However, subgroup analysis by sex found significantly less mortality in men, but found no significant difference in women (men: RR 0.63, 95% CI 0.42 to 0.93; women: RR 1.03, 95% CI 0.64 to 1.63).
One review reported that there were “serious adverse cardiovascular effects” with vitamin E in one RCT, and increased risk of intracerebral haemorrhage with beta carotene (data not reported). The review also reported that, in two RCTs, risk of lung cancer was increased with beta carotene in cigarette smokers. One subsequent RCT found significantly more epistaxis with vitamin E compared with placebo (RR 1.06, 95% CI 1.01 to 1.11). Four systematic reviews and one RCT did not report on adverse effects.
One systematic review found no significant difference between omega 3 supplementation or advice to increase omega 3 intake compared with placebo in mortality.
We found one systematic review (search date 2002, 48 RCTs, 36 913 people), which compared high-dose omega 3 supplementation or dietary advice for at least 6 months versus control or placebo. If found no significant difference in mortality between omega 3 supplements or advice compared with placebo or no advice (44 RCTs, 36,195 people; RR 0.87, 95% CI 0.73 to 1.03; including 1 RCT of people at high and low risk of cardiovascular disease [CVD], and including RCTs of fish- and plant-based omega 3 fatty acids). There was also no significant difference for the same comparison in people with a low risk of CVD (17 RCTs, 14,599 people; RR 1.07, 95% CI 0.70 to 1.64). There was no significant difference for the same comparisons in RCTs with a low risk of bias, but any risk of CVD (RR 0.98, 95% CI 0.70 to 1.36). There was no significant difference in mortality in subgroup analyses of RCTs of advice to eat more oily fish compared with no advice, or fish oil capsules compared with placebo (advice: RR 0.91, 95% CI 0.57 to 1.44; capsules: RR 0.90, 95% CI 0.76 to 1.07).
We found no evidence of increase in cancers with increased omega 3 intake, but estimates were imprecise and a clinically important effect could not be excluded. Trials reported fewer than 400 cancers (mainly cancer deaths) overall and suggested no effect of omega 3 fatty acids on cancer (10 RCTs, cancer death or diagnoses: 391/17,433 [2%]; RR 1.07, 95% CI 0.88 to 1.30). The review of cohort studies found no increase in cancers in those taking greater quantities of omega 3 fatty acids (10 cohort studies, number of people not reported; RR 1.02, 95% CI 0.87 to 1.19).
Lack of protective effects of high doses of omega 3 fatty acids in the general public does not mean that oily fish at more usual dietary levels are not healthy. General population advice in the UK is that people should eat two portions of fish a week, one of which should be oily. The UK Committee on Toxicity, which reviewed the effects of oily fish and their contaminants with the Scientific Advisory Committee on Nutrition, set safe upper limits for intake of oily fish on the basis of a wide review of the evidence.
We found insufficient evidence on the effects of Mediterranean diet on risk factors for cardiovascular disease, cardiovascular events, or deaths in the general population.
We found no systematic review or RCTs on the effects of Mediterranean diet on risk factors for cardiovascular disease, cardiovascular events, or deaths in the general population. We found one systematic review (search date not reported; 1 prospective cohort study, 22,034 “middle-aged and older adults” in Greece). It found that, with a 2/9 increment in Mediterranean diet score (not described), there was a 25% reduction in total mortality and 33% reduction in heart disease mortality (after adjustment for sex, smoking, education, body mass index, or physical activity) in people over 55 years old only (no further data reported).
The Mediterranean diet is a pattern of eating characterised by use of olive oil for cooking and little use of dairy fats. Other characteristics of this diet include increased intake of oily fish, fruit, vegetables, nuts, pulses (and occasionally low-fat dairy products and red wine), and a lower intake of processed foods, meat, meat products, and hard fats, compared with a typical European or US diet.