What is cardiovascular disease?
Cardiovascular disease (CVD) is the leading cause of death worldwide. Smoking, diabetes and being overweight are risk factors for CVD, which means that they increase your chances of developing CVD. CVD can often be prevented by a healthy lifestyle, such as keeping to a healthy weight or losing weight if you need to.
Following a diet
Some people choose to lose weight by following a diet; for example, by eating less fat, or by reducing the number of calories they eat. Intermittent fasting is a type of diet involving patterns of eating and fasting (not eating foods); it does not limit what foods you eat, but limits when you can eat them. Eating patterns in intermittent fasting include: fasting for one or two days each week; fasting every other day; or eating only during certain hours and fasting for at least 12 hours every day.
Why we did this Cochrane Review
Diets that involve intermittent fasting are becoming popular. We wanted to find out if intermittent fasting could reduce or prevent CVD.
What did we do?
We searched for studies that tested intermittent fasting against 'usual eating' (someone eats whatever foods they want whenever they like), or against 'energy restriction' diets (someone limits the number of calories they eat).
We wanted to find out whether intermittent fasting affected mortality, cardiovascular mortality, risk of stroke, heart attack or heart failure. We also looked at whether intermittent fasting affected body weight and blood sugar levels.
Search date: we included evidence published up to 12 December 2019.
What we found
We found 26 relevant studies; we then used the results from 18 of the studies to compare the different diets. The 18 studies included 1125 adults (aged over 18 years). Some people in the studies had risk factors for CVD and some people had no risk factors. Most studies were funded by universities and research centres; two studies were funded by companies that make diet foods.
The studies compared intermittent fasting against usual eating (in seven studies); energy restriction diets (eight studies); and usual eating and energy restriction diets (three studies). The studies lasted from four weeks to six months. Results were reported after three months (short-term), and between three and 12 months (medium-term).
We didn't find any data on mortality, cardiovascular mortality or risk of stroke, heart attack or heart failure.
We found that people may lose more weight by intermittent fasting than by usual eating over three months (evidence from 7 studies in 224 people); but not when compared against energy restriction diets for three months (10 studies; 719 people) or longer (3 to 12 months; 4 studies; 279 people).
We also found that intermittent fasting did not appear to affect blood sugar levels when compared against usual eating over three months (3 studies; 95 people); energy restriction diets over three months (9 studies; 582 people); or energy restriction diets over 3 to12 months (4 studies; 279 people).
The weight losses and changes in blood sugars reported in the studies were small. These changes were not deemed to be clinically significant.
Only four studies reported unwanted effects of intermittent fasting: some people taking part reported mild headaches. Only one study reported on people's well-being, showing a small increase in scores for physical well-being.
Our confidence in our results
We are not confident in our results. We found limitations in the ways that the studies were designed, conducted and reported; and in some studies, the results varied widely, or were not consistent. Our results are likely to change if more evidence becomes available.
We did not find enough good certainty evidence to know whether intermittent fasting could prevent CVD. We found that intermittent fasting may help people to lose more weight than 'eating as usual' (not dieting) but was similar to energy restriction diets. We need further research to test the benefits and potential harms of intermittent fasting, and to test if it might affect how many people die or develop CVD.
We are uncertain about the effects of intermittent fasting on clinical events such as mortality, myocardial infarction and heart failure due to lack of data for these outcomes. The individual meta-analyses show that intermittent fasting may be effective in reducing weight when compared to ad libitum feeding and may be as effective as continuous energy restriction. Despite this, these changes appear to be clinically insignificant at short-term follow-up. The quality of the available evidence is low to very low which means that many areas of uncertainty remain. Further research is needed to understand which patient groups would and would not benefit from intermittent fasting (e.g. patients with diabetes or eating disorders) as well as the effect on longer-term outcomes such as all-cause mortality and myocardial infarction.
Cardiovascular disease (CVD) is the leading cause of death worldwide. Lifestyle changes are at the forefront of preventing the disease. This includes advice such as increasing physical activity and having a healthy balanced diet to reduce risk factors. Intermittent fasting (IF) is a popular dietary plan involving restricting caloric intake to certain days in the week such as alternate day fasting and periodic fasting, and restricting intake to a number of hours in a given day, otherwise known as time-restricted feeding. IF is being researched for its benefits and many randomised controlled trials have looked at its benefits in preventing CVD.
To determine the role of IF in preventing and reducing the risk of CVD in people with or without prior documented CVD.
We conducted our search on 12 December 2019; we searched CENTRAL, MEDLINE and Embase. We also searched three trials registers and searched the reference lists of included papers. Systematic reviews were also viewed for additional studies. There was no language restriction applied.
We included randomised controlled trials comparing IF to ad libitum feeding (eating at any time with no specific caloric restriction) or continuous energy restriction (CER). Participants had to be over the age of 18 and included those with and without cardiometabolic risk factors. Intermittent fasting was categorised into alternate-day fasting, modified alternate-day fasting, periodic fasting and time-restricted feeding.
Five review authors independently selected studies for inclusion and extraction. Primary outcomes included all-cause mortality, cardiovascular mortality, stroke, myocardial infarction, and heart failure. Secondary outcomes include the absolute change in body weight, and glucose. Furthermore, side effects such as headaches and changes to the quality of life were also noted. For continuous data, pooled mean differences (MD) (with 95% confidence intervals (CIs)) were calculated. We contacted trial authors to obtain missing data. We used GRADE to assess the certainty of the evidence.
Our search yielded 39,165 records after the removal of duplicates. From this, 26 studies met our criteria, and 18 were included in the pooled analysis. The 18 studies included 1125 participants and observed outcomes ranging from four weeks to six months. Of quantitatively analysed data, seven studies compared IF with ab libitum feeding, eight studies compared IF with CER, and three studies compared IF with both ad libitum feeding and CER. Outcomes were reported at short term (≤ 3 months) and medium term (> 3 months to 12 months) follow-up.
None of the included studies reported on all-cause mortality, cardiovascular mortality, stroke, myocardial infarction or heart failure.
Body weight was reduced with IF compared to ad libitum feeding in the short term (MD -2.88 kg, 95% CI -3.96 to -1.80; 224 participants; 7 studies; low-certainty evidence). We are uncertain of the effect of IF when compared to CER in the short term (MD -0.88 kg, 95% CI -1.76 to 0.00; 719 participants; 10 studies; very low-certainty evidence) and there may be no effect in the medium term (MD -0.56 kg, 95% CI -1.68 to 0.56; 279 participants; 4 studies; low-certainty evidence).
We are uncertain about the effect of IF on glucose when compared to ad libitum feeding in the short term (MD -0.03 mmol/L, 95% CI -0.26 to 0.19; 95 participants; 3 studies; very-low-certainty of evidence) and when compared to CER in the short term: MD -0.02 mmol/L, 95% CI -0.16 to 0.12; 582 participants; 9 studies; very low-certainty; medium term: MD 0.01, 95% CI -0.10 to 0.11; 279 participants; 4 studies; low-certainty evidence).
The changes in body weight and glucose were not deemed to be clinically significant.
Four studies reported data on side effects, with some participants complaining of mild headaches. One study reported on the quality of life using the RAND SF-36 score. There was a modest increase in the physical component summary score.