Four diets with exercise – which did best?

Executive summary

* This week's paper used a technique that is generally considered to be the highest level of evidence (systematic review and meta-analysis). It turned out to be a good example of the technique only being as good as the way in which it is used.

* The study aimed to review the outcomes (weight, BMI, fat/lean mass) when different exercises (aerobic exercise, resistance training or mixed exercise) were combined with different diets (calorie restriction, 5:2 intermittent fasting, time-restricted feeding, or the ketogenic diet).

* It didn’t review the different exercises in the main paper or for the abstract results.

* It claimed that 78 trials were found and reviewed. This number was wrong and that was before a number of trials were found to fail the researchers’ inclusion criteria.

* Results fell into two categories:

1) Each diet plus exercise being reviewed against exercise alone.

There were many issues here – the main one being that barely a quarter of the trials used to claim a result actually examined the diet plus exercise vs exercise alone. Most examined diet plus exercise vs diet alone. That tells us the impact of exercise, not diet.