Aerobic exercise cut fat and boosted fitness but barely touched inflammation

C-reactive protein is one of the most commonly measured markers of inflammation in the blood, and it is often quoted alongside cardiovascular risk. Exercise is routinely promoted as a way to bring that marker down, which has made it a selling point for aerobic training programmes. Individual trials had disagreed, with some reporting reductions and others none at all. Pooling those trials is the only sensible way to see whether the average effect is real or simply noise from small studies.

The authors searched electronic databases, cross-referenced and hand searched for randomised controlled trials published in English between January 1, 1990 and January 1, 2006. To qualify, a trial had to measure C-reactive protein, use an aerobic exercise intervention lasting 4 weeks or more, and enrol adults aged 18 or older. Five studies met those criteria, covering 323 men and women, with 171 people in exercise groups and 152 in control groups, and 6 separate outcomes for C-reactive protein available for pooling. Secondary outcomes were body weight in kilograms, percentage of body fat and maximum oxygen consumption. Results were combined using a random-effects model.

The primary outcome barely moved. C-reactive protein fell by roughly 3% in the exercise groups, an average of 0.11 mg/L, and the range around that estimate stretched from a fall of 0.39 to a rise of 0.17 mg/L, so it was not statistically significant. Body composition told a different story: body weight dropped by 3.4 kg and body fat by 1.4%, both statistically significant reductions of around 4%. Maximum oxygen consumption rose by 3.3 mL per kilogram per minute, an increase of about 12%.

The result complicates a popular claim rather than demolishing it. Fitness and body composition responded exactly as decades of exercise research would predict, which suggests the trials were capable of producing change where change was there to find. The inflammation marker simply did not follow, at least across these five trials and this length of intervention. Open questions include whether people with higher starting inflammation, longer programmes, or larger samples would show something the pooled estimate here cannot detect.

This is a reason to be sceptical of the specific promise that a few weeks of cardio will lower your C-reactive protein reading. The same evidence is a reason to expect meaningful improvements in aerobic capacity, weight and body fat from the same training. If you have had an inflammation marker measured, a single blood value is not the best way to judge whether your training is working. Judging by fitness and body composition is better supported by what these trials actually recorded.