When studying the effects of various interventions on human health, it is notoriously difficult to glean actionable information from any one single scientific study. Often these are observational studies that are subject to many confounding factors and limited by small sample sizes, poorly controlled conditions, and short observation periods. Even well-controlled, randomized clinical trials can often reach opposing conclusions.
We can reach better conclusions, however, by performing a systematic literature review: examining the scientific literature as a whole by using strict criteria to select the highest quality studies and evaluating the results of the all of the studies as a whole. Although there may be individual studies within the review that support either side of an assertion, if a large majority of the studies support one side of an argument, we can be much more confident about their conclusions.
The 2007 report, Food, Nutrition, Physical Activity, and the Prevention of Cancer by the World Cancer Research Fund and American Institute for Cancer Research does just this. I will summarize the relevant findings of the review below. The full report is available here: http://www.dietandcancerreport.org/
Because obesity is an established cause of cancer, the report examined the risk factors for weight gain, overweight, obesity. The systematic literature review of this area covered 207 publications investigating the determinants of weight gain, overweight and obesity. The report concludes:
"The epidemiological evidence on physical activity is substantial and consistent. There is robust mechanistic evidence, particularly in relation to its impact on appetite regulation and energy balance. Overall, the evidence that all types of physical activity protect against weight gain, overweight, and obesity is convincing. It has this effect by promoting appropriate energy intake. Conversely, the evidence can be interpreted as showing that sedentary living is a cause of weight gain, overweight, and obesity."
The panel gives the evidence linking physical activity to a decreased risk of weight gain, overweight and obesity its highest ranking of convincing. The panel ranks the strength of evidence linking energy-dense foods, sugary drinks, and fast food to an increased risk of weight gain, overweight and obesity as probable.
Furthermore, while Siv has claimed that changes to appetite that accompany physical activity cancel the effects of exercise, the report claims otherwise:
"Control [of appetite] seems to be least effective at relatively low levels of physical activity, meaning that sedentary people tend to gain weight more readily than active people. Conversely, although high levels of physical activity increase energy requirements and appetite, the likelihood of consuming more than is needed is lower."
Therefore, although some may be able to produce studies showing otherwise, the preponderance of high quality scientific evidence strongly supports the assertion that physical activity can prevent weight gain, overweight and obesity.
Since this topic was originally on cancer, I will quote the panel's conclusions about the effects of sugar consumption on the risk for cancer:
"The evidence is hard to interpret. There is limited evidence suggesting that sugar is a cause of colorectal cancer."
This does not mean that there is convincing evidence that sugar is not a cause of cancer. Rather, the report concludes that there is not sufficient high quality evidence to reach a solid conclusion.