Aerobic exercise vs weights. Who will win?
There is no question that diet and exercise both play a role in fat/weight loss. Schools of thought range from the “it matters more what you eat” camp to the “it matters more how much you move with subgroups ranging from the “it matters how many calories you eat” camp to the “it matters if you do weights” camp and then the all-popular, “just move more” vs. “move, but move really really fast in short intervals of time” camps.
Jen Sinkler of Experience Life magazine brought this study to my attention, because I belong to the, “it matters if you do weights and probably doesn’t matter much if you move more, whether at a steady pace or really really fast in burst intervals.” camp ((mostly out of laziness and abhorrence of “cardio”)
Slentz CA et al. The effects of aerobic versus resistance training on visceral and liver fat stores, liver enzymes and insulin resistance by HOMA in overweight adults from STRRIDE AT/RT: A randomized trial. Am J Physiol Metab. doi:10.1152/ajpendo.00291.2011, 2011.
One of the first factoids a person learns about when the topic of “weight loss” is the difference between “losing fat” and “losing weight”. Despite the misplaced metric of the scale, however, one quickly comes to realize that the goal of “weight loss” is actually “fat loss”, and that the two losses are not, in fact, equivalent.
As the journey into complexity continues, the difference between visceral and subcutaneous fat eventually surfaces; subcutaneous fat being the fat that makes you..well, visibly fat, and visceral fat being the fat that exists around your organs, deep to the abdominal fascia. It is also sometimes vilified as “the dangerous fat”.
Visceral fat as well as fat _in_ your liver are linked to several disease states, the existence of which are debated (such as Syndrome X) and not (such as type 2 diabetes). High levels of visceral fat are also correlated with dying (from any cause, which, epidemiologically is somewhat problematic, but we won’t get into the specifics of all-cause mortality).
Circulating liver enzymes (most notably, alanine aminotransferase, or ALT) have also been linked to diabetes and non-alcoholic fatty liver disease; though it is unlikely ALT is directly in the casual pathway of either of these two conditions.
We know that aerobic exercise improves insulin sensitivity and that it decreases visceral fat. What isn’t clear, according to these researchers is the part that resistance training plays in visceral fat levels as well as circulating ALT levels.
This was a randomized controlled trial (yay!) comparing the effects of aerobic training alone vs. resistance training alone vs. aerobic and resistance training on subcutaneous and visceral fat levels as well as circulating liver enzyme levels.
The authors looked for adults between the age of 18 and 70, with BMI’s between 26-35, mild to moderate cholesterol problems who lead sedentary lives (defined as physically active less than 2 times per week). Subjects also had to be non-smokers and without a history of diabetes, hypertension or heart disease.
After screening 3145 potential subjects, 234 were found to meet the inclusion and exclusion criteria. These individuals were then asked to maintain their existing lifestyle for 4 months prior to the actual study. This is also known as a run-in period. There are many reasons to do a run-in period, but in this case, the authors wanted to weed out people that would not be committed to the study to reduce the number of drop-outs after the study got started. They lost 38 people during the run-in period, which left them with 196 subjects in the study itself.
Unfortunately, the authors did not report how they randomized these 196 subjects into the three treatment groups, but they randomly assigned each subject to one of the three groups.
Subjects were all weighed (average of three weights taken over 2 weeks on different days), height measured once. What sounds like a VO2max test or at least some sort of maximal stress test was performed at baseline and after the intervention period (since it’s unlikely sedentary people can actually meet the criteria for a true VO2 max).
The amount of visceral and subcutaneous fat was determined by a CT scan. An image of a single cross-section of the abdomen was taken at the level of the L4 pedicle.
Total weight lifted during each workout was recorded either by a supervising personal trainer or electronically.
The resistance-training only group worked out 3 days per week, 3 sets of 8-12 reps for 8 exercises.
The aerobic-training only group didn’t appear to have “days per week”, but rather did approximately 12 miles (19.2km) per week of milage at 85% peak VO2. These were done on treadmill, elliptical trainers, and or/exercise bikes, or a combo of the three. Subjects were instructed to stay within a certain heart rate zone.
The combined group did both workout protocols each week.
The trial time was 8 months.
ALT and AST were measured for each time point by blood draw.
I didn’t care much for their statistical approach, but in the end, I’m not sure it makes that much of a difference. Each group had 13 paired t-tests to compare the pre- with the post- values with no correction for multiple comparisons.
I do not agree (for reasons that I have stated over and over again in this blog) that a non-significant p-value indicates “no difference” between groups, which these authors use liberally.
The average BMI in this study was around 30.
In terms of program success, all groups showed both statistically and practically relevant changes in their workout indicators. The subjects in the aerobic training group increased their measured peak VO2 as a group, and the subjects in the resistance-training group increased the amount of total weight lifted per workout over the trial time.
With respect to the variables of interest though, results were lacklustre. The authors commonly interpreted a p-value of less than 0.10 as a “trend towards significance”, which has been condemned by biostatisticians across the world as being misleading and inaccurate. If we actually corrected for 13 multiple comparisons (there were actually 45), the p-value required to interpret a difference as “statistically significant” would be around 0.004.
When you separate out all the smoke and mirrors though, what you end up with not that many relevant changes. On average, the aerobic and combined groups lost 2kg (+/- 3ish kg) over 8 months. On average, the resistance group gained 0.7kg (+/- 2.4kg) of body weight.
There were only two between-group differences detected under than 0.05 level, but above the 0.01 level. No differences were under the 0.01 level. Visceral fat surface area was less in the aerobic-only group when compared to the resistance-only group; and changes in the liver/spleen ratio were higher in the aerobic group compared to the resistance-only group (but this analysis was not performed on all subjects, as only 67% of the scans had both spleen and liver in them). In terms of interpreting the liver/spleen ratio itself, a denser liver, compared to a spleen, might indicate less fat in the liver itself.
So where does this leave us on the original question?
I think there are a few angles that are important to consider in this study:
1) If you are not sedentary with a BMI between 26 and 35 with mild to moderate cholesterol problems, this study doesn’t apply to you. It doesn’t really help inform you on whether you should or should not be incorporating aerobic or resistance training in your current regime. So if things are working, just keep on truckin’.
2) It’s really hard to tell what the importance of losing cross-sectional area of fat actually is. While some of these differences between aerobic and resistance training were statistically significant (if we are generous about the multiple comparison oversight), there’s no data to help us figure out if losing 16 square centimetres of visceral fat in a single CT slice at the L4 pedicle level means anything in terms of the stuff we’re actually interested in, such as diabetes, insulin resistance and dying.
3) Despite the increase in peak VO2 and a nicely progressive increase in total weight lifted in a session of resistance-training, I’d say an average loss of roughly 5 pounds of bodyweight with a range of 11 pounds of weight gain to 20 pounds of weight loss over EIGHT months is pretty dismal for any program with, or without resistance training.
Albeit this was not a diet study; and it does isolate some of the factors that might inform us on how to prescribe exercise to people trying to lose weight. However, if 144 people only managed to lose, at most, 20 pounds over eight months (with most clustered around the FIVE pound area), I’d say the take-home message here is that our primary focus should NOT be on exercise prescription for the purposes of decreasing body weight, fat loss (visceral or subcutaneous) or liver enzymes. I’d even go so far as to argue that if you’re around a BMI of 30 and consuming around 2000 calories a day, you probably can’t exercise your way to weight loss without eating less.
The bottom line: The motto, “Eat less, move more,” is often used by fitness/diet professionals as a simple message on how to lose weight. I think what this study highlights in the end, is that this message should probably also change, probably to, “EAT LESS, move however much you want to,” or even just, “EAT LESS. Waaaay less.”