TAKING A SCIENTIFIC APPROACH TO THE AFFILIATE NUTRITION CHALLENGE Introduction As more members of the CrossFit community adopt a Paleo diet, there is a great need for data quantifying the benefits of a Paleo diet. We hear plenty of anecdotal evidence for Paleo-related improvements, from weight loss and leaner body composition, to improvements in cholesterol and other blood markers, to increased strength and metabolic conditioning, to reduced pain and improved immune response, to better energy and mood. However, how reliable are these effects? Those recommending a switch to Paleo need a foundation of results and critical evaluation. To encourage their members to try out the Paleo diet, many CrossFit affiliates are embarking on Paleo “challenges”, often with a competitive element to help with motivation. At our affiliate, we tried that, but only a few people stuck with the diet long-term, so the second time around, we tried a different approach. We decided to embark on a Paleo study rather than a challenge. There was no competitive element; rather, affiliate members agreed to participate in a scientific investigation, following the standard guidelines for experimentation with human participants. This meant that they committed to participating, filled out consent forms, followed an experimental protocol, provided data at regular intervals, and participated in an interview and debriefing at the end of the study. We chose to do a study rather than a challenge for several reasons: 1) to add to the body of data evaluating the effects of a Paleo diet, 2) to provide a template/example for other affiliates wishing to add to the body of data, and 3) to encourage our members to try Paleo in a non-competitive, supportive context and judge the results for themselves. Participants agreed to eat a Paleo-Zone diet for 8 weeks. The Paleo diet is based on the diet of our human ancestors from the Paleolithic era. The modern interpretation of Paleolithic nutrition includes meat, vegetables, fruit, seeds, and nuts. It omits all grains, dairy, refined sugar, soy, and starches like potatoes and legumes. The Zone part (see Enter the Zone, Sears, 1995) refers to portion sizes and macronutrient ratios. The Zone ratio is 30 percent protein, 40 percent carbohydrates, and 30 percent fat. The Zone diet uses a concept called blocks, which refer to certain benchmark portion sizes, to save you from doing calculations all the time. Paleo-Zone simply refers to a diet based on Paleo foods with Zone portions/ratios. In addition, participants agreed to track their compliance and several other factors in a daily log, train regularly, get body composition measures taken every two weeks throughout the study, and get performance measures and blood work (a lipid profile) taken at the beginning and the end of the study. We hypothesized that 1) cholesterol and triglyceride measures would improve, particularly for participants outside the normal range, 2) strength and metabolic conditioning performance would improve on average, 3) subjective wellness measures (energy and affect) would improve throughout the study, and 4) body composition would improve (weight loss and lowered percent body fat).
Participants All participants were members of our affiliate who volunteered for the study. The sample we chose certainly shows a self-selection bias: motivation, dedication, health, and fitness levels of CrossFitters are likely higher than in the general population, and this was a subset of CrossFitters who were motivated and dedicated enough to change up their diet for eight weeks and make a number of study-related commitments. Recruitment Participants were recruited with announcements on our affiliate’s website. All participants received a thorough written description of the study, attended a lecture on eating Paleo-Zone given by our resident nutrition expert, and gave informed consent to participate. Potential benefits of participating included:
Participant Background Information 21 participants volunteered for the study, ranging in age from 18 to 59, with an average age of 37.5 (s=9.8). There were 13 men (average age 40.4) and 8 women (average age 32.8). There were no existing medical conditions other than high cholesterol (5) and exercise-induced asthma (1). At the beginning of the study, participants averaged
We asked them to rate their current diet on a scale of 1-10, with 1 being healthy and 10 being unhealthy, and ratings averaged 6.9 (s=1.4, range: 4 to 9). We also asked them to rate their motivation to change their current diet, and ratings averaged 8.0 (s=2.0, range: 2 to 10). Eleven of the 21 participants reported that someone else in their household was also eating Paleo. 15 participants completed the study, and 6 dropped out. Six of those 15 participants were missing one or more measures: two were missing blood work, three were missing end-of-study performance measures (strength and metcon) due to illness or injury, and one was missing both. The 15 participants who completed the study ranged in age from 25 to 59, with an average age of 38.7 (s=10.0). There were 11 men (average age 41.7) and 4 women (average age 30.3). At the beginning of the study, these 15 participants averaged
Their healthiness-of-diet ratings averaged 6.4 (s=1.3, range: 4 to 9). Their motivation-to-change diet ratings averaged 8.5 (s=1.1, range: 7 to 10). Six of these 15 participants reported that someone else in their household was also eating Paleo. Reasons for Participating We asked participants why they chose to volunteer for the study, and they gave several reasons. Of the original 21 participants, 11 mentioned performance, nine mentioned health, and three specifically mentioned cholesterol. Seven mentioned body composition – wanting to lean out or lose weight.
Measures We collected a wide range of measures to investigate the effects of switching to a Paleo diet. They included body composition measures, physiological measures (blood work), performance measures (strength and metabolic conditioning), subjective measures, and compliance measures. Body composition measures included weight, percent body fat, and waist circumference in inches. Percent body fat was measured with a scale which required inputs for height, sex, age, and activity level. Scales that measure body composition are not entirely accurate, but because we were interested in before and after differences for individuals and not absolute values of body composition, the scale suited our purpose and was more convenient for multiple measurements than some of the more involved and expensive methods of body fat measurement. Physiological measures included total cholesterol, HDL, LDL, VLDL (when available), and triglycerides. These measures are part of a standard lipid profile. Participants went to their usual doctor or ordered a kit from bloodtestsathome.com. Performance measures included both strength measures and metabolic conditioning (metcon) measures. For strength, we used the CrossFit total (deadlift, back squat, and press), and for metcon, we used the CrossFit workout “Christine,” which is 3 rounds for time of: row 500 m, 12 body weight deadlifts, and 21 box jumps. Subjective measures included daily self-ratings of energy and affect (“How did you feel today, overall?”). Participants rated these on a scale of 1 to 10, with 1 being worst and 10 being best. Compliance measures included daily self-reports of Paleo compliance and training compliance. For the study, participants agreed to consume at least 90% of their calories from Paleo foods and to train at least 3 days a week. Participants also recorded hours of sleep per night to track overall wellness. Subjective measures, compliance measures, and sleep were recorded in an online log (a Google spreadsheet) that was shared only with the study administrator, to protect the participants’ privacy.
Method Participants were told that the study would require an eight week commitment. They were told that by agreeing to participate in the study, they agreed to:
Participants were told that all their information would remain confidential and that their data would not be individually identifiable in any public documents.
Results Reported results exclude two identified outliers. In addition, results for different measures are based on different numbers of participants. This is because some participants provided incomplete information; for example, some were injured during the study and unable to complete a second set of performance measures, but they still provided body composition and physiological measures. A few participants opted not to get blood work completed. Complete descriptive statistics, including measures of central tendency and variability, are available in Tables A-D at the end of the paper. Body composition measures (13 participants) On average, participants lost 7.1 pounds over eight weeks, ranging from a minimum of 2.6 pounds to a maximum of 15.8 pounds. They also lost up to four inches from waist circumference, with an average of 1.4 inches and up to 12.8 percent body fat, with an average loss of 3.0 percent and at worst, a gain of 0.6 percent body fat. Physiological measures (11 participants) On average, participants’ total cholesterol dropped 17 points, with a maximum drop of 106 points. HDL rose on average 1 point, with a maximum gain of 55; LDL dropped on average 18 points, with a maximum drop of 85; and triglycerides dropped on average by 1, with a maximum drop of 42. We had VLDL measures for four participants; it dropped on average 1 point, with a maximum drop of 9. We would expect the largest changes from participants who started with above-normal cholesterol levels (over 200). There were five participants who fit these criteria. When we look at only these five, their total cholesterol fell on average 48 points, HDL rose by 1 point, LDL fell by 46 points, and triglycerides fell by 15. Performance measures (9 participants) Relative strength measures, which are weights lifted divided by body weight, all improved. On average, relative deadlift improved by 8 percent, relative back squat improved by 9 percent, and relative press improved by 3 percent. Relative CrossFit Total improved by 20 percent, on average, ranging from -1 percent to + 38 percent. All participants improved on the metcon workout (“Christine”) with an average improvement of 69 seconds, with improvements ranging from 29 to 165 seconds. Subjective measures Energy and affect ratings were similar for participants and across time. Energy averaged a rating of 6.7 across the study, and affect averaged 6.8. Average ratings stayed within a small window (6-7) throughout the study, rose slightly over the first 3 weeks and then stayed fairly steady. Compliance and other measures Compliance with the eating guidelines was good, with participants averaging 6.1 days per week of staying 90% Paleo. Paleo compliance stayed close to 6 days/week throughout the study. Training compliance was also good, with number of training days averaging 3.8 (we set a minimum of 3), and it stayed steady throughout the study, with the exception of participants who experienced an illness or injury during the study period. Hours of sleep per night averaged 7.4. Sleep increased a bit from week 1 to week 2 but stayed fairly flat throughout the rest of the study.
Two Examples In addition to the group results, we would like to highlight the experience of two of our participants. Both participated in the study because of concerns about health and, in particular, cholesterol levels. Both saw the improvements they were hoping for at the end of just eight weeks. In fact, their improvements exceeded their expectations. The first participant lost nearly 16 pounds and 2 inches off his waist, and also dropped his percent body fat. At just over six feet tall, he went from 205 pounds to 189 pounds and from a 37 to a 35 inch waist. His percent body fat went from 17.4 to 15.7. His given reason for participating was to lower his cholesterol, based on a recommendation from his doctor, and he did just that. He started with a total cholesterol measure of 240 (over 200 is considered "high") and dropped to 197, his HDL ("good cholesterol") went up from 43 to 55, his LDL ("bad cholesterol") dropped from 173 to 127, his VLDL cholesterol dropped from 24 to 15, and his triglycerides dropped from 118 to 76. The second participant, at 5 feet, 9.5 inches tall, went from 186 to 178 pounds, from a 36 inch waist to a 34 inch waist, and he dropped his percent body fat by more than 2. His bloodwork also revealed big changes. He started the experiment with the highest total cholesterol of all the participants, at 339. By the end of the 8 weeks, his total cholesterol had dropped to 233, an improvement of over 100 points, putting him much closer to the 200-point cutoff used by physicians. His LDL cholesterol dropped from 225 to 140, and his triglycerides dropped from 292 to 250. These were impressive changes over just 8 weeks, especially keeping in mind that the amount of food to be consumed was not limited in any way, only the types of food. While there is debate about the value of blood cholesterol levels in predicting health outcomes (see extensive review and discussion in Good Calories, Bad Calories, Taubes, 2007), most of us must answer to our family physicians' concern about elevated cholesterol. To show that a Paleo diet not only does not increase cholesterol levels but can actually drastically reduce them provides people with some justification for trying a nutrition plan that is not advocated or even accepted within mainstream medicine. For these two participants, impressive numbers overcame the skepticism of their doctors.
Feedback from Participants Twelve participants provided written feedback about their participation in the study. We asked them to evaluate their experience, about their perceived improvements, about support during hte study, and about their future diet plans. Evaluation of the Study
Improvements
Support
Future Plans
Discussion In summary, we saw positive results across the board. Body composition improvements were moderate in most participants and large in a few. On average, over the eight-week study, participants lost about 7 pounds, 1.5 inches off their waistlines, and 3 percent body fat. The largest losses were 16 pounds, 4 inches, and 13 percent body fat. While cholesterol, LDL, and triglycerides all fell on average, the largest drops were for the five participants who started the study with what is considered above-normal cholesterol (over 200). Their total cholesterol fell on average 48 points, LDL fell by 46 points, and triglycerides fell by 15, with maximum drops of 106, 85, and 42, respectively. Relative strength measures all improved; the relative CrossFit Total improved on average by 20 percent (and up to 38 percent). All participants improved on the metcon workout with an average improvement of over a minute. Of course, we would expect to see these improvements over time with consistent training, and without a control group, we cannot be certain that the improvements in performance were due to the change in diet. However, it is encouraging to see improvements in these measures over just eight weeks. Self-reported energy and affect stayed relatively stable across the study, on average. Compliance with the eating guidelines was good throughout the study, with participants staying 90% Paleo more than six days a week. Training compliance was excellent; we set a minimum of 3 days per week, and participants averaged 3.8 days per week (not including periods of injury or illness) In addition, our attrition rate was fairly low, losing only six of the 21 affililate members who volunteered for the study. Twelve of the 15 participants who completed the study provided written feedback about the study, and all of them indicated that they planned to continue to eat a Paleo diet, which is extremely encouraging. Because we don’t have a random or representative sample from the general population, we can’t claim that our study results generalize to the general population. However, we believe they likely do generalize to other CrossFitters, and so we wanted to publish them here to make them available to other CrossFitters. In addition, we hope that this study can lay the foundation for other researchers to conduct more rigorous investigations of the effects of a Paleo diet in a broader population. We hope that other CrossFit affiliates can use this study as an example if they wish to embark on in-house Paleo challenges. We found that taking a study approach fostered a supportive and positive atmosphere rather than a competitive one, and this resulted in better commitment and morale among our members. Overall, our pariticpants saw improvements across a variety of measures, including body composition, physiological measures, and athletic performance. A large proportion of participants had a positive experience and plan to stick with the new eating habits they learned. The study also piqued interest in members who did not participate in the study. To conclude, we would like to share some of our favorite quotes from participants:
Table A. Body composition results
Table B. Relative strength results
Table C. Blood cholesterol and triglycerides results
Table D. Blood cholesterol and triglycerides for participants starting with total cholesterol>200
|