Paleo Then and Now: A Five-Year follow-up Survey of the Ancestral Health Community A Five-Year follow-up

The Ancestral Health (or Paleo) movement is a collection of people who apply an evolutionary perspective to improve their health through lifestyle changes (e.g., diet, physical actively, sleep). In 2013, a survey of this population revealed that its participants did not align with media stereotypes; in contrast, survey participants were predominantly white, female, and middle aged. Since diet and health trends change over time, the goal of this survey study was to collect data on the members of the movement and compare the survey responses from five years earlier. In addition, we surveyed people who have left the movement to determine why they left and if they follow a particular diet now. Broadly, we hypothesized that participant demographics would remain the same while lifestyle characteristics (i.e., how they participate in the movement) would change. The survey was distributed online through various platforms (social media, blog posts, and mailing lists). A total of 1506 (54% female) participants completed the survey and a majority (75%) were still a member of the movement. Overall, numerous participant responses differed from the 2013 survey (18 questions; p<0.05 for all). One of the main changes between survey years was a general aging of the participants as demonstrated by an increase in participant age, income, and education level. This suggests that the movement is not capturing new and/or younger individuals. Further, the main reason participants left the paleo movement was a change to a new specific diet (16%; two most reported diets: low-carbohydrate and ketogenic diet), although 39% of participants who left do not currently follow any particular diet. We conclude by hypothesizing reasons for the changes and potential implications for the future of the movement.


Introduction
The 2002 book, The Paleo Diet, is credited as starting the contemporary 'paleo' or 'ancestral health' movement [1]. The paleo diet, and lifestyle pattern (e.g., physical activity and sleep) that quickly followed, is produced by viewing modern health from an evolutionary perspective [2]. Clinical research has demonstrated beneficial effects of following a paleo diet [3,4]. Meanwhile, the paleo diet has since been labeled as a fad diet [5,6] and is surrounded by controversy [7][8][9]. Numerous concerns of the paleo diet have been presented: minimal clinical support, negative side effects, low calcium intake, and cost [6]. In addition, the often associated exercise program, CrossFit, has been found to be comparable to other workout programs [10]. Moreover, the ability the ability of this movement to go mainstream has been called into question, suggesting that this movement, as a whole, is simply a passing fad [11].
In 2013, the first survey of the ancestral health community revealed information about the population and addressed common stereotypes of the movement [12]. Certain stereotypes about the paleo movement dominated the media and the public's imagination at the time. Followers of a paleo or ancestral lifestyle were often thought to be rather vain "modern cavemen": athletic, single, meat-eating, young, white, and male [12]. Despite these common stereotypes, the survey suggests that the majority of participants were actually: white, female, middle aged (mean 38 years old), in a committed relationship, highly educated, relatively affluent, and motivated by weight loss and health concerns. Thus, while a few of the common preconceptions may have held up, others did not [12].
Since then, more research has been conducted on the various aspects of the ancestral health movement. That, plus the continued presence of other popular diets fad diets [5], suggest it is possible that the ancestral health community has changed. Thus, a follow up survey of the ancestral health community was warranted. By comparing the 2018 survey results to the 2013 data, the goal of this survey is to determine if, and how, the ancestral health movement has changed over the past five years. Specifically, we try to demonstrate who currently chooses to adopt an ancestral health lifestyle, how and to what degree they follow the lifestyle, and the obstacles they face while maintaining this lifestyle. In addition, data were collected on why some left the paleo movement and what kind of dietary approaches they now follow. Generally, we hypothesized that the basic demographic characteristics (e.g., age, gender, race, education) between the two data sets would remain the same while lifestyle characteristics (i.e., how they participate in the movement, adherence, motivation) would change.

Study Design
A cross-sectional survey was distributed using a self-administered online questionnaire [13]. While distribution was directed at our target audience, any individual (except for those under the age of 18) who came upon the survey link could complete the questionnaire. Inclusion criteria included: access to the survey, being 18 years or older, and agreeing to informed consent. The study population consisted of adults who identified themselves one of two ways: currently following a paleo diet or no longer following a paleo diet. Most of the survey instrument was duplicated from a 2013 survey on the same topic [12]. The responses for identical questions between the current survey and the 2013 survey were compared to determine difference between survey populations. Additional questions were added to capture information from participants who no longer follow a paleo lifestyle. Survey questions were used to collect information about: demographics, lifestyle choices (food, exercise, and sleep), attitude toward evolution, perceived health outcomes and diet/lifestyle change motivation, obstacles, and duration. The survey was managed through the internet-based management system (qaultrics.com). See Supplemental Text 1 for survey for details and questions.

Survey Distribution
Survey distribution was modelled after the 2013 survey [12] to minimize difference in the population who had access to the survey. In particular, social media platforms (Instagram, Facebook, Twitter, and Reddit) were used for distribution of the survey link. Prominent influencers and organizations in the ancestral health and paleo community were contacted (n=26) and requested to assist with survey promotion via social media or email lists. Specifically, the leading paleo organization, the Ancestral Health Society (ancestralhealth.org), also promoted the survey. Survey distribution began October 23 rd , 2018 and ended November 4 th , 2018 (13 days). No incentive was included for participation.

Participants
A total of 1,506 participants consented to the study. Only 1,463 answered the first question which identified their current relationship with the paleo diet (currently follow the paleo diet: n=1,133; no longer follow the paleo diet: n=251; never followed the paleo diet: n=79). A response rate was unattainable due to distribution methods, as it is impossible to know how many received the survey. The average duration for those who completed the survey was 18.16 minutes. For those who consented and started the survey, the completion rate was 93.38% for all participants. The dropout rate, those who did not submit the last survey question, was 9%. Data from our 2013 survey [12] was used as a comparison group to determine difference in the ancestral health community over five years.

Use of Human Subjects
This survey was exempt from the Institution Review Board approval from the State University of New York, New Paltz, Human Research Ethics Board. Participant confidentially was ensured since no identifying information was collected.

Statistical Analysis
A chi-square test was used to test categorical variables between the 2018 and 2013 survey data. An unpaired t-test was used to compare numerical data between the 2018 and 2013 survey data. If the numerical data failed normality, a Mann-Whitney Rank Sum Test was used. Alpha was set to 0.05 for all tests. Statistical tests were applied using SigmaStat 10.0 (Systat Software, San Jose, CA).

General Survey Statistics and Demographics
A total of 1,508 participants from 55 countries consented to the study. Demographic data that differed from the prior 2013 survey year are presented in Figure 1. Non-significant or noncomparable data are presented in Supplemental Table 1. The median age increased 5 years (36 to 41 years of age; t-test: p=<0.001) between the two survey years. Overall, the total number of participants who have children or are married/with a partner have both increased (33% to 51%, 1.55-fold change and 57% to 74%, 1.29-fold change, respectively). Further, the proportion of participants without children had decreased (67% to 47%, 0.73-fold change). The largest change in the reported education level was an increase in the proportion of participants with a master's degree (22% to 25%, 1.14-fold change). The proportion of participants who reported an income greater than $100,000 increased between survey years (41% to 50%, 1.22-fold change). Participants were from the same top four countries and the same top five US states as the 2013 survey.

Diet, Exercise, and Sleep
Data on participant lifestyle choices that differed from the prior 2013 survey year are presented in Figure 2. Non-significant or non-comparable data are presented in Supplemental Table 2. Only 38% (n=412) of participants reported that they avoid all three food groups (Grains, Legumes, and Dairy). Whereas only 22% reported avoiding all the three food groups and alcohol. The majority (88%) of participants reported taking at least one supplement and the two largest changes between survey years was an increase in reported magnesium consumption (49% to 69%, 1.41-fold change) and a decrease in fish oil consumption (65% to 52%, 0.80-fold change). Overall, the proportion of participants who partake in strength training decreased (81% to 78%, 0.96-fold change) whereas the proportion of participants who partake in aerobic training increased (73% to 77%, 1.05-fold change). The overall proportion of participate who participate in CrossFit decreased (16% to 11%, 0.69-fold change) between survey years.

Motivation, Obstacles, and Duration
Data on participants' motivation, obstacles, and duration that differed from the prior 2013 survey year are presented in Figure 3. Non-significant or non-comparable data are presented in Supplemental Table 3. The two biggest changes in the reasoning for going paleo were an increase in recovery from illness (21% to 25%, 1.19-fold change) and a decrease in weight loss (31% to 28%, 0.90-fold change). The mean reported difficulty of maintain a paleo lifestyle (1-Very Difficult, 7-Very Easy) was significantly lower (2013 mean: 4.997 (SD: 1.429), 2018 mean: 5.163 (SD: 1.443), T-test: p=<0.001) for the 2018 participants. The two biggest changes in the main obstacle of following a paleo diet were an increase in "giving up favorite non-paleo food" (21% to 24%, 1.14-fold change) and decrease in "too expensive" (13% to 10%, 0.77-fold change). Of those who chose "other" (n=283), 29% reported that they have no obstacles.

Perceived Health Outcomes
Data on participants' perceived changes in health outcomes are presented in Figure 4. Five categories came back significantly different compared to the 2013 survey year. Overall, the number of participants who reported they were aware of any change (i.e., worsened, stayed the same, or improved) in their blood chemistry increased (57% to 89%; 1.56-fold change). Within this, the proportion of participants who reported improvements in their blood chemistry increased (48% to 68%, 1.42-fold change).

Past Paleo Followers
Data on those who reported they no longer follow the paleo lifestyle (n=251) can be found in Table  1. Of the n=38 participants who selected "Changed to a new specific diet" for their reason for no longer following a paleo lifestyle, n=17 (44%) reported they currently follow the Ketogenic Diet.

Discussion
By comparing survey responses from 2018 and 2013, we identified differences throughout the survey. Interpretation and implications of the differences and similarities are discussed below.

Changes in participation and demographics:
One of the most notable changes between the two surveys is that survey participants are nowon averagefive years older (median age went from: 36 to 41 years old). This could suggest fewer younger people may be joining the movement, while the movement, as a whole, continued to age. For example, the average age of a new members who joined the movement in the three-year period prior to the survey increased (from 37.7 years old to 40.4 years old; 2.7 years older). It is unclear why fewer younger individuals are being drawn to the paleo movement. Perhaps, they are turning to other popular diet trends [5] as the debate of which diet is best for health continues.
Similar to our 2013 study, the majority of the 2018 participants were middle-aged Generation Xers (born approximately from 1962 to 1982), not young Millennials (born approximately from 1982 to 2002). As summed up in our 2013 study, the median age of participants in both surveys of the ancestral health community may also be explained in part by: "older individuals being sicker, and thus in need of alternative health and wellness paradigms. In other words, younger members of society have yet to feel the full effects of the "Diseases of Civilization" (diabetes, heart disease, and cancer, among others), which often accompany the process of aging. Young people may have simply experienced less frustration with ill health and our current health care system, and, as a result, are less motivated to seek out and experiment with alternative ways of caring for their bodies." (Schwartz and Stapell, 2013;page 15) Other demographic data suggests that our sample represents a rather privileged segment of society. While the 2013 participants reported very high levels of educational attainment, and a remarkably high level of income, compared to the national averages, five years later, the participants responses for these categories increased even further. For example, in 2018, 77% of participants held a 4year degree or higher, and 50% of participants reported earning $100,000 per year or more. Some of these changes may reflect the fact that, on average, the recent sample population is now older and further into peak earning years. These results may also be related to the fact that race and wealth are strongly correlated in the United States [14], so it follows from the SES data that the vast majority of participants are white (91% in both 2013 and 2018). Similar to the 2013 results, the 2018 sample had a higher frequency of those who believed in evolution without god compared to the national average (66% vs. 33%) [15].
Geographically, participants were from the same top four countries (United States [US], Canada, Australia, United Kingdom) and five US states (California, Washington, Texas, Colorado, New York) as the 2013 survey. Since the survey was conducted and promoted through English-language blogs, organizations, and social media located in the United States, this likely explains the Englishspeaking countries at the top of the list. Within the United States, geographic location may play a role in both SES and the decision to follow a paleo lifestyle, as many participants are located in predominately high-income "blue" and "purple" states: California, New York, Washington, Colorado, and Texas. As stated in our previous 2013 study, these five states may have dominated both surveys due to their large cosmopolitan cities. Further, a similar percentage of participants came from outside the US (31% vs 30%) and a similar number of countries (55 vs 60) between the two survey years.
New to the 2018 survey, we asked participants about their political party affiliation. Democrat was the most common response. However, there was a relatively equal distribution among the four main political parties in the United States (in descending order): Democrat, Independent, Libertarian, and Republican. This undermines the common belief that most paleo followers are Libertarians due to the paleo-libertarian connection [16]. Therefore, the ancestral health community includes quite a diverse political community.

Changes in lifestyles choices: diet, exercise, and sleep
As in the 2013 survey, we found high avoidance of some non-paleo foods, namely grains and legumes, but not of dairy and alcohol among participants. However, only 22% of participants reported avoiding all three food groups (grains, legumes, and dairy) and alcohol. In other words, there is substantial variation in the food groups individuals avoid and many do not follow the typical standard paleo diet. Further, a majority of participants regularly consume starchy carbohydrates (e.g., white rice, potatoes, sweet potatoes, etc.). The majority reported regularly taking one or more supplements, with magnesium and Vitamin D being the most popular, and some differences were found compared to the 2013 study (participants reported taking less fish oil and more taking magnesium). As with the 2013 results, the majority of survey participants (78%) reported participating in any form physical activity, which is much higher compared to the estimated 80% of the US being insufficiently active [17]. Further, the reported hours of sleep (7.4 hours) is higher than the US average (6.8 hours [18]). The overall percentage of participants who participate in CrossFit also decreased over the five-year period. The exact reasons for these differences over the five years is unclear and could simply reflect variation in health trends or other observed changes in demographics.

Changes in responses to motivation, obstacles, and duration
Motivation is one key factor in adopting and maintaining a new diet and lifestyle. As in the 2013 survey, the same two most popular factors for adopting a paleo lifestyle were reported: weight loss and recovery from illness. However, the biggest change between the 2013 and 2018 in regards to reasons for adopting a paleo lifestyle was a decrease in those who reported "weight loss" and an increase "recovery from illness". This suggests more people are now motivated to adopt a paleo diet to address unresolved health problems, and fewer individuals are eating this way to lose weight. As suggested below, other popular diets, such as the ketogenic diet, may be competing for the attention of those interested in losing weight, which could explain the decrease seen here. Of note, during this time of increased discussion surrounding the possible perils of climate change, in both surveys, not a single participant chose environmental sustainability as the primary reason for going paleo. This suggests that concerns about the environment simply are not critical motivators for adopting a paleo lifestyle.
In terms of obstacles to maintaining a paleo diet and lifestyle, 'social pressure' and 'giving up nonpaleo foods' were most often cited as the major obstacle to living a paleo lifestyle in both survey years. The two biggest changes were a decrease in survey participants indicating it was 'too expensive' and an increase 'giving up non-paleo foods.' As in 2013, under the category of "Other" in the 2018 survey, participants described a wide variety of obstacles, including: eating out of the house, preparing and finding food, and time constraints. Despite these obstacles, still the majority of participants (70%) said it was easy to maintain a paleo lifestyle (66% from the 2013 survey).
The reported increase of ease may also be due to the increasing number of paleo-friendly foods founds both in restaurants and supermarkets. Further, given the increased financial cost of following the paleo diet [19], the reported increase in average income between 2013 and 2018 survey data may be responsible for the increase in reported ease.
One of the most striking differences between the two survey years relates to adoption and duration.
In 2013, 67% of participants reported joining the movement within the past two years, and 85% in the last three years. The current survey suggests a slowdown in the recent rate of adoption as only 33% of 2018 participants joined in the past three years. We believe this difference reflects the movement's rapidly growing popularity in 2013 compared to 2018. Further, the difference between the two survey years could reflect the novelty of the movement is now over since most participants had only been living that way for a short amount of time in the 2013 survey. This difference suggests the rapid period of growth followed by a sharp decline in growth rates which is characteristic of fads. Moreover, to further illustrate the five-year aging of participants between the two survey years, 45% of the 2018 participants reported living this lifestyle for more than five years, compared to only 4% in the 2013 survey.

Changes in perceived health outcomes
As in the 2013 study, the 2018 survey participants perceived themselves to be better off physically and mentally after adopting a paleo diet and lifestyle as they reported improvements in five of the seven health categories were reported. Also, similar to 2013, the majority of participants in 2018 reported that they were unsure whether their fertility had improved or not. The largest change in perceived health outcomes between the two surveys can be found in the blood chemistry category: more participants were aware of any change in the blood chemistry, and the proportion of participants who reported improvements in their blood chemistry increased. This difference between the two surveys may be due to increased awareness about the potential benefits of routine lab work, the increase in household income, and greater access to blood work testing. Furthermore, improved access to medical care after 2013 with the Affordable Care Act (also known as Obamacare), in the US, may have had an influence on these numbers as well. While these responses are self-reported, and many are completely subjective, participants from both survey years reported overall improvement in a broad range of health outcomes by transitioning to a paleo diet and lifestyle.

Past paleo followers
In an effort to understand why some people choose to abandon a paleo lifestyle, we asked two questions on the 2018 survey addressed to those who left the movement. While, the largest reason for stopping paleo was a change to new specific diet (16%), all other responses suggest no switch to a different diet (e.g., limited food choices, lack of motivation or no specific reason). Furthermore, when asked what diet they followed now, the largest response (39%) was "no particular diet." This could suggest that those who left paleo are not continuing their interest in promoting health via diet. The two most popular diets currently followed by participants who left paleo were low carbohydrate and ketogenic diet (selected by 23%). Both of these diets have the added advantage of not being associated with the common paleo media images of egotistical "modern cavemen." Further, these approaches are not associated with the theory of evolution and thus do not have the baggage of debates around evolution common in the United States [15]. Together, the responses to these two questions suggest 1) the popularity of other fad diets: low carbohydrate and ketogenic diets, and 2) a large amount of paleo followers are not continuing to use a specific diet as a way to promote health. This further demonstrates the "fad" aspects of the paleo diet, as many seem to only adopt a paleo diet for a short period of time, and not continue their interest in dieting itself. Moreover, in a 2018 survey study, 76% of surveyed paleo diet followers consider their diet a fad diet [20]. While we did not ask this question in our survey, we have no reason to believe our sample would differ on perception of paleo as a fad diet.

Limitations and future research
Given the cross-sectional, internet-based, and self-selected anonymous design of this study, we are limited in our ability to comment on reasons for the changes seen. While valuable, the results of our two surveys only capture two time points of the paleo story; they do not offer us a complete story of paleo adherents over time. Future qualitative interviews could capture data that could potentially explain the differences seen in this study. This study allows us to capture broader trends in the movement, thus we can only hypothesize the reasons for the changes. Our data collection strategy was such that we are unable to judge the change in size of the ancestral movement, if any, over the past five years. Although, far fewer individuals completed our survey in 2018, we are unable to say definitively if that reflects less overall participation in the movement, or even less enthusiasm for a paleo diet and lifestyle today. However, this current study's sample size is similar to a 2015 online social media diet survey study (n=1,326 paleo diet followers) [21]. It is possible the movement's participants have moved elsewhere and were no longer in the survey distribution path. To our eye, the channels we relied upon for survey distribution remain central to the paleo movement, but we lack reliable data on the outlets most frequented by paleo adherents. In addition, the low sample size of those who no longer follow the paleo diet (n=242), limits the interpretation and generalization of these data.
Future research could build upon this second survey in several ways. First, as described in our 2013 paper, the examination of the ancestral health movement would be well served by an estimate of the total number of paleo adherents. Such an accurate study of the composition of the movement is only possible with a representative national survey. Second, our understanding of the movement would be improved by the collection of longitudinal data. A panel study would help identify, 1) how individuals develop or sever ties to the movement, 2) how individual practice of the paleo diet changes over time in response to new needs, and 3) how consistently paleo is practiced while individuals claim adherence to the lifestyle. Furthermore, how this diet trend reflects other fad diet trends would be of interest. For example, while low-carbohydrate diets (e.g., Atkins) were initially labeled as fad diets, low-carbohydrate diets have withstood the test of time, leading some to question if the low-carbohydrate diet should have been labeled a "fad" at all [22]. While research on the ancestral health movement, specifically the paleo diet, has been building over the past five years, there are still a number of unaddressed questions, 1) is the paleo lifestyle and demographic composition reflective of our results?, 2) are these demographics similar to those of other dietary trends/fads?, and 3) in what ways, if any, does the paleo movement represent more than just a fad diet? In a time when massive amounts of conflicting information about diet and lifestyle flood our screens daily, more detailed information about diet fads/trends could help people navigate the noise.

Conclusions and overall implications
From 2013 to 2018, the ancestral health movement appears to have changed in some significant ways. Overall, the results from our 2018 survey suggest a maturing of the paleo movements; participants are in more committed relationships, older, more highly educated, more affluent. Further, as demonstrated by the decrease in the number of people joining the movement in the past three years, plus the aging of those who joined in the past three years, this suggest that fewer young people, and total people, are joining the paleo movement. Like our 2013 survey, these results further distance the ancestral health movement from the common young, vain, male, and single "modern caveman" media stereotypes; participants were still predominately white, female, and motivated by health and weight loss concern. They also suggest that the ancestral movement represents a very well-resourced segment of society. Collectively, this survey, and the prior 2013 study, allows us to document the population of individuals following a popular diet that is unique because it is based on the central keystone of all of biology: evolution. A diet and health movement based on a scientific theory with as much explanatory potential as evolution seems as though it should not be another passing fad; however, the results from this survey suggest that this may indeed be the case.

Supplemental Tables
Supplemental Table 1 All Other States (including Washington DC; n=44) 5% Note: statistically different demographics between participants from 2013 and 2018 are presented in Figure 1. Statistical difference was not tested for country and state of origin data. Data listed in descending order. Italicized items are created by the authors and were not answer options on survey.

Start of Block: Main Question Block
Q28 <div style="text-align: center; "> <span style="font-size:16px;"><strong>For the purposes of this survey, please take the word "paleo" to mean: <br> <br>The effort to optimize human health today by examining our evolutionary or ancestral past. (  o God created human beings pretty much in the present form at one time within the last 10,000 years or so.