Flexitarian Diets and Health: A Review of the Evidence-Based Literature

A flexitarian or semi-vegetarian diet (SVD) is one that is primarily vegetarian with the occasional inclusion of meat or fish. Of late, there appears to be an increasing movement toward this practice. There has not been a recent update on these diets from a health perspective. Using the National Centre for Biotechnology Information PubMed database, a search was made for all studies published between 2000 and 2016 that met defined inclusion criteria. A total of 25 studies were located with 12 focusing on body weight and diet quality. There was emerging evidence suggestive of benefits for body weight, improved markers of metabolic health, blood pressure, and reduced risk of type 2 diabetes. SVD may also have a role to play in the treatment of inflammatory bowel diseases, such as Crohn’s disease. Given that there is a higher tendency for females to be flexitarian yet males are more likely to overconsume meat, there is a clear need to communicate the potential health benefits of these diets to males.

iNTRODUCTiON "Flextarianism" is a neoteric term that has been emerging in the scientific and public sectors recently. Added to the Oxford English Dictionary in 2014, flexitarian is a portmanteau of "flexible" and "vegetarian, " referring to an individual who follows a primarily but not strictly vegetarian diet, occasionally eating meat or fish (1). Despite the global demands for meat, it appears that there are now a growing number of flexitarian consumers who abstain from eating meat regularly (2).
Most consumers can be grouped into meat consumers, meat avoiders, or meat reducers (3). The trend toward flexitarian diets (FDs) appears to reflect consumers who are "meat-reducers, " eating meat within meals on some but not every day of the week (3), as with typical "meat-eaters". This definition is most closely in line with that of semi-or demi-vegetarianism. Subsequently, the terms are often used inter-changeably in the literature. For example, in one publication semi-vegetarian diets (SVDs) are defined as those significantly reducing meat intake on at least 3 days of the week (4).
The FD seems to recognize the fact that meat is an important source of protein, fat, and micronutrients (5,6), yet also considers the ethical sides, such as the need to avoid intensification and improve animal welfare (7). It also considers evidence that long-term consumption of increasing amounts of red meat and particularly processed meat may increase the risk of mortality, cardiovascular disease, type 2 diabetes, and certain forms of cancer such as colon cancer (6). Recently, the International Agency for Research on Cancer classified red meat as probably carcinogenic and processed meat carcinogenic to humans (8).
Research from NatCen's British Social Attitudes survey found that 29% of people in Britain have reduced the amount of meat that they ate in the past 12 months (9). The definition of meat reducers included reductions in all meats except fish. In particular, women (34%) were most likely to reduce their meat intake. Similarly, 39% of 65-to 79-year olds had reduced their red meat intake compared with 19% of 18-to 24-year olds. The report also showed that men (23%) were shifting and reducing their meat intake. Over half (58%) cited health reasons along with saving money, concerns over animal welfare, and food safety (9).
The aim of the present paper was to review the evidence looking at the FD/SVDs and health, using defined inclusion criteria. While some discussion papers have been published about vegetarianism (10) and low-meat diets and health (11), no publications have focused on the current trend of flexitarianism or SVDs. The present paper set out to evaluate the evidence-based looking as FD/SVDs from a public health perspective.

MeTHODS
The National Centre for Biotechnology Information (NCBI) search engine (PubMed) was used to extract relevant Englishlanguage papers published between January 2000 and June 2016. Data files were extracted from the NCBI collection depository and imported into Covidence software used to create systematic reviews.
For inclusion, studies needed to clearly define the terms flexitarian, semi-vegetarian, or demi-vegetarian used in their study. Data results and findings also needed to be reported and analyzed separately from other forms of dietary patterns. Randomized controlled trials (RCTs) and observational studies were included. Articles were excluded if they were published before 2000, a pilot study, or focused solely on vegan diets or vegetarianism.

ReSULTS
The NCBI search identified 46 papers and after an adjustment for replica papers, 39 articles remained for assessment. Of these, 14 papers were discarded after reviewing the abstracts and article content as they did not meet the inclusion criteria. This left 25 articles for general review. This included 21 epidemiological studies and 4 randomized controlled or clinical trials. The algorithm of qualifying publications is shown in Figure 1. Of these, 12 studies were conducted in the US or Canada, 6 in Europe, 5 in Asia, and 2 in Australasia.

Definitions
Definitions of flexi-semi-vegetarianism were extracted from the identified papers (Table 1). Definitions tended to vary between studies. For example, some authors specified that participants on an SVD restricted their intake of red meat (12,13), while others restricted fish (14). In studies using data from the Adventist Health Study-2, SVDs were defined as those consuming dairy products and/or eggs and meat (red meat and poultry ≥1 time/ month and <1 time/week) (15)(16)(17)(18)(19)(20).
In a more general sense, other authors defined SVDs as those containing moderate levels of animal products (21,22), though it was not specified what "moderate" was. One paper defined these as eating red meat, poultry, or fish no more than once a week (23), while other studies reported participants as being semivegetarian (SV) if they excluded red meat from their diet but ate other meats (13).

Body weight
As shown in Table 2, six studies focused on SVDs and body weight. Two RCTs looked at the effects of different plant-based diets in relation to weight loss. In one study, authors undertook a 6-month RCT, where overweight adults were allocated to five different plant-based diets. Vegan diets were associated with significantly higher levels of weight loss by the end of the study (22). A Korean study reported that postmenopausal women maintaining an SVD over 20 years had a significantly lower body weight, body mass index (BMI), and percentage of body fat compared with non-vegetarians (NVs) (12).
Cross-sectional data from 71,751 participants taking part in the Adventist Health Study-2 (2002)(2003)(2004)(2005)(2006)(2007) showed that BMI was highest in NVs (mean 28.7 kg/m 2 ), slightly lower in SVs (mean 27.4 kg/m 2 ), and lowest in strict vegetarians (mean 24.0 kg/m 2 ) (18). These findings are similar to earlier trends (2002-2006 analysis) showing that mean BMI was lowest in vegans (23.6 kg/ m 2 ) and incrementally higher in lacto-ovo vegetarians (LOVs) (25.7 kg/m 2 ), pesco-vegetarians (PVs) (26.3 kg/m 2 ), SVs (27.3 kg/ m 2 ), and NVs (28.8 kg/m 2 ) (15). Cross-sectional research on 9,113 young Australian women (22-27 years) identified that SVs had a lower BMI and tended to exercise more than NVs (13).  Intake of red meats, poultry, or fish, but not only fish was more than or equal to once per month but less than once per week Tantamango Cancer Four studies were found to fulfill the inclusion criteria. One study examined vegetarian patterns in relation to breast cancer (BC) risk using data from 96,001 adults taking part in the prospective Adventist Health Study-2 (2002)(2003)(2004)(2005)(2006)(2007). Findings showed that vegans had a significantly lower risk of BC compared with vegetarian and NVs (24). With regard to prostate cancer risk, data from 26,346 males taking part in the Adventist Health Study-2 found that only vegan diets were associated with reduced prostate cancer risk (25). Equally, an earlier analysis of this study showed that vegan diets had statistically significant protection for overall cancer incidence (hazard ratio = 0.84; 95% confidence interval: 0.72, 0.99) (26). Other work using data from the same North American study discovered that PVs had the lowest risk of colorectal cancer, followed by LOVs, vegans, and SVs when compared with NVs (20).

Diabetes and Metabolic Syndrome
Six studies were found looking at SVDs in relation to markers of metabolic health or risk of diabetes. In one study, authors observed that postmenopausal women following an SVD for more than 20 years had significantly lower glucose, insulin levels, and homeostatic model assessment of insulin resistance compared with NV controls (12).
Other research using data from India's third National Family Health Survey 2005-2006 (n = 156,317) from adults aged 20-49 years found that the consumption of lacto-, lacto-ovo-, and SVDs were associated with a reduced likelihood of diabetes compared with NV diets, after data adjustments (14). Equally, research from the Adventist Health Study-2 found that cases of diabetes developed were lowest in vegans (0.54%), followed by SVs (0.92%), LOVs (1.1%), and PVs (1.3%) compared with 2.1% in NVs (16).
Furthermore, cross-sectional data from the Adventist Health Study-2 (n = 773) showed that the prevalence of metabolic syndrome was highest in NVs (39.7%), intermediate in SVs (37.6%), and lowest in vegetarians (25.2%). Data for vegans were not reported (17). Findings from the same study also highlighted that diabetes prevalence was 2.9% in vegans, 4.8% in SVs, and 7.6% in NVs (15). With regard to blood pressure, research involving 26 postmenopausal women from convents found that omnivores had significantly higher systolic and diastolic pressures than the SVs (27).

Diet Quality
Six studies focused on dietary quality, nutritional intakes, and/or status. In a randomized trial, overweight and obese adults were allocated to different plant-based diets. Participants allocated to the vegan and vegetarian diets had significantly improved   Vegan diets showed statistically sig. protection for overall cancer incidence (HR = 0.84) in both genders combined and for female-specific cancers (HR = 0.66). LOV were associated with decreased risk of cancers of the GI system

Diabetes and metabolic health
Kim and Bae (12)   The concentration of C-reactive protein was normal at the final visit in more than half of the patients in remission who were on the SV diet. There was no untoward effect of SV diet  macronutrient profiles and Diet Inflammatory Index (DII) scores (a tool for assessing the inflammatory potential of a diet). Vegan, vegetarian, and PV groups all had significant improvement in the DII score at 2 months but not at 6 months (22). Remaining studies were mainly observational in nature. For example, an internet-based survey using data collection tools from the Belgian Food Consumption Survey and a convenience sample of subjects showed that vegan diets had the lowest total energy and highest fiber intake compared with omnivores (23). Alongside this, results from 96,335 adults in the Adventist Health Study-2 showed found that those eating SVDs had the lowest caloric intakes (1,713 kcal/day) but other than this there were few dietary differences (18).
Cross-sectional research from the Australian Longitudinal Study on Women's Health (n = 9,113) showed that rates of low iron, iron-deficiency, or anemia were highest in vegetarians (42.6%), followed by SVs (38.6%) and then NVs (25.5%) (13). Other research focusing on the omega-3 profile of diets showed that vegan and vegetarian diets led to reductions in eicosapentaenoic acid and docosahexaenoic acid levels compared with semi-omnivores (28).

Additional Health Outcomes
Three studies focused on other health outcomes. One study observed that SVDs helped to prevent relapse of symptoms in patients with inflammatory bowel disease (IBD) (29). Additional research by the same team concluded that up to 32.4 g dietary fiber, delivered via an SVD could be given to patients with IBD, indicating that these could be used as a supportive treatment for Crohn's disease patients (30).
Other work involving 73,308 adults from the Seventh-day Adventist study-2 demonstrated that vegetarian diets, including SVDs were associated with lower all-cause mortality, with results appearing to be more robust in males (19).

DiSCUSSiON
Flexitarian diets have been gaining popularity-a transition that seems to have been fueled by a combination of health, environmental, and animal welfare concerns. The present paper has identified that flexitarian/SVDs could have potential health benefits with strongest evidence appearing to be in relation to weight loss and metabolic health benefits, including reduced diabetes risk and blood pressure ( Table 2).
There is also emerging evidence that SVDs could be an option for patients with IBD, such as Crohn's disease (30). While the mechanisms behind this are yet to be confirmed, it has been speculated that a plant-based diet may be effective for gut inflammation, namely, through the actions of dietary fiber (31).
Physicians focused on non-invasive and cost-effective interventions could help their patients improve health outcomes by encouraging a shift toward diets higher in vegetables, whole grains, legumes and fruits, and fewer animal products (32) with an FD being a useful tool in this transition. Furthermore, it has been estimated that making a transition toward plant-based diets that are aligned with standard dietary guidelines could help to reduce global mortality by 6-10% (33).
Flexitarian diets/semi-vegetarian diets could also be useful in helping those with high meat intakes to fall in line with recommended guidelines. For example, data from the UK National Diet and Nutrition Survey showed that red and processed meat intakes were 84 g/day for men and 47 g/day for women (34). Scientific Advisory Committee on Nutrition guidelines are set at 70 g/day for adults for red and processed meat (35). Subsequently, there is a greater tendency for males to exceed red and processed meat guidelines. Interestingly, looking at the studies reviewed, around 70% of conscious flexitarians were educated females (15,18,20,26). Considering this, there is clear scope to educate males about the health benefits of FDs.
In order to align with meat intake guidelines other approaches can also be taken. These include a greater use of simulated meatlike products with a similar flavor, texture, and color to meat (36). As it is the International Year of the Pulse, there has been much interest in these from a health perspective. Pulses include beans, peas, and lentils, which have been eaten for at least 10,000 years, providing protein, fiber, and essential micronutrients, including iron, folate zinc, magnesium, as well as phytochemicals, such as saponins and tannins (37). It has been estimated that eating just half a cup of beans or peas daily can significantly enhance diet quality and nutrient density, helping consumers to meet dietary recommendations (37).

Limitations
On the whole, FDs appear to have emerging health benefits. However, it should be considered that before any formal recommendations about FDs can be made, official definitions of these diets are needed. For example, the inclusion criteria used in the present article did not always bring up publications using the same definition of SVDs. In Germany, legal definitions of vegetarian and vegan have been compiled by the German Federation of Food Law and Food Science and the Vegetarierbund Deutschland to facilitate the categorizing and labeling of foods, though these definitions are yet to be formalized by the European Commission (38). Similarly, a formal classification of FDs is needed that can be put into appropriate use in future research.
Presently, large cohort investigations also appear to overlook FD/SVDs. For example, the recent European Prospective Investigation into Cancer and Nutrition did not include SVDs when comparing the metabolic profile of meat eaters, vegetarians, and vegans (39). It is also worth pointing out that large dietary surveys, such as the UK National Diet and Nutrition Survey, could look to include a definition and analysis of flexitarian or SVDs.
Equally, impending RCTs looking to investigate the health benefits of FD/SVDs need to align methodologies with the Consolidated Standards of Reporting Trials guidelines for RCTs (40). In the case of observational studies, these should follow the Strengthening the Reporting of Observational studies in Epidemiology guidelines (41). It was a limitation in the present review that the most research came from the US and Canadian Adventist Health Study. The SV dietary approaches taken by these Adventists may be different that of other flexitarians or SVs. More research in the UK and Europe is needed to examine this.
One a final note, in the present review, the focus was on "flexitarianism" as this terminology has been increasingly used by the public press. Nevertheless, it should be considered that terms such as "meat reduction" or "Mediterranean diets" which may also constitute an FD were not included in this review. These could form the basis of future publications, along with an anticipated future growth in studies looking at FD and markers of health.

CONCLUSiON
The trend of flexitarianism does not appear to be subsiding. This review provides a first line of evidence that FDs may have emerging health benefits in relation to weight loss, metabolic health, and diabetes prevention. While most flexitarians presently seem to be female, there is a clear need to communicate the potential health benefits of these diets to males. As not everyone and in particular men might not want to exclude meat altogether, FDs offer a path that includes their dietary preferences yet could improve public health outcomes.

AUTHOR CONTRiBUTiONS
This review has been researched and written by an Independent Nutrition Consultant.

FUNDiNG
This review was supported by Marlow Foods (Quorn Foods) Limited, Stokesley, England. The views expressed are those of the author alone and Marlow Foods had no role in writing the review.