This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
There is increasing interest in Pinterest as a method of disseminating health information. However, it is unclear whether the health information promoted on Pinterest is evidence-based or incorporates behavior change theory.
The objective of the study was to determine the presence of health behavior theory (HBT) constructs in pins found on Pinterest and assess the relationship between various pin characteristics and the likelihood of inclusion of HBT.
A content analysis was conducted on pins collected from Pinterest identified with the search terms “nutrition infographic” and “healthy eating infographic.” The coding rubric included HBT constructs, pin characteristics, and visual communication tools. Each HBT construct was coded as present or not present (yes=1, no=0). A total theory score was calculated by summing the values for each of the 9 constructs (range 0-9). Adjusted regression analysis was used to identify factors associated with the inclusion of health behavior change theory in pins (
The mean total theory score was 2.03 (SD 1.2). Perceived benefits were present most often (170/236, 72%), followed by behavioral capability (123/238, 51.7%) and perceived severity (79/236, 33.5%). The construct that appeared the least was self-regulation/self-control (2/237, 0.8%). Pin characteristics associated with the inclusion of HBT included a large amount of text (
These findings suggest that current Pinterest infographics targeting healthy eating contain few HBT elements. Health professionals and organizations should create and disseminate infographics that contain more elements of HBT to better influence healthy eating behavior. This may be accomplished by creating pins that use both text and images of people and food in order to portray elements of HBT and convey nutritional information.
Chronic disease is a major public health concern. In the United States, heart disease, cancer, and stroke cause over 50% of all deaths [
In an attempt to alter these unhealthy eating trends, health professionals are increasingly using the Internet to improve dietary behaviors in populations [
Pinterest is beginning to impact health educators and the way they share information [
Pinterest may be a useful tool to disseminate information about dietary behaviors, but there is no research about the content of nutrition and diet pins. In particular, it is of interest whether they contain constructs of health behavior theory (HBT), as health promotion materials containing more elements of HBT have been demonstrated to be more effective in changing behavior [
The study protocols met the exemption criteria of the university’s institutional review board. No human subjects were involved in this study and only existing, publicly available data were collected for analysis.
This content analysis evaluated HBT in healthy eating infographic pins selected from Pinterest. Two public health graduate students trained in HBT and nutrition coded the pins to evaluate the extent to which constructs of the health belief model and social cognitive theory, 2 major theories of behavior change, were included. KJ, JLW, and KS are graduate students studying HBT. As it relates to nutritional qualifications, KJ is a registered dietitian nutritionist and was responsible for the nutrition-specific codes in the instrument. JLW has a BS in nutritional science, and KS works professionally as a nutrition coach.
The sample was collected from Pinterest in September and October 2015. The study authors created a new Pinterest user account so that no search history would influence the search results. The study sample was identified using the following terms independently entered into the main Pinterest search bar: “nutrition infographic” and “healthy eating infographic.” The first 250 eligible pins that were returned for each term were saved for analysis, resulting in 500 initial pins. Eligible pins included English-language pins that addressed some aspect of nutrition or healthy eating. Examples of this included pins that explained the benefits of vitamins and minerals or provided tips for a healthy diet. Examples of ineligible pins included those addressing physical fitness or the use of food as a beauty product. An Excel spreadsheet (Microsoft Corp) was used to save the URL for each pin and to record the number of repins, likes, and comments. Finally, duplicate pins were deleted resulting in a final sample of 238.
The methodology for identifying pins was adapted from previous eHealth content analysis research [
The researchers coded each pin in the study sample using an HBT-based coding rubric adapted from previous content analysis studies [
JHW is a senior health communications researcher and trained the other study authors in content analysis research during biweekly sessions over the course of 4 months. First, all authors met to define and reach a common understanding of all the study variables. Second, a coding instrument was developed and pilot tested by jointly coding pins and then resolving any discrepancies in codes. As part of the biweekly training sessions and in response to discrepancies in codes, which were either resolved until agreement was reached or they were removed from the instrument, JHW trained the other study authors to be able to identify the HBT constructs that were measured in this study. Third, the authors revised the instrument to remove coding options for which agreement could not be achieved or that were otherwise determined to be not applicable. Lastly, interrater reliability was established using a subset of the study sample.
The data were coded into an electronic spreadsheet and then exported for analyses. The coding rubric included 4 primary categories: (1) pin characteristics, (2) visual communication tools, (3) health belief model constructs, and (4) social cognitive theory constructs.
Pin characteristics included the URL and the number of repins, likes, and comments for each pin. The variables repins, likes, and comments were not normally distributed and underwent square root transformations for analyses. Pin affiliation/author (business, government, or individual) and the pin category of healthy eating (macronutrient, micronutrient, disease management, portion control/weight management, other) were recorded. Each pin was also coded with respect to whether or not it addressed a particular diet trend, such as paleo (eating only whole unprocessed foods) or avoiding genetically modified foods. Lastly, each pin was assessed to determine whether or not the coder would recommend the pin for use in promoting healthy eating. The coders were uniquely qualified to evaluate this aspect of pins.
The visual communication aspects of each pin were coded. These characteristics included the amount of text in the pin (no text, text light [covering <50% of the pin], or text heavy [covering >50% of the pin]); whether or not there was a person depicted in the pin (yes/no, and if yes, whether or not it was a photo of a real person or a cartoon); whether or not food was depicted in the pin (yes/no, and if yes, whether or not it was a picture of real food or a cartoon); and finally, the dominant colors of the pin (vibrant colors, muted colors, or black and white).
Constructs from social cognitive theory and health belief model were coded. Each HBT construct was coded as present or not present (yes=1, no=0).
To ensure intercoder reliability, 2 coders evaluated a common 5% (26/500) of the study sample, which is considered adequate in cases of a large sample [
All analyses were conducted using Stata version 12 (StataCorp). Descriptive statistics were computed and summarized in aggregate. A total theory score was calculated by summing the values for each of the 9 construct codes (range 0-9). Combining multiple constructs to form a total theory score has been done previously [
A total of 238 infographic pins from Pinterest were analyzed. Characteristics of the pins are described in
Pin characteristics.
Variables | na (%) | |
Business | 198 (83.2) | |
Individual | 3 (15.6) | |
Government | 37 (1.3) | |
Macronutrientsc | 25 (10.6) | |
Micronutrientsd | 36 (15.3) | |
Disease managemente | 30 (12.7) | |
Portion control/weight managementf | 47 (19.9) | |
Otherg | 98 (41.5) | |
Text heavy (>50%) | 109 (46.2) | |
Text light (<50%) | 127 (53.8) | |
Vibrant | 155 (65.7) | |
Muted | 78 (33.1) | |
Black and white | 1 (0.4) | |
No person depicted | 157 (66.2) | |
Cartoon | 66 (27.9) | |
Photograph | 14 (5.9) | |
No food depicted | 15 (6.4) | |
Cartoon | 126 (53.4) | |
Photograph | 95 (40.3) | |
Referencesj | 100 (43.7) | |
Professional recommendationk | 82 (34.8) |
aNot all categories in every variable will sum to 238 due to some instances of missing data.
bPin affiliation: who authored the pin.
cMacronutrients: carbohydrates, proteins, fats.
dMicronutrients: vitamins and minerals.
eDisease management: cancer, obesity, arthritis, cardiovascular disease, etc.
fPortion control/weight management: identifying and promoting healthy portion sizes.
gOther: included highly specific nutrition topics including top 10 healthiest foods, lists of superfoods, harms of specific foods, and nutritional content of specific fruits and vegetables.
hText: pins were categorized as text heavy if more than 50% of the infographic contained text and text light if less than 50% of the infographic contained text.
iColor: pins were categorized as vibrant colored, muted in color, or black and white.
jReferences: presence of references or citations.
kProfessional recommendation: whether or not the pin provided accurate information and advice to recommend to clients.
Engagement metrics.a
Engagement metrics | Average | SD | Range |
Comments | 2.35 | 4.37 | 0-30 |
Repins | 1318 | 2267 | 3-15,848 |
Likes | 215 | 440 | 0-5027 |
aVariables were not normally distributed and underwent transformations for analyses. The mean and SD are presented here in raw form for interpretation.
The prevalence of each construct [
The number of repins (
Prevalence of health behavior theory constructs among pins.
Construct | Description [ |
n (%) |
Perceived benefits | Belief about the potential positive aspects of a health action | 170 (72.0) |
Perceived barriers | Belief about the potential negative aspects of a health action | 37 (15.6) |
Perceived susceptibility | Belief about getting a disease or condition | 38 (16.1) |
Perceived severity | Belief about the seriousness of a condition or the consequences of leaving it unaddressed | 79 (33.5) |
Self-efficacy | Belief that one can achieve the behavior required to execute the outcome | 11 (4.7) |
Self-regulation/control | Controlling oneself through self-monitoring, goal-setting, feedback, self-reward, self-instruction, and enlistment of social support | 2 (0.8) |
Behavioral capability | Providing tools, resources, or environmental changes that make new behaviors easier to perform | 123 (51.7) |
Observational learning/modeling | Beliefs based on observing similar individuals or role models perform a new behavior | 3 (1.3) |
Subjective norm | An individual's perception of social norms or his/her peers' beliefs about a behavior. A function of an individual's normative beliefs and motivation to comply with beliefs | 17 (7.2) |
Summary of health behavior theory scores.a
Mean | SD | Frequency | |
Overall | 2.03 | 1.20 | 238 |
Business | 2.09 | 1.21 | 189 |
Government | 2.67 | 0.577 | 3 |
Individual | 1.67 | 1.12 | 36 |
aVariables were not normally distributed and underwent transformations for analyses. The mean and SD are presented here in raw form for interpretation.
Correlation between pin engagement metrics and health behavior theory score.
Total theory | Likes | Repins | ||||
Coefficient | Coefficient | Coefficient | ||||
Repins | 0.13 | .04 | 0.95 | <.001 | 1.00 | — |
Likes | 0.16 | .01 | 1.00 | — | 0.95 | <.001 |
Comments | 0.17 | .01 | 0.66 | <.001 | 0.64 | <.001 |
Shown in
Regression analysis of pin characteristic and presence of health behavior theory.
Theory square root | Coefficient | SE | 95% CI | ||
Professional recommendation | .197 | .0577 | 3.4 | .001 | .0827 to .3104 |
Text heavy | .141 | .0547 | 2.58 | .011 | .0334 to .2491 |
Color | −.110 | .0578 | −1.89 | .060 | −.2237 to .0044 |
Photograph of person | .196 | .0581 | 3.38 | .001 | .0816 to .3107 |
Cartoon of food | .132 | .0551 | 2.40 | .017 | .0235 to .2407 |
References | .184 | .0567 | 3.25 | .001 | .0727 to .2964 |
The purpose of this study was to determine the level of integration of HBT in nutrition infographics on Pinterest. Secondarily, we identified factors associated with its inclusion. The secondary interest was largely exploratory and was intended to inform future research and efforts that might use Pinterest to promote healthy eating. Overall, it was found that HBT constructs are integrated into nutrition pins only minimally. The low levels of HBT are not surprising, as pins can be created and shared by lay parties who may not have training in HBT or know why its inclusion may improve pin impact on behavior. This may be a recurring problem for health professionals wishing to adopt and use social media and other health technology in their health promotion efforts. Payne et al [
Interestingly, there were very few pins created by government organizations such as Let’s Move! and the US Department of Agriculture, although they had the highest average HBT scores. It seems plausible that pins created by government health organizations would contain higher levels of HBT because individuals working in these settings may be more likely to have training in a related discipline. In general, however, these results may indicate that businesses and individuals who create nutrition infographics for Pinterest lack the training to effectively incorporate HBT into social media campaigns. While research elsewhere indicates that health professionals currently use and understand social media in vocational roles only minimally [
The majority of pins were focused on very specific nutrition topics, such as lists of superfoods, health foods for babies, and nutritional content of specific fruits and vegetables. The second most common infographic category was portion control and weight management. It is promising that the latter category was a relatively large percentage of the sample, especially considering that portion sizes have dramatically increased over the past several decades [
While Pinterest is dominated by young middle- to upper-middle-class white females, Pinterest use grew significantly among individuals living in rural locations, those with an annual salary of less than $30,000 per year, and those aged 50 years and older from 2013 to 2014 [
The most common HBT constructs in this sample were perceived benefits, behavioral capability, and perceived severity. The remaining 6 constructs appeared in 15% or less of the sample, with self-regulation and control being the least common. It is concerning that self-efficacy was one of the least incorporated constructs because self-efficacy has been shown to be a significant predictor of behavior change [
Factors positively associated with HBT included the presence of heavy amounts of text, a photograph of a real person, or a cartoon image of food. This suggests that depicting HBT constructs on Pinterest is more likely to be accomplished by incorporating a combination of text, people, and images rather than text alone. Indeed, research has demonstrated that images, rather than text, are the most desirable way to communicate information [
Social media engagement is a key performance indicator that links social media usage to action [
The researchers only assessed the image and did not evaluate the websites the pin linked to. Additionally, there are many duplicates of each pin shared on Pinterest, so engagement metrics may be spread among pins. While the researchers made note of information accuracy, it was not the primary focus of the study. A more rigorous analysis of content accuracy could be done to determine accuracy of messages.
Promoting dietary modifications is an important public health strategy for preventing chronic disease. Pinterest as a social media platform has the potential to communicate health information and influence healthy eating behavior through infographics. However, current Pinterest infographics targeting healthy eating contain few HBT elements.
It is recommended that health professionals and organizations create and disseminate infographics that contain more elements of HBT to better influence healthy eating behavior and be more effective in changing behavior [
This study may also help dietitians, public health workers, and health educators make an informed decision about whether or not to recommend Pinterest as a health information source. Because many health professionals use infographics to disseminate health information [
chronic obstructive pulmonary disease
health behavior theory
None declared.