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 Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
Despite efforts made by public health organizations to improve consumption of fruits and vegetables, populations in developed countries usually eat less than the minimum recommended. Social media, such as blogs, represent a unique opportunity for improving knowledge translation in health care because they facilitate interactive communication between the public and health professionals. However, no studies have yet evaluated the effect of blogs to promote dietary behavior changes.
Our study aims to conduct a preliminary assessment before undertaking a full randomized controlled trial (RCT) of the feasibility of using an evidence-based healthy eating blog promoting the consumption of fruits and vegetables among adult women.
A total of 80 women aged 18 years and older (mean 42, SD 13 years) eating less than five servings per day of fruit and vegetables (mean 2.75, SD 1.84 servings) were recruited. Participants were randomized to the healthy eating blog group (n=40), which included a weekly blog post over a 6-month period, or to a control group (n=40) that had no exposure to the healthy eating blog. Blog posts were written by a registered dietitian and focused on the improvement of fruit and vegetable consumption. We targeted four main determinants of the behavior that were identified as the best predictors for fruit and vegetable intake by two systematic reviews: (1) knowledge, (2) attitude, (3) self-efficacy, and (4) motivation. The intervention was considered feasible if (1) more than 70% of questionnaires were completed, (2) attendance rate was more than 90% for in-person appointments with the research coordinator, (3) participants accessed at least 75% of the blog posts, and (4) the attrition rate was less than 25%. Blog access was assessed by collecting the blog browsing history data for each participant.
During the intervention, 26 posts were published on the blog. Pre- (baseline) and postintervention (6 months) questionnaires were completed by mean 97% (SD 3%) of participants. All participants attended their in-person appointments. Women accessed mean 87% (SD 7%) of the posts published during the intervention. Only 3% (2/80) of participants dropped out of the study. Between the healthy eating blog and control groups, a difference of 1.0 servings of fruits and vegetables (
These results suggest that an intervention using a healthy eating blog meets preestablished feasibility criteria. A larger-scale RCT using the same methodology will be conducted to assess the impact of a healthy eating blog on the dietary habits of women.
The incidence of chronic diseases has dramatically increased worldwide [
Although it is clear that the adoption of preventive behaviors such as a healthy diet is associated with health benefits, nonadherence rates to medication and lifestyle changes are estimated to be 50% in developed countries [
Chronic disease prevention and management require sustained lifestyle behavior changes and a long-term commitment, which can require help from a health professional. However, dietary counseling may not be accessible for some patients, such as those living in rural areas, having inflexible working hours and schedule, or having reduced mobility. To overcome these barriers and to inform preventive health care quality improvement, social media interventions in nutrition, such as blogs, could be an effective strategy to reach a large proportion of the Internet population with diverse sociodemographic characteristics, independently of education, race/ethnicity, or health care access [
Despite the fact that health blogs are proliferating at high speed, there is no empirically supported knowledge on the impact of health blogs on consumers’ health behaviors and outcomes to promote healthy dietary behavior changes (eg, increase fruit and vegetable intake). To our knowledge, no study has yet evaluated the effects of an evidence-based healthy eating blog on women’s dietary and eating behaviors. Moreover, attrition rates, which refers to the proportion of users who drop out before completion of the study [
As a secondary objective, we intended to collect clinical data such as fruit and vegetable consumption and anthropometric measurements. Although this feasibility trial was not powered to detect differences in these outcomes, their assessments will be useful to evaluate the data collection tools. It will also provide an indication of the variance in measurement (effect size) to be used for the power calculation for the definitive RCT.
This study was a randomized feasibility trial comparing two groups: control and healthy eating blog (access to the healthy eating blog). The sample size for feasibility trials is typically determined pragmatically, with recommendations of a minimum of 30 participants per group [
An advertisement was sent to a list of people who had indicated interest in participating in the Institute of Nutrition and Functional Foods clinical studies. Also, some members of the research team posted the ad on their personal Facebook page. The eligibility of participants was assessed over the phone based on the following inclusion criteria: (1) a woman aged 18 and older living in the Quebec City metropolitan area, (2) has Internet access as well as an active email address, and (3) consumes less than five servings of fruits and vegetables per day (assessed with a 24-hour recall performed by a registered dietitian). We chose to focus the intervention on women because they are primarily responsible for food purchase and preparation in households. Therefore, nutrition promotion strategies that target women have the potential for reach that goes beyond the individual and can thereby improve their family’s health [
Eligible participants were randomized by the research coordinator, who generated a random order list using the Institute of Nutrition and Functional Foods (INAF)’s Web platform. Participants were then scheduled for an in-person individual first visit at the research center to complete a baseline clinical outcomes assessment. All participants gave written informed consent and received Can $100 financial compensation at their final visit at INAF. This project was approved by the ethics committee of Laval University (2014-084 on May 21, 2014).
As a preliminary step, in 2013 we performed a qualitative study exploring women’s views and expectations regarding healthy eating blogs as a means to improve their dietary behaviors [
The development of the intervention was inspired by the steps of the Intervention Mapping protocol [
Based on clinical experience of the registered dietitians on the research team and on national and international public health campaigns [
Conceptual framework of the intervention. F&V: fruits and vegetables.
The 6-month intervention started on January 13, 2015. Participants received an email to inform them of their allocation (control group or healthy eating blog group). Women in the healthy eating blog group received an identification code and a password to access the healthy eating blog as well as instructions on how to navigate the experimental blog website entitled “Salsa Etcetera” (
Blog posts and comments were read and reviewed by the team of researchers and dietitians before they were published online. Each post included a step-by-step recipe developed by the research team that featured fruits and vegetables. Recipes were described using text and pictures and/or video. Twenty-six posts were created over the 6-month intervention, including the two that were already on the blog when the participants first logged on. Because the majority of women who participated in our preliminary phase mentioned they would like a weekly post to be published (as opposed to twice a week or once every two weeks), participants received a new blog post once a week for 24 weeks. An email was sent every week to inform them that a new post was on the blog. We also sent an email reminder to participants who did not log on the blog for two consecutive weeks.
Home page of the intervention healthy eating blog “Salsa Etcetera.”.
All participants were seen at the INAF twice during the study: once at the beginning and once at the end of the 6-month intervention to complete Web-based questionnaires and for anthropometric measurements using standardized procedures (height, weight, and waist circumference) [
Feasibility was assessed by evaluating compliance rates, participation rates, and attrition rates. Compliance rates were measured for the 80 participants by the (1) completion of questionnaires assessing clinical outcome data measures for our definitive RCT (change in fruit and vegetable intake, changes in eating behaviors, social support, body weight) and (2) attendance at in-person appointments with the research coordinator. Based on a previous study at INAF that involved the completion of a similar number of questionnaires [
This feasibility study was not designed to achieve sufficient statistical power to address behavioral outcomes. However, these outcome measures were useful to provide better guidance with regard to development of the definitive RCT. Among the behavioral outcomes of interest, fruit and vegetable intake was assessed with a validated Web-based FFQ [
The program SAS was used to analyze data obtained from the questionnaires and to calculate descriptive statistics of the healthy eating blog group and the control group. We used
Recruitment took place from October 28, 2014 to December 15, 2014 (7 weeks). During this time, eligibility was assessed for 128 women (
Sociodemographic characteristics, eating habits, and Internet use characteristics of participants (N=80).
Sociodemographic, eating habits, and Internet use characteristics | Healthy eating blog (n=40) | Control (n=40) | ||
Age (years), mean (SD) | 42.0 (13.7) | 42.2 (13.4) | .95 | |
>.99 | ||||
White | 36 (90) | 37 (93) | ||
Other | 4 (10) | 3 (8) | ||
.56 | ||||
High school | 6 (15) | 4 (10) | ||
College | 15 (38) | 13 (33) | ||
University | 18 (45) | 23 (58) | ||
Did not answer | 1 (2.5) | 0 (0) | ||
.42 | ||||
0-19,999 | 4 (10) | 2 (5) | ||
20,000-49,999 | 15 (38) | 11 (28) | ||
50,000-99,999 | 12 (30) | 12 (30) | ||
100,000-149,999 | 4 (10) | 7 (18) | ||
150,000-199,999 | 1 (3) | 1 (3) | ||
≥200,000 | 1 (3) | 0 (0) | ||
Did not answer | 3 (8) | 7 (18) | ||
BMI (m/kg2), mean (SD) | 27.7 (5.2) | 27.1 (6.4) | .21 | |
Fruit and vegetable daily intake (servings), mean (SD) | 2.45 (1.94) | 3.05 (1.70) | .43 | |
.39 | ||||
<1 | 0 (0) | 1 (3) | ||
1-2 | 5 (13) | 7 (18) | ||
3-4 | 12 (30) | 12(30) | ||
5-10 | 17 (43) | 10 (25) | ||
≥10 | 6 (15) | 10 (25) | ||
Computer | 26 (65) | 30 (75) | ||
Smartphone | 4 (10) | 4 (10) | ||
Tablet | 9 (23) | 6 (15) | ||
Did not answer | 1 (3) | 0 (0) | ||
.85 | ||||
Home | 34 (85) | 36 (90) | ||
Work | 4 (10) | 3 (8) | ||
Car or bus | 1 (3) | 1 (3) | ||
Other | 1 (3) | 0 (0) | ||
Read a blog before, n (%) | 36 (90) | 34 (85) | .50 | |
Read a nutrition blog before,a n (%) | 22 (55) | 20 (50) | .65 | |
Read comments on a blog,a n (%) | 29 (73) | 27 (68) | .63 | |
Commented on a blog,a n (%) | 11 (28) | 8 (20) | .43 |
a Only among participants who had already read a blog before.
Participants were mostly white women aged between 22 and 71 years of age (mean 42, SD 14 years) and ate less than five portions of fruits and vegetables per day (mean 2.75, SD 1.83 servings). The majority of these women received college- or university-level education and more than half of respondents (38/70) had a family income greater than Can $50,000. Mean BMI was 27.5 (SD 4.6), and 31% (25/80) were obese according to Canadian Health Risk Classification (BMI ≥30) [
Preintervention and postintervention questionnaires were completed by all participants in both the healthy eating blog and control groups. The monthly questionnaire was completed by 97% (37/38) of participants on average. The lowest completion percentage (92%, 35/38) was observed on month 4. All participants in both groups attended their in-person appointments (100%, 38/38). Therefore, compliance rates all reached our feasibility criteria.
Flowchart of participants. F&V: fruits and vegetables.
As shown in
Participation rates on the healthy eating blog during the 6-month intervention (n=38).a
Participation | Month | Total | ||||||
1 | 2 | 3 | 4 | 5 | 6 | |||
Date range | Jan 13-Feb 11 | Feb 12-Mar 11 | Mar 12-Apr 8 | Apr 9-May 6 | May 7-Jun 3 | Jun 4-Jul 1 | ||
Articles posted each month, n | 6 | 4 | 4 | 4 | 4 | 4 | 26 | |
Total comments,b n | 141 | 74 | 80 | 73 | 71 | 75 | 514 | |
Comments/post, mean (SD) | 23.5 (9.9) | 17.5 (2.0) | 16.5 (2.0) | 17.0 (5.5) | 15.0 (1.5) | 16.8 (2.8) | 17.7 (4.7) | |
Comments/participant,b mean (SD) | 3.7 (0.3) | 1.9 (0.2) | 2.1 (0.1) | 1.9 (0) | 1.9 (0.1) | 2.0 (0.0) | 13.5 (0.7) | |
Total printed recipesb,c | 35 | 19 | 6 | 10 | 30 | 27 | 127 | |
Total video viewsb,c,d | 0 | 0 | 0 | 0 | 6 | 4 | 10 |
a n=38 because two participants dropped out.
b For the month’s new posts and precedent posts.
c n=37 because data logs were impossible to collect for one participant who had a firewall installed on her computer.
d First video was posted on May 6 and the second video was posted on June 4. All four views during month 6 were for video 2.
Percentage of participants (n=37) who viewed each post. Note data logs were impossible to collect for one participant who had a firewall installed on her computer. The bold blue line represents the feasibility criterion of 75% of participants viewing each post.
The attrition rate was 3% (2/80), which is less than our preestablished 25% criteria. The two participants who dropped out were in the healthy eating blog group and the reasons were external to the study as mentioned previously.
We observed no significant differences in fruit and vegetable consumption between the healthy eating blog group (mean 2.44, SD 1.91 portions/day) and the control group (mean 3.05, SD 1.70 portions/day) at baseline (
Mean fruit and vegetable consumption for the two groups before and after the 6-month intervention (n=78).
Our study is the first to evaluate the feasibility of an intervention using an evidence-based healthy eating blog to improve women’s dietary and eating behaviors. According to our preestablished criteria, the intervention is deemed feasible.
Measured completion of pre- and postintervention questionnaires (100%) and in-person appointment attendance (100%) met our preestablished compliance rate criteria. Unsurprisingly, all questionnaires were completed because participants filled them in during their in-person appointment, which everyone attended. Compliance rates were high for both the healthy eating blog and control groups. The incentive of Can $100 given at the end of the study may have contributed to compliance. Completion of the monthly questionnaires by the healthy eating blog group (mean 96.5%, SD 3.2%) also met our compliance rate criteria of 70%. The reminder sent to participants when they did not complete the monthly questionnaire may have contributed to achieving this high completion rate.
It is known that increased participant use of and engagement with an intervention platform are associated with greater behavior change [
In concordance with studies included in Williams et al’s systematic review of RCTs examining the use of social media to promote healthy diet and exercise in the general population [
Participants in our preliminary study [
We did not have a criterion for the number of comments on the blog, but compared to existing blogs, the participation rate was very high [
Only a few people watched published videos (16% for video 1 and 11% for video 2). This is consistent with findings from Strekalova et al [
Attrition rate was very low (3%, 2/80). This could be due to the honorarium given to participants in both the control group and healthy eating blog group at the end of the study (Can $100). We also think that this study did not imply much time and involvement because participants could log on to the blog from home and were not forced to write comments and interact with the dietitian-blogger or other participants. A recent study examined the feasibility of delivering a group Web-based and face-to-face weight-loss intervention to 40 young adults with low income and reported that 30% of participants completed the 5-month intervention [
Our sample may not be representative of the general population. According to a 2014 report by the Pew Research Center, users of social media in the United States are predominantly aged between 18 and 34 years [
Finally, we faced some difficulties in obtaining records of log-ins. Depending on the WordPress updates (the Web software used to create the blog), we noticed that statistics were erroneous for a small number of participants whose log-in statistics were not recorded for some posts, although they had commented on these same posts. This is why we chose to focus on the number of comments and the number of page views to assess participation rates, and not the number of log-ins, which was a less reliable statistic.
This feasibility study provided valuable information about how to optimize and implement a healthy eating blog intervention. It also contributed to inform the conduct of a future definitive RCT to assess the efficacy of a healthy eating blog to improve dietary and eating behaviors. Because preestablished feasibility criteria were met, characteristics of the intervention to be used in the RCT will remain unchanged. Given the novelty of this intervention, this project paves the way for designing and evaluating the effects of other social media tools on consumers’ health outcomes. This study was the first to provide an empirically supported basis for the design of interventions using social media, more specifically blogs, to improve the quality of health promotion and disease prevention health care services through enhanced bidirectional exchange of evidence-based and experiential nutrition-related knowledge.
To date, there is very modest evidence that interventions using online social media may be effective. Because the Web is used more and more and this field of research is very recent [
body mass index
food frequency questionnaire
Institute of Nutrition and Functional Foods
randomized controlled trial
We would like to thank all participants for their time and dedication, the Direction des technologies de l’information de l’Université Laval for developing the blog platform on Wordpress and the Service de Consultation en Statistiques de l’Université Laval for their help with statistical analysis.
MEC wrote a first draft of this paper. SD, VBM, VP, MPG, and SS designed the study. MEC and VBM developed the blog content under the supervision of SD and VP, VBM kept a weekly record of log-in statistics, and MEC answered participants’ questions and comments. All authors reviewed the manuscript.
None declared.