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 the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Recent studies have shown the potential of Web-based interventions for changing dietary and physical activity (PA) behavior. However, the pathways of these changes are not clear. In addition, nonusage poses a threat to these interventions. Little is known of characteristics of participants that predict usage.
In this study we investigated the users and effect of the Healthy Weight Assistant (HWA), a Web-based intervention aimed at healthy dietary and PA behavior. We investigated the value of a proposed framework (including social and economic factors, condition-related factors, patient-related factors, reasons for use, and satisfaction) to predict which participants are users and which participants are nonusers. Additionally, we investigated the effectiveness of the HWA on the primary outcomes, self-reported dietary and physical activity behavior.
Our design was a two-armed randomized controlled trial that compared the HWA with a waiting list control condition. A total of 150 participants were allocated to the waiting list group, and 147 participants were allocated to the intervention group. Online questionnaires were filled out before the intervention period started and after the intervention period of 12 weeks. After the intervention period, respondents in the waiting list group could use the intervention. Objective usage data was obtained from the application itself.
In the intervention group, 64% (81/147) of respondents used the HWA at least once and were categorized as “users.” Of these, 49% (40/81) used the application only once. Increased age and not having a chronic condition increased the odds of having used the HWA (age: beta
Respondents did not use the application as intended. From the proposed framework, a social and economic factor (age) and a condition-related factor (chronic condition) predicted usage. Moreover, users were healthier and more knowledgeable about healthy behavior than nonusers. We found no apparent effects of the intervention, although exploratory analyses showed that choosing to use or not to use the intervention led to different outcomes. Combined with the differences between groups at baseline, this seems to imply that these groups are truly different and should be treated as separate entities.
Trial ID number: ISRCTN42687923; http://www.controlled-trials.com/ISRCTN42687923/ (Archived by WebCite at http://www.webcitation.org/5xnGmvQ9Y)
The increasing prevalence of overweight is a problem in modern society. It is closely related to a number of chronic conditions such as type 2 diabetes mellitus and places a great burden on the health care system. Losing weight and especially preventing weight regain is challenging. It might be more cost-efficient to prevent people from becoming overweight by focusing on healthy dietary and physical activity (PA) behavior [
The problem of attrition [
We incorporated the WHO framework and behavior theories in a study of use of the Healthy Weight Assistant (HWA), a Web-based lifestyle intervention. We considered the influence of social and economic factors (demographics), condition-related factors (ie, general practitioner [GP] visits, having a chronic condition, and self-reported and self-rated dietary and PA behavior), patient-related factors or constructs identified by behavior change theories (ie, knowledge, attitude, and self-efficacy) [
Additionally in this study, we assessed the effectiveness of the intervention using self-reported dietary and PA behavior as primary outcome measures because the intervention was aimed at improving health behavior. We included secondary outcome measures that are known determinants of behavior change. We also chose to include measures of knowledge, attitude, and self-efficacy [
Consequently, our research questions were: What characteristics of participants are related to the use of the HWA intervention? What effects does the HWA intervention have on the primary and secondary outcome measures?
Participants were recruited through advertisements about an online lifestyle intervention in local newspapers, supermarkets, and on health-related websites. Permission of an ethics review board for the study was not required because, according to the Dutch law, nonintrusive interventions conducted with healthy adults do not require approval from an ethics board. In total, 297 respondents expressed interest in using an online lifestyle intervention and satisfied our inclusion criteria (body mass index [BMI] 18.5 - 28.0 kg/m2, Dutch-speaking). The inclusion criterion for BMI was chosen to reflect the target group of the intervention under investigation. The sample used in this study was a self-selected convenience sample. Enrollment took place beginning November 1, 2008, and ending December 31, 2008. All participants were randomly assigned to either the Web-based lifestyle coach or a waiting list. A total of 150 participants were allocated to the waiting list group, and 147 participants were allocated to the intervention group. Participants filled out online questionnaires before the 12-week intervention period started and again after the intervention period ended. The posttest questionnaire was available for all respondents for a period of 3 weeks beginning February 27 and ending April 16. After the intervention period, respondents in the waiting list group could use the intervention. The flowchart of the study can be found in
Flowchart of the study
Randomization took place 1 week before the start of the intervention period. We used block randomization with blocks of 4 participants, stratified on age, sex, and education. The randomization scheme was created by a computer application and carried out by a member of the research team. Participants who filled out demographic information were randomized. Only respondents who completed the pretest questionnaire were included; therefore, 28 respondents were excluded. Participants were not blinded to randomization outcome but received an email with information on when and how they were able to access the Healthy Weight Assistant (HWA) after filling out the pretest questionnaire.
The Healthy Weight Assistant (HWA) is a Web-based lifestyle intervention developed by the Netherlands Nutrition Centre, which is a government-funded organization focusing on increasing the knowledge of consumers about the quality of food and encouraging consumers to eat healthily and safely. The goal of the HWA is to support people with a healthy weight and people who are slightly overweight (ie, BMI 18.5-28.0 kg/m2) to maintain and achieve a healthy weight. The aim is not to achieve a given weight loss, but to support the achievement of healthy dietary and PA behavior. Therefore, the focus was broader than only energy balance-related behavior. The target group was selected by the Netherlands Nutrition Centre according to their BMI classification. The theoretical basis for behavior change via the HWA is the transtheoretical model [
The Healthy Weight Assistant
The HWA consists of 4 steps, which are marked in the application by a “to-do list” and tabs in the “diary” (
We made use of a waiting list control group. Participants randomized in this group received an email newsletter every 3 weeks, but no access to the HWA during the intervention period. The newsletter contained general information about the study and about the University of Twente. Furthermore, it contained leisure tips, but it contained no information on healthy lifestyle. After the intervention period, participants in the waiting list group received access to the HWA. Participants in the intervention group also received the newsletter every 3 weeks.
Online questionnaires were used to assess pretest and posttest values. Education was self-reported and recoded into the following three categories: low (primary and lower vocational education), moderate (secondary and middle vocational education), and high (higher vocational and university education). BMI (kg/m2) was calculated using self-reported weight and length. Dietary behavior was measured using a 14-item self-report questionnaire of the Netherlands Nutrition Centre, based on the Netherlands classification model [
In addition to the online questionnaires, the HWA stored every log-on by a participant. These log files were used to attain the usage of the HWA, that is, the number of times each respondent logged on to the HWA within the intervention period.
SurveyMonkey was used for the electronic data collection [
Our format of data collection was an “open survey” [
We were not able to record unique site visitors or survey visitors. The completion rate was 90% (269/297). To prevent multiple entries from the same person we used cookies that were stored when visiting the first page and were valid for 14 days. Also, we checked IP addresses. Entries from the same address with identical sex and birth date were checked for completeness. The most complete entry was saved, or, in case of equal completeness, the first entry was saved.
Previous research on the HWA using the same research instrument on self-reported dietary behavior yielded information on the mean and standard deviation of this primary outcome measure (mean 62.9, SD 8.43) [
Statistical analyses were performed using SPSS Statistics 17.0 (IBM Corporation, Somers, NY, USA). We used the multiple imputation (MI) feature of SPSS Statistics 17.0 to handle missing data of posttest nonrespondents. Demographic variables and baseline outcome measures were used as predictors in the imputation model. We used an iterative Markov chain Monte Carlo method, which is the fully conditional specification. In addition, five imputed datasets were generated on which the effectiveness analyses were performed. When possible, pooled outcomes were used for the analyses; otherwise, the five estimates were combined into a single overall estimate following the MI inference rules of Rubin [
Differences between users and nonusers within the intervention group were assessed using Pearson's chi-square and analysis of variance testing. Furthermore, regression analysis was used to see whether characteristics predicted use of the intervention.
Effectiveness of the intervention was assessed by intention-to-treat (ITT) using effect sizes and odds ratios. Additionally, exploratory analyses were performed on pretest and posttest scores of all participants combined and separately for the control group, the users, and the nonusers of the intervention using regression analyses and effect sizes. All reported
Of the 269 enrolled respondents (those who completed the pretest questionnaire), 159 respondents filled out the posttest questionnaire (response rate = 59%, 159/269). The response was significantly lower in the intervention group (51%, 65/127) than in the control group (66%, 94/142) (
Baseline differences on outcome variables between responders and dropouts
Variable | Responders |
Dropouts |
|
|
BMI (kg/m2), mean (SD) | 24.0 (2.5) | 23.9 (2.5) | .83 | |
|
.18 | |||
Healthy | 48 (30) | 26 (24) | ||
Improvable | 99 (62) | 69 (63) | ||
Unhealthy | 12 (8) | 15 (14) | ||
Healthy PA, n (%) | 64 (42) | 41 (37) | .46 | |
Knowledge, mean (SD)a | 7.9 (1.1) | 7.7 (1.2) | .19 | |
Attitude, mean (SD)b | 4.1 (0.4) | 3.9 (0.5) | .001 | |
Self-efficacy, mean (SD)c | 2.1 (0.6) | 2.2 (0.6) | .55 | |
Self-rating, mean (SD)d | 6.8 (1.1) | 6.4 (1.5) | .02 | |
Realistic insight, diet, n (%) | 92 (60) | 69 (63) | .35 | |
Realistic insight, PA, n (%) | 88 (58) | 70 (64) | .60 |
a Scale from 1 (very poor) to 10 (excellent)
b Scale from 1 (very unfavorable) to 5 (very favorable)
c Scale from 1 (very high) to 5 (very low)
d Scale from 1 (very poor) to 10 (excellent)
As shown in
Baseline demographics and reasons for use
Variable | Total |
Intervention |
Control |
|
|
Age (years), mean (SD) | 41.5 (13.5) | 41.2 (13.5) | 41.7 (13.6) | .73 | |
Sex, n female (%) | 177 (66) | 85 (67) | 92 (65) | .80 | |
|
.71 | ||||
High, n (%) | 143 (53) | 69 (54) | 74 (52) | ||
Moderate, n (%) | 87 (32) | 42 (33) | 45 (32) | ||
Low, n (%) | 39 (15) | 16 (13) | 23 (16) | ||
Chronic condition, n (%) | 48 (18) | 19 (15) | 29 (20) | .27 | |
|
|||||
Insight into lifestyle, n (%) | 161 (60) | 80 (63) | 81 (57) | .38 | |
Living healthier, n (%) | 120 (45) | 61 (48) | 59 (42) | .33 | |
Fun, n (%) | 112 (42) | 55 (43) | 57 (40) | .62 | |
Lose weight, n (%) | 107 (40) | 56 (44) | 51 (36) | .21 |
a Multiple answers possible so cumulative percentages do not equal 100%
Respondents in the waiting list (control) condition reported to have opened a mean of 3.4 (SD 1.2) out of 5 newsletters. From the log files of the HWA, we know that 81 of the 127 (64%) respondents in the intervention group used the HWA at least once, while 49% (40/81) of these used the application only once. The respondent that used the HWA most frequently used it 13 times during the intervention period of 12 weeks. The median number of times HWA was used was 1.0. Of the 127 respondents in the intervention group, 4 (3%) used the application at least the intended number of times within the intervention period (ie, once a fortnight or 6 times during the 12-week period). Satisfaction with the application was assessed within the posttest questionnaire. We used only the data provided by 50 respondents who filled out the posttest questionnaire and who had accessed the HWA at least once in the intervention period. These results are depicted in
Satisfaction with the Healthy Weight Assistant (n = 50)
Item | Mean (SD) | Disagree, n (%) | Neutral, n (%) | Agree, n (%) |
Easy to use | 3.3 (0.83) | 8 (16) | 22 (44) | 20 (40) |
Useful | 2.9 (0.87) | 13 (26) | 25 (50) | 12 (24) |
Recommend to others | 3.0 (0.90) | 12 (24) | 22 (44) | 16 (32) |
Keep using | 2.7 (0.89) | 20 (40) | 22 (44) | 8 (16) |
Baseline differences between respondents in the intervention group who used the application (users) and the respondents in this group who did not use the HWA at least once (nonusers) are depicted in
Baseline differences between users and nonusers in the intervention group
Variable | Users (n=81) | Nonusers (n=46) | F or χ2 |
|
|
Age (years), mean (SD) | 42.6 (13.2) | 38.8 (13.8) | F1,125= 2.307 | .13 | |
Sex, n female (%) | 58 (72) | 27 (59) | χ2 1 = 2.2 | .17 | |
|
χ2 2 = 0.7 | .70 | |||
High, n (%) | 46 (57) | 23 (50) | |||
Moderate, n (%) | 26 (32) | 16 (35) | |||
Low, n (%) | 9 (11) | 7 (15) | |||
Chronic condition, n (%) | 8 (10) | 11 (24) | χ2 1 = 4.5 | .04 | |
BMI (kg/m2), mean (SD) | 24.2 (2.5) | 23.7 (2.3) | F1,125= 0.900 | .35 | |
|
χ2 = 8.4 | .015 | |||
Healthy, n (%) | 28 (35) | 6 (13) | |||
Improvable, n (%) | 46 (57) | 31 (67) | |||
Unhealthy, n (%) | 7 (9) | 9 (20) | |||
Healthy physical activity level, n (%) | 28 (37) | 19 (41) | χ2 1 = 0.2 | .70 | |
Knowledge, mean (SD) | 7.9 (1.1) | 7.4 (1.4) | F1,125 = 4.194 | .04 | |
Attitude, mean (SD) | 4.0 (0.4) | 3.9 (0.5) | F1,125 = 2.665 | .11 | |
Self-efficacy, mean (SD) | 2.3 (0.6) | 2.2 (0.6) | F1,125 = 0.274 | .60 | |
Self-rating, mean (SD) | 6.6 (1.4) | 6.5 (1.5) | F1,125= 0.037 | .85 | |
|
χ2 2 = 8.2 | .02 | |||
Underestimation, n (%) | 17 (21) | 2 (4) | |||
Realistic insight, n (%) | 52 (64) | 31 (67) | |||
|
χ2 2 = 2.1 | .36 | |||
Underestimation, n (%) | 1 (1) | 1 (2) | |||
Realistic insight, n (%) | 47 (58) | 32 (70) |
Overall, at baseline, users were healthier (scored better on dietary behavior and had a chronic condition less often) and were more knowledgeable about healthy behavior. Furthermore, users seemed to underestimate their behavior more often than nonusers, and nonusers seemed to overestimate their behavior more often than users.
To assess whether variables of the framework proposed in the introduction could be used to predict if respondents were going to use the HWA, we performed an exploratory logistic regression using the factors from the framework (social and economic, condition-related, patient-related or constructs from behavior change theories, and reasons for use). Results of this logistic regression (
Logistic regression model to predict usage of the HWA
Included | Coefficient B |
|
Odds Ratio (OR) |
|
Constant | -12.63 (4.013) | .002 | ||
Factor | Variable | |||
Social and economic | Age | 0.04 (0.018) | .02 | 1.04 (1.00 - 1.08) |
Internet use | 0.18 (0.131) | .17 | 1.20 (0.93 - 1.55) | |
Sex | 0.50 (0.504) | .32 | 1.65 (0.62 - 4.44) | |
Education | 0.13 (0.353) | .71 | 1.14 (0.57 - 2.28) | |
Condition-related | Self-rating | −0.35 (0.379) | .36 | 0.71 (0.34 - 1.49) |
GP visits | 1.19 (0.647) | .07 | 3.30 (0.93 - 11.72) | |
Chronic condition | 2.24 (0.749) | .003 | 9.40 (2.17 - 40.82) | |
Diet | 0.71 (0.688) | .31 | 2.03 (0.53 - 7.80) | |
PA | 0.80 (0.948) | .40 | 2.22 (0.35 - 14.26) | |
Insight, diet | 0.56 (0.667) | .40 | 1.76 (0.48 - 6.48) | |
Insight, PA | −1.00 (0.818) | .22 | 0.37 (0.07 - 1.83) | |
Patient-related | Knowledge | 0.03 (0.213) | .91 | 1.03 (0.68 - 1.56) |
Attitude | 0.57 (0.681) | .41 | 1.76 (0.46 - 6.69) | |
Self-efficacy | 0.26 (0.458) | .57 | 1.30 (0.53 - 3.18) | |
Reasons for use | Insight into lifestyle | 0.47 (0.531) | .37 | 1.60 (0.57 - 4.55) |
Live healthier | −0.03 (0.281) | .93 | 0.98 (0.56 - 1.69) | |
Fun | 0.13 (0.165) | .44 | 1.14 (0.82 - 1.57) | |
Lose weight | 0.16 (0.122) | .18 | 1.18 (0.93 - 1.50) |
Furthermore, we performed a linear regression to investigate whether satisfaction with the intervention HWA predicted the number of logins (
Linear regression on satisfaction predicting number of log-ins to the Healthy Weight Assistant
B (SE) | Beta | |
Constant | −2.61 (1.17) | |
Satisfaction | 0.70 (0.38) | 0.23a |
a
In addition, ITT analyses were performed on all outcome variables (
Complementary to the ITT analyses, we performed analyses comparing the differences of the control group with the differences of the users (results not shown). These analyses did not yield any significant effects and were comparable to the results of the ITT analyses, although the effect sizes were generally larger.
Intention-to-treat (ITT) analyses
Variable | Intervention (n=127) | Control (n=142) | Effect Sizea (ES) or OR (95% CI) | |||
Pretest | Posttest | Pretest | Posttest | |||
BMI, mean (SD) | 24.0 (2.4) | 24.1 (2.5) | 23.9 (2.5) | 24.0 (2.5) | ES: 0.07 (-0.10 – 0.24) | |
|
OR: 0.84 (0.44 – 1.58) | |||||
Healthy, n (%) | 34 (27) | 45 (35) | 40 (28) | 46 (32) | ||
Improvable, n (%) | 77 (61) | 73 (58) | 91 (64) | 89 (63) | ||
Unhealthy, n (%) | 16 (13) | 9 (7) | 11 (8) | 7 (5) | ||
Healthy PA, n (%) | 49 (38.6) | 58 (46) | 58 (41) | 69 (49) | OR: 1.10 (0.60 – 2.01) | |
Knowledge, mean (SD) | 7.7 (1.2) | 7.7 (1.3) | 7.9 (1.1) | 7.7 (1.3) | ES: 0.15 (−0.13 to 0.42) | |
Attitude, mean (SD) | 4.00 (0.45) | 4.03 (0.45) | 4.01 (0.44) | 4.02 (0.45) | ES: 0.08 (0.00 – 0.16) | |
Self-efficacy, mean (SD) | 2.2 (0.61) | 2.3 (0.70) | 2.1 (0.59) | 2.2 (0.64) | ES: 0.04 (−0.01 to 0.17) | |
Self-rating, mean (SD) | 6.5 (1.4) | 6.9 (1.2) | 6.8 (1.2) | 6.9 (1.2) | ES: 0.18 (−0.04 to 0.40) | |
Realistic insight, diet, n (%) | 83 (65) | 71 (56) | 83 (59) | 87 (61) | OR: 0.74 (0.35 – 1.56) | |
Realistic insight, PA, n (%) | 79 (62) | 83 (65) | 84 (59) | 88 (62) | OR: 0.78 (0.35 – 1.74) |
aEffect size for ratio variables presented as Cohen’s d, that is, the number of standard deviations the intervention group (I) improved more than the control group (C) (mean improvement I – mean improvement C)/pooled SD of improvement. Effect size for ordinal variables is presented as the odds ratio.
For the group as whole (independent of randomized condition), there were significant differences between pretest and posttest scores. With respect to diet (effect size
Pretest and posttest values on outcome variables for control group, nonusers, and users
Variable | Control (n=142) | Nonusers (n=46) | Users (n=81) | ||||
Pretest | Posttest | Pretest | Posttest | Pretest | Posttest | ||
BMI, mean (SD) | 23.9 (2.5) | 24.0 (2.5) | 23.7 (2.3) | 23.9 (2.5) | 24.2 (2.5) | 24.2 (2.5) | |
|
|||||||
Healthy, n (%) | 40 (28) | 46 (32) | 6 (13) | 11 (24) | 28 (35) | 34 (42) | |
Improvable, n (%) | 91 (64) | 89 (63) | 31 (68) | 30 (65) | 46 (57) | 43 (53) | |
Unhealthy, n (%) | 11 (8) | 7 (5) | 9 (20) | 5 (11) | 7 (9) | 4 (5) | |
Healthy pysical activity level, n (%) | 58 (41) | 69 (49) | 19 (41) | 16 (35) | 30 (37) | 42 (52) | |
Knowledge, mean (SD) | 7.9 (1.1) | 7.7 (1.3) | 7.4 (1.4) | 7.3 (1.4) | 7.9 (1.1) | 7.9 (1.2) | |
Attitude, mean (SD) | 4.0 (0.44) | 4.0 (0.45) | 3.9 (0.46) | 4.0 (0.45) | 4.0 (0.44) | 4.0 (0.44) | |
Self-efficacy, mean (SD) | 2.1 (0.59) | 2.2 (0.64) | 2.2 (0.62) | 2.4 (0.77) | 2.3 (0.61) | 2.3 (0.65) | |
Self-rating, mean (SD) | 6.8 (1.2) | 6.9 (1.2) | 6.5 (1.5) | 6.9 (1.4) | 6.6 (1.4) | 6.9 (1.1) | |
Realistic insight, diet, n (%) | 83 (59) | 87 (61) | 31 (67.4) | 25 (54.3) | 52 (64.2) | 46 (56.8) | |
Realistic insight, PA, n (%) | 84 (59) | 88 (62) | 32 (69.6) | 27 (58.7) | 47 (58.0) | 56 (69.1) |
Effect size (ES) of the differences between pretest and posttest values on outcome variables for control group, nonusers, and users
Variable | Control (n = 142) | Nonusers (n = 46) | Users (n = 81) | |||
ESa |
|
ES a |
|
ES a |
|
|
BMI | 0.02b | CI: −0.39 to 0.44 | 0.06a | CI: -0.64 – 0.77 | 0.03a | CI: -0.51 to 0.57 |
Diet | −0.09c |
|
−0.23b |
|
−0.13b |
|
PA | −0.10c |
|
−0.07b | Z = −0.71 (.48) | −0.17b |
|
Knowledge | −0.15b | CI: −0.35 to 0.04 | −0.08a | CI: −0.49 to 0.34 | 0.04a | CI: −0.20 to 0.29 |
Attitude | 0.01b | CI: −0.06 to 0.08 | 0.28a | CI: 0.15 – 0.41 | −0.05a | CI: −0.15 to 0.05 |
Self-efficacy | 0.14b | CI: 0.03 – 0.24 | 0.33a | CI: 0.13 – 0.53 | 0.05a | CI: −0.09 to 0.19 |
Self-rating | 0.15b | CI: −0.05 to 0.35 | 0.25a | CI: −0.18 to 0.68 | 0.27a | CI: 0.00 – 0.54 |
Insight, diet | −0.03c |
|
−0.13b |
|
−0.07b |
|
Insight, PA | −0.01c |
|
−0.13b |
|
−0.11b |
|
a Effect sizes for ratio variables are presented as Cohen’s d, while effect sizes for ordinal variables are presented as
b Effect size (ES) presented as Cohen’s d: (meanpost - meanpre)/SDpooled
c Effect size presented as
d Wilcoxon signed-rank test
e In this column the reliability of the effect size is presented as the confidence interval for Cohen’s d for ratio variables and as
The results showed that the HWA was not used as often as intended. Increased age and not having a chronic condition increased the odds of having used the application at least once. Moreover, users were healthier and more knowledgeable about healthy behavior than nonusers. The ITT analyses showed no apparent effects of the intervention; however, there were differences in the effect of the intervention on users and nonusers. With respect to dietary behavior and attitude, nonusers improved more than users, while with respect to physical activity and self-rated behavior the users improved more than nonusers. On self-efficacy, the control group and the nonusers showed deterioration from baseline to posttest.
Only 64% (81 out of 127) of the participants who received access to the HWA actually used the application. This finding is not unique to this study; for example, see [
Additionally, users more often underestimated their dietary behavior (respondents who did meet the criteria for healthy behavior but who rated their behaviour as unhealthy were classified as underestimators), while nonusers more often overestimated their behavior. This shows that the people who could have benefited most from the HWA were less likely to use the application. Of the patient-related factors or constructs from behavior change theories, only knowledge showed a significant difference between users and nonusers. Users knew more about healthy behavior, which supports the notion that the people who could have benefited most from the HWA were least likely to use the application.
There were no differences related to the reasons for use between users and nonusers, and the different reasons do not explain whether respondents used the HWA or not. However, the reasons for use might play a role in the frequency of use. The most frequently mentioned reason for wanting to use the intervention was to gain insight into one’s own behavior (60%). It might be that this goal was reached after using the HWA once, and participants might not have felt the need to use the HWA again.
Interestingly, the intervention was specifically not made to help people lose weight, but this goal was mentioned by 40% of respondents. Respondents seemed to want a quick and short-term effect (to gain insight) and might not have been willing to use the intervention frequently to work on a long-term goal (eg, a healthier lifestyle). Satisfaction with the HWA was not associated with the frequency of use. However, overall, participants were not very satisfied with the HWA, which might have contributed to the relative low usage rates. To summarize, one of the social and economic factors (ie, age), condition related factors (ie, chronic condition, self-reported behavior, and insight into behavior), and one of the patient-related factors (ie, knowledge) were related to use of the system. Satisfaction and reasons for use provided more in-depth information related to the causes of the lack of adherence to the intervention.
At baseline, the intervention and control groups showed a significant difference in attitude. The absolute value of the difference was small, however, and we don’t consider it to be a meaningful difference. Therefore, we can argue that the groups were comparable at baseline. We found no meaningful significant effects of the intervention using ITT analyses. We did find that both the waiting list group and the intervention group showed significant improvement on behavior and a significantly more favorable self-rated behavior. This well-known Hawthorne effect [
In this study, we were faced with substantial dropout and nonusage rates. High dropout rates are not uncommon in this field of research and have been said to be a major challenge [
Our results showed that the users of the HWA were healthier than nonusers, which is an unfortunate finding not unique to this study [
In this study, the frameworks used to predict usage and to study effectiveness seem to have been insufficient. From the WHO framework [
A limitation of this study is the use of self-reported behaviors. Although we used questionnaires used in previous studies, there is a chance of biased results due to social desirability or lack of insight into behavior. As a consequence, a possible change in insight into behavior might not be reflected in our results. It could be that at baseline, participants provided optimistic self-reported behavior. Due to the intervention, the users might have provided more realistic self-reported behavior at posttest. Unfortunately, this potentially positive effect of the HWA could not be tested in this study. A second limitation is related to the participants in this study. Most respondents were female and highly educated. Various studies have reported overrepresentation of this group [
Usage is a major issue in research into the effects of eHealth applications. More research is needed into transforming potential users into actual users and into keeping them engaged with the application and, thereby, stimulating them to keep using the intervention. Moreover, long-term research on the use of eHealth applications is needed to provide insight into the way usage fluctuates over time. From this study, we have gained insight into differences between users and nonusers, which can be seen as a first step to decreasing attrition. The next step might be found when looking at the opportunities technology has to offer. For example, several recent studies have shown beneficial effects of adding mobile technology [
We are grateful for funding from the Netherlands Nutrition Centre.
None declared
Translated questionnaire dietary behavior
Translated questionnaire physical activity behavior
body mass index
confidence interval
effect size
general practitioner
Healthy Weight Assistant
intention-to-treat
odds ratio
physical activity
standard error
World Health Organization