JMIR J Med Internet Res Journal of Medical Internet Research 1438-8871 JMIR Publications Inc. Toronto, Canada v17i5e115 25963607 10.2196/jmir.3932 Original Paper Original Paper Who Follows eHealth Interventions as Recommended? A Study of Participants' Personal Characteristics From the Experimental Arm of a Randomized Controlled Trial Eysenbach Gunther Meadel Cecile Guertler Diana Reinwand Dominique A MSc 1
CAPHRI School for Public Health and Primary Care Department of Health Promotion Maastricht University P Debyeplein 1 Maastricht, 6229 HA Netherlands 31 433882435 31 433671032 d.reinwand@maastrichtuniversity.nl
2 http://orcid.org/0000-0002-1567-1005 Schulz Daniela N PhD 1 http://orcid.org/0000-0002-5126-9320 Crutzen Rik PhD 1 http://orcid.org/0000-0002-3731-6610 Kremers Stef PJ PhD 3 http://orcid.org/0000-0002-7406-2779 de Vries Hein PhD 1 http://orcid.org/0000-0002-3640-2517 1 CAPHRI School for Public Health and Primary Care Department of Health Promotion Maastricht University Maastricht Netherlands 2 Jacobs Center for Lifelong Learning and Institutional Development Jacobs University Bremen Germany 3 NUTRIM School for Nutrition, Toxicology and Metabolism Department of Health Promotion Maastricht University Maastricht Netherlands Corresponding Author: Dominique A Reinwand d.reinwand@maastrichtuniversity.nl 05 2015 11 05 2015 17 5 e115 08 10 2014 06 11 2014 10 12 2014 14 01 2015 ©Dominique A Reinwand, Daniela N Schulz, Rik Crutzen, Stef PJ Kremers, Hein de Vries. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 11.05.2015. 2015

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.

 Background

Computer-tailored eHealth interventions to improve health behavior have been demonstrated to be effective and cost-effective if they are used as recommended. However, different subgroups may use the Internet differently, which might also affect intervention use and effectiveness. To date, there is little research available depicting whether adherence to intervention recommendations differs according to personal characteristics.

 Objective

The aim was to assess which personal characteristics are associated with using an eHealth intervention as recommended.

 Methods

A randomized controlled trial was conducted among a sample of the adult Dutch population (N=1638) testing an intervention aimed at improving 5 healthy lifestyle behaviors: increasing fruit and vegetable consumption, increasing physical activity, reducing alcohol intake, and promoting smoking cessation. Participants were asked to participate in those specific online modules for which they did not meet the national guideline(s) for the respective behavior(s). Participants who started with fewer than the recommended number of modules of the intervention were defined as users who did not follow the intervention recommendation.

 Results

The fewer modules recommended to participants, the better participants adhered to the intervention modules. Following the intervention recommendation increased when participants were older (χ2 1=39.8, P<.001), female (χ2 1=15.8, P<.001), unemployed (χ2 1=7.9, P=.003), ill (χ2 1=4.5, P=.02), or in a relationship (χ2 1=7.8, P=.003). No significant relevant differences were found between groups with different levels of education, incomes, or quality of life.

 Conclusion

Our findings indicate that eHealth interventions were used differently by subgroups. The more frequent as-recommended intervention use by unemployed, older, and ill participants may be an indication that these eHealth interventions are attractive to people with a greater need for health care information. Further research is necessary to make intervention use more attractive for people with unhealthy lifestyle patterns.

 eHealth Web-based intervention intervention use computer tailoring personal characteristics health lifestyle multiple health behaviors intervention adherence socioeconomic status Introduction

New eHealth interventions are an important tool to improve public health by providing people with information, skills, and support needed for a positive health-related lifestyle change [1,2]. These eHealth interventions provide the opportunity to use computer tailoring to provide highly personalized information to a respondent without face-to-face counseling [3-5]. With the use of computer tailoring, participants receive information derived from an individual assessment, attuned to their individual answers, and, in our case, aimed at motivating individuals to adopt 1 or more healthy behaviors [6]. Consequently, computer-tailored interventions provide feedback that is more relevant to the individual, contains less redundant information, and is more likely to be processed and remembered than generic information [7-10]. Furthermore, eHealth interventions are easily accessible and have the potential to reach a wide population [11,12].

Numerous studies have demonstrated that computer-tailored interventions are effective in motivating individuals to adopt health behaviors [1,13-15], including increased physical activity [16-20], healthy nutrition [5,20-24], smoking prevention and cessation [25-28], and decreasing alcohol intake [29-31]. Moreover, interventions to change multiple health behaviors have also been shown to be effective [32-35]. Studies also showed that computer-tailored interventions are more cost-effective than typical health care [36-38].

People of a low socioeconomic status (SES) often have unhealthy lifestyle behaviors [39,40]. They often eat fewer fruits and vegetables [41], are less physically active (eg, [42]), consume more alcohol (eg, [43]), and smoke more tobacco (eg, [44]) compared to people with higher SES.

Although there is evidence that Web-based interventions are effective in improving health behavior, these interventions come with high dropout rates and the problem that participants often do not use the intervention as recommended [45-47]. Although studies have investigated characteristics of dropout and nonusage of eHealth interventions (eg, [45,48]), it is equally important to know more about the participants who use these interventions as recommended. It is conceivable that people with different sociodemographic profiles may use these Web-based interventions differently because of general differences in online behavior and Internet usage between certain groups. This digital divide, for example, refers to unequal access to and use of the Internet among people with a lower SES [49-53].

More than 90% of the general Dutch population has access to the Internet and 86% of Dutch people use the Internet every day [54,55]. Although the gap between people with and without Internet access seems to be closing, there is still a difference in Internet use between certain sociodemographic groups [56]. People with a higher SES use the Internet more often than people with a lower SES to achieve personal development (eg, getting a new job), whereas people with a lower SES use the Internet primarily for other purposes, such as entertainment [57,58].

Education is often used as a proxy to measure SES; therefore, the literature about education and Internet use is extensive. People with a higher educational level have been found to use the Internet more frequently to gain health-related information, for work, and for shopping or product information. People with a lower educational level use the Internet more often with other objectives, such as browsing the Web or playing online games [59-62]. Educational level might also play a role in online behavior because most information on the Internet is written at a high literacy level whereas nearly half of the people do not understand this level of written information [63]. Moreover, employed people spend less of their leisure time online [57]. People with a lower income use the Internet more often for entertainment purposes, such as downloading music [64-66], whereas people with a higher income spend more of their time online searching for news or information [57].

Age- and gender-based differences in Internet use also exist. Although the majority (80%) of Dutch people aged 65 years or older have access to the Internet [67], they are less familiar with routine daily use of the Internet [56,68] and use the Internet primarily as an information source [69] in contrast to younger Internet users who primarily understand the Internet as an entertainment medium [70]. Males have been found to use the Internet more often, are more experienced with Internet use, and feel more comfortable with it [56,71]. This is in contrast to women, who spend less time online when having to take care of their family [72], but seek health information online more often than men [64]. In addition to these, other personal factors, such as a lower quality of life [73,74] or being married, are related to less Internet usage [75].

Because SES is an important predictor of how people use the Internet [56], it is conceivable that people with a lower SES may not implement eHealth interventions as intended by the intervention developer and may be unlikely to follow intervention recommendations, which makes behavior change less likely [76]. However, because people with a lower SES are a high-risk group for unhealthy behaviors [44,77-79], they are a highly relevant target group that might benefit from eHealth behavior change interventions. The same reasoning might hold for other personal characteristics, such as age, perceived health, or quality of life.

Therefore, this study investigates whether people are following the recommendations of how to use eHealth interventions. The purpose of our study is to identify personal and socioeconomic characteristics associated with recommended eHealth intervention use. Based on findings from the literature, we hypothesize that people with a higher education, who do not have paid work, those who have a lower income, who are younger, female, have a high perceived quality of life, and are not in a relationship are more likely to use the intervention as recommended.

 Methods Overview

A detailed description of the study protocol has been published elsewhere [80]; only those study methodology details relevant to the study at hand are described here.

 Participants, Procedure, Study Design, and Intervention Content

This study is part of a randomized controlled trial that was conducted in the Netherlands between 2009 and 2012. The study received ethics approval from the Medical Ethics Committee of Maastricht University and the University Hospital Maastricht (MEC) and has been registered by the Dutch Trial Register (NTR 2168). Participants were recruited through different Dutch Regional Health Authorities (RHAs) in the Netherlands [80]. These RHAs periodically monitor the health status, health behaviors, and related aspects of the adult population. At the end of this monitor, people were asked if they were interested in participating in this study. They were told that they would be invited to take part in a free online program that provides participants with tailored feedback about their health behavior. Internet access, a computer, basic Internet skills, and sufficient Dutch language skills were required preconditions for participating. The intervention consisted of 2 parts and focused on 5 health behaviors: fruit consumption, vegetable consumption, physical activity, smoking behavior, and alcohol intake. During the first part of the intervention, participants had to answer questionnaires about their health behaviors. The answers were used to provide participants with their personalized risk appraisal, which provided feedback by comparing the respondents’ behavior to the Dutch guidelines defined for the 5 behaviors, such as (1) being physically active for at least 30 minutes on at least 5 days a week, (2) eating at least 200 grams of vegetables, (3) eating at least 2 pieces of fruit each day, (4) drinking no more than 2 glasses of alcohol a day (for men; 1 glass for women), and (5) not smoking at all.

The second part of the intervention consisted of 5 lifestyle modules. Participants who were interested in participation in the program received an email with their personalized link to log on to the computer-tailoring program. Based on the first part, the questionnaire assessment as part of the RHA monitor, participants were provided with tailored feedback concerning their behavior. They received an overview about all 5 behaviors and whether they met the guidelines or not. In the second part of the program, participants were asked to complete all modules for which they did not meet the guidelines. For example, in case a participant reported smoking and eating less than 2 pieces of fruit a day, he/she was advised to participate in the modules for smoking and fruit consumption. All modules included tailored feedback based on the determinants specified in the I-Change Model [81]: attitudes, social influence, self-efficacy, and preparatory and coping planning. The order of the modules was counterbalanced, either starting with preventive behaviors and addiction behaviors (ie, physical activity/vegetable consumption/fruit consumption and alcohol intake/smoking), or vice versa (alcohol intake/smoking followed by physical activity/vegetable consumption/fruit consumption).

 Measures Demographic Information

The following demographic information were assessed: age, gender (1=male; 2=female), education (1=low: no education, primary, or lower vocational school; 2=middle: secondary vocational school or high school; 3=high: higher professional education or university), monthly income (1≤€1751; 2=€1751-€3050; 3≥€3050), work situation (1=no paid job; 2=paid job) [82], family status (1=single; 2=relationship), number of persons living in the household, and country of birth (1=the Netherlands; 2=other).

 Health Status

Participants were asked whether they suffered from (any of) the following diseases: diabetes mellitus, myocardial infarction, stroke, high blood pressure, other cardiovascular diseases, and/or cancer. Participants were categorized as ill (1=suffering from at least 1 of the diseases) or healthy (0). To assess quality of life, the Short Form Health Survey (SF-12) questionnaire was used [83-85] (ranging from 18 to 48; based on a mean split of 24, we defined 0=a low quality of life score and 1=a high quality of life score).

 Health Behaviors

All 5 health behaviors were assessed with the use of validated questionnaires. Physical activity was assessed with the Short Questionnaire to Assess Health-Enhancing Physical Activity (SQUASH) [86]. Weekly vegetable intake (raw, boiled, baked, or salad), weekly fruit intake, and fruit juice consumption were assessed with the Food Frequency Questionnaire (FFQ) [87]. Alcohol consumption was assessed with the 5-item Dutch Quantity-Frequency-Variability (QFV) questionnaire [88]. Smoking behavior was assessed by asking if, what (eg, cigarettes, shag), and in what quantities participants smoke and their answers were converted into a score for tobacco consumption according to the recommendations by Mudde and colleagues [89].

 Intervention Use

We defined someone as using the intervention as recommended if he/she started with the suggested number of lifestyle modules based on his/her assessed behavior. Answering the first question within the specific module was defined as starting the module. For example, if a participant did not meet the guideline for smoking, vegetable consumption, and physical activity, this person was expected to start with 3 lifestyle modules (more than 3 modules were also counted as using the intervention as recommended) to be classified as a participant who uses the intervention as recommended. If this participant only started 2 or less modules, he/she was classified as not meeting the intervention recommendation.

 Statistical Analyses

The data was analyzed with SPSS 20.0 (IBM Corp, Armonk, NY, USA). Descriptive statistics were used to describe participants’ characteristics.

Frequency analyses were performed to identify the number of participants who used the intervention as recommended (1) and those who did not (0), as defined by the intervention recommendation. Only participants who got the advice to complete at least 1 module were included in analysis of intervention use as recommended. Differentiations were made between different subgroups for age, gender, education, income, working situation, health status, family status, and quality of life. We used chi-square tests to explore the differences among these groups.

Logistic regression analyses using the Enter method were used to predict intervention use as recommended among different personal characteristics. The number of the received intervention modules, based on the amount of health behaviors that did not met the national guidelines, was used as a dependent variable. Separate logistic regression analyses were carried out dependent on the amount of lifestyle modules participants were advised to complete. One logistic regression analysis including all respondents was conducted with as-recommended program use as a dependent variable to detect characteristics that predict intervention use in general. Age, gender, level of education, income, employment status, family status, household size, country of birth, diseases, and quality of life were included as predictors in all models. The recommended number of modules was also taken into account in the model. Tests were performed at alpha=.05 to indicate statistical significance.

 Results Sample Characteristics

A total of 1638 participants were included in this study (Table 1). Variables with missing data were not filled up (maximum missing values of 2.4%). The mean age was 43.9 years (SD 12.6) and slightly more men (53.60%, 878/1638) than women (46.40%, 760/1638) participated in the trial. A minority of the participants had a low educational level (10.84%, 174/1605). The majority had a middle income (46.83%, 767/1606), were in a relationship (75.94%, 1215/1600), and came originally from the Netherlands (95.27%, 1531/1607). Most participants were healthy (79.84%, 1283/1607) and reported a high quality of life (58.51%, 935/1598).

 Intervention Use

Figure 1 represents the percentages of participants who used the intervention in the recommended way. For example, 414 of 585 participants (70.8%, red bar) who were advised to complete at least 2 lifestyle modules did not do so, 162 participants (27.7%, blue bar) followed the recommendation, and a minority (9/585, 1.5%, green bar), attended more than 2 modules. These percentages indicate that the healthier their lifestyle and the fewer modules participants were advised to complete, the more participants followed the intervention guideline.

Percentage of participants who used the intervention in the recommended way.

Sample characteristics (N=1638).

Characteristics n (%) Mean (SD) Range
Age (years) 43.94 (12.57) 19-65
Gender
Male 878 (53.60)
Female 760 (46.40)
Educational level
High 700 (43.16)
Middle 731 (45.55)
Low 177 (10.84)
Income (€)
<1750 373 (22.77)
1751-3050 767 (46.83)
>3051 466 (28.45)
Working situation
Paid job 1240 (77.26)
Nonpaid job 365 (22.74)
Family status
Single 385 (24.06)
In relationship 1215 (75.94)
Number of people in household 2.89 (1.37) 1-11
Country of birth
The Netherlands 1531 (95.27)
Other 76 (4.73)
Disease status
Ill 324 (20.16)
Healthy 1283 (79.84)
Quality of life (SF-12) 40.19 (5.08) 18-48
High 935 (58.51)
Low 663 (41.49)
Number of modules recommended
0 174 (10.62)
1 451 (27.53
2 585 (35.71)
3 315 (19.23)
4 100 (6.11)
5 13 (0.79)
Intervention Use by Different Subgroups

The table in Multimedia Appendix 1 gives an overview of the number of unhealthy behaviors and the number of started modules, differentiated by several personal variables. Figure 2 graphically summarizes the difference between the subgroups and the number of participants who did not comply with the intervention recommendation compared to those who used the intervention as recommended.

Significantly more older (39.50%, 361/914) than younger participants (26.3%, 191/726; χ2 1=44.8, P<.001) and significantly more women (50.66%, 385/760) than men (38.8%, 341/878; χ2 1=23.0, P<.001) used the intervention as recommended.

People with low education (42.0%, 73/174) adhered best to the recommendation to participate in the suggested modules depending on their health behavior compared with participants with middle (33.2%, 243/174) or high education (32.6%, 228/700). However, no significant differences among these 3 educational levels were found with regard to recommended intervention use (χ2 1=2.9, P=.23).

Participants with a low income (33.24%, 124/373), middle income (34.29%, 263/767), and high income (33.26%, 155/466) also did not differ significantly from one another with regard to intervention use (χ2 2=0.6, P=.72).

However, participants without a paid job (40.0%, 146/365) followed the recommendation of the intervention significantly more often than participants with a paid job (32.02%, 397/1240; χ2 1=7.9, P=.01).

Those participants who were ill (37.7%, 123/324) did use the intervention as recommended significantly more often compared to healthy participants (32.74%, 420/1283; χ2 1=4.5, P=.02).

Participants in a relationship (46.34%, 563/1215) followed the intervention recommendation significantly more often than single participants (38.2%, 147/385; χ2 1=7.8, P=.003).

Finally, no significant differences were found for participants with a high quality of life (43.5%, 407/935) compared to those with a low quality of life (45.9%, 304/663; χ2 1=0.9, P=.19).

Percentage of participants who followed the recommendation to start with the correct number of intervention modules differentiated by education, income, work, age, gender, and disease status. Age was categorized as 1=young and 2=old based on a mean split of 44 years. *P<.05, **P<.001.

 Predictors of Intervention Use

As indicated in Table 2, higher age at baseline was a significant predictor of following the relevant recommendation to start with 1 or more modules. This was found for those who were recommended to follow 1, 2, and 3 modules, and in general. Age did not predict intervention use as recommended for respondents who were recommended to follow 4 lifestyle modules.

Logistic regression results for the relationship between socioeconomic variables, personal characteristics, and following the intervention recommendation.

Predictora Number of modules recommended to start with
1 module (n=427) 2 modules (n=556) 3 modules (n=302) 4 modules (n=108)
β P OR (95% CI) β P OR (95% CI) β P OR (95% CI) β P OR (95% CI)
Age (cont) 0.04 <.001 1.05 (10.02-10.07) 0.04 <.001 1.04 (1.02-1.05) 0.05 .002 1.06 (1.02-1.09) –0.03 .30 0.97 (0.90-1.04)
Gender (ref=female) 0.17 .62 1.13 (0.71-10.78) 0.54 .007 1.16 (1.24-2.54) 0.41 .22 1.60 (0.76-3.24) –0.91 .24 0.41 (0.05-3.07)
Diseases (ref=healthy) 0.08 .82 1.09 (0.57-2.04) –0.17 .50 0.84 (0.51-1.39) 0.40 .24 1.72 (0.69-4.2) –1.36 .37 0.26 (0.05-1.41)
Country of birth (ref=other than NL)b –0.32 .52 0.73 (0.27-1.95) –0.46 .31 0.63 (0.27-1.53)
Family status (ref=relationship) –0.68 .03 0.50 (0.27-0.94) 0.35 .21 1.42 (0.82-2.25) 0.68 .16 2.12 (0.71-6.03) –1.24 .41 0.29 (0.04-2.18)
Household (cont) –0.10 .25 0.90 (0.76-1.07) –0.04 .69 0.96 (0.83-1.13) –0.12 .42 0.88 (0.66-1.21) –0.80 .07 0.45 (0.20-1.05)
Income (ref=low)
Low .33 .25 .59 .12
Middle 0.46 .21 1.59 (0.77-3.29) –0.48 .15 0.62 (0.33-1.2) –0.16 .53 0.67 (0.26-2.85) –1.25 .15 0.06 (0.02-4.18)
High 0.37 .17 1.45 (0.85-2.50) –0.36 .13 0.70 (0.43-1.11) 0.21 .81 1.12 (0.51-2.99) 1.01 .55 1.86 (0.39-19.44)
Work situation (ref=unemployed) –0.23 .45 0.79 (0.44-1.44) –0.001 .99 0.99 (0.62-1.61) –0.27 .68 0.84 (0.36-1.9) –1.26 .14 0.20 (0.04-1.99)
Education (ref=low)
Low .65 .74 .70 .08
Middle 0.15 .73 1.16 (0.50-2.70) –0.26 .44 0.77 (0.40-1.5) 0.40 .40 1.60 (0.57-4.67) 2.84 .08 10.99 (1.38-212.83)
High 0.24 .36 1.27 (0.77-2.08) –0.06 .79 0.94 (0.62-1.49) 0.08 .71 1.17 (0.52-2.54) –0.26 .65 0.65 (0.13-4.65)
QOL (cont) –0.02 .37 0.98 (0.93-1.03) –0.07 .002 0.93 (0.89-0.98) .22 0.96 (0.9-1.03) –0.09 .24 0.92 (0.8-1.06)

aCont=continuous; ref=reference group for categorical variables.

b Analysis of country of birth not possible for those in 3 and 4 modules because number of participants not from the Netherlands<10.

Being single significantly predicted recommended intervention use, but only for those participants who were advised to start with 1 module. However, none of the socioeconomic variables (education, income, and work) had a significant influence on the intervention use behavior regardless of the number of unhealthy behaviors.

A low quality of life (SF-12) was associated with being more likely to use the intervention as recommended for only people who received the advice to start with 2 modules. It should be noticed that the analysis of country of birth was not possible for the model with 3 and 4 recommended modules because the number of participants not from the Netherlands was less than 10.

The regression analysis of intervention use in general indicated that being older, female, having a lower quality of life, and given the recommendation to complete fewer lifestyle modules were significant predictors of intervention use according to recommendations (see Table 3).

Logistic regression results for the relationship between socioeconomic variables, personal characteristics, and following the intervention recommendation within the complete sample (N=1586).

Predictor β P OR (95% CI)
Age (cont) 0.04 <.001 1.04 (1.02-1.05)
Gender (ref=female) 0.34 .02 1.40 (1.08-1.80)
Diseases (ref=healthy) –0.18 .71 0.94 (0.67-1.31)
Country of birth (ref=not NL) –0.26 .40 0.77 (0.42-1.41)
Family status (ref=in relationship) 0.01 .97 1.0 (0.67-1.46)
Household (cont) –0.09 .07 0.91 (0.82-1.01)
Income (ref=high)
Low .67
Middle –0.16 .47 0.86 (0.56-1.31)
High 0.01 .98 1.06 (0.74-1.34)
Work situation (ref=unemployed) –0.15 .35 0.86 (0.62-1.19)
Education (ref=high)
Low .78
Middle 0.06 .79 1.06 (0.68-1.67)
High 0.08 .58 1.08 (0.82-1.43)
QOL (cont) –0.04 .002 0.96 (0.93-0.98)
Module recommendation (cont) –1.59 <.001 0.20 (0.17-0.24)

a Cont=continuous; ref=reference group for categorical variables.

 Discussion Principal Findings

Because eHealth intervention use as recommended increases the effectiveness of behavior change [76], it is of high importance that people at a high risk of unhealthy lifestyle behaviors use those interventions in such a way. Our analysis of a sample of the general Dutch population revealed that there was a difference in intervention use among people grouped by different personal characteristics.

Contrary to earlier findings regarding Internet use and age (eg, [56,59]), we found that more older than younger participants used the intervention as recommended. It might be possible that older people were less familiar with eHealth interventions and, therefore, gained more information that was new and relevant to them resulting in more frequent use of the intervention modules.

Women in our study used the intervention as recommended more often, which could be explained by the fact that women use the Internet to seek health information more frequently than men [90]. Females tend to be more interested in health topics [91,92] and rely more often on the Internet as a trustful source [59,93]. These explanations might be possible reasons why women used the intervention as recommended more frequently compared to men in our study. In addition, males have been found to evaluate the Internet as a less valuable source of health information than women do [59], which might include eHealth interventions, and this may be another reason for the lower intervention adherence by men in our study.

Participants within a relationship have been found to use the intervention as recommended more frequently compared to singles. People within a relationship have been found to have healthier behavior and health might be something in their interest which could explain why they are more interested in using the intervention [94]. Further research should explore the importance of family status as well as health behaviors of other family members in more depth because it might be that family members are more likely to behave alike, which might ultimately affect (the need for) intervention use.

People with a lower educational level used more intervention modules than those with a higher education level did. This result is surprising because it is known from the literature that higher-educated people spend more time online to seek health information [62,95]. One explanation is that the lower-educated participants may lack prior knowledge and may have used this intervention to gain more knowledge about a healthy lifestyle [59]. Another explanation could be that people with a lower SES use the Internet primarily to gain information, whereas people with a high SES make use of different sources, including professionals or their social environment, and thus rely less on the Internet for information [93]. But these results must be interpreted with caution because although we found higher as-recommended intervention use, education was not a significant predictor within the regression analysis. Furthermore, our data indicated that income is not a predictor for recommended intervention use, which might indicate that income level might not be important with regard to intervention use as recommended.

This study revealed that unemployed people used the intervention as recommended more frequently. Participation in the intervention is time-consuming and it may be that employed people adhered less to the intervention recommendations because they had less leisure time. Previously, van Deursen and van Dijk [57] reported that unemployed people spent more time online than employed people did.

The fact that participants who reported having a disease used our intervention more frequently is in-line with previous literature findings. Individuals who perceive themselves as more ill have been found to use the Internet as a source of health information [96,97]. This might also be an explanation for our finding that participants with a lower level of quality of life used the intervention as recommended. It might be plausible that these participants look for health information and tips about how to change their lifestyle to gain a better health condition and a higher quality of life.

We also found the more modules recommended to complete, the fewer were done by participants. Following the recommendations of an eHealth intervention requires a significant investment of time for reading and processing information and interacting with the program. Participants who received the recommendation to use many modules might be at greater risk of being overwhelmed by those requirements. If an eHealth program demands too much cognitive effort from their participants, ego depletion [98,99] can arise and participants might be more inclined not to use the program as recommended.

To summarize, we have found differences in intervention use as recommended among participants with different personal characteristics. We know that especially younger people, males, people who have a job, people with illnesses, and singles did not use the intervention as recommended. Furthermore, our analysis revealed that being older, female, having a low quality of life, and a healthier lifestyle are predictors of intervention use as recommended when all personal characteristics are taken into account.

 Strengths, Limitations, and Further Implications

One of the strengths of this study is the multiple-behavior approach because previous studies have demonstrated that those interventions have a high impact on behavior change [2]. Furthermore, we assessed several indications to measure SES, which allows us to compare the impacts of education, income, and occupational status.

In addition to the randomization of the started behaviors, either preventive or addictive modules, one of the limitations of this study is that participants could not choose on their own which module they wanted to begin with in the given module block. This might have increased the risk of participants not using the intervention because they might have disliked a given sequence. Furthermore, a predefined order of the modules may have led to reduced feelings of choice. We were not able to analyze potential consequences of this reduction and further studies may explore this issue at greater depth. Participants could have also misinterpreted a predefined order as an order of importance. Providing the module about fruit consumption first and then smoking might have created the impression that changing fruit consumption is more important than smoking cessation. Another limitation of the predefined order of the modules might be that some participants had to start with a behavior that they experienced as difficult, which may have led to reduced motivation to complete the full program. Future intervention studies should study under which conditions the utilization of a simultaneous strategy favoring multiple behaviors simultaneously or sequentially are preferred by participants.

Although this study used the term “following interventions as recommended,” we can only say for sure that participants started with the recommended amount of modules and not whether they finished the modules or how they used the modules. Next, the aim of the study was to look at differences in people of this intervention group in regards to following intervention recommendations. Yet, participants of an intervention are often preselected, implying that people who are healthier and more interested in health are more willing to participate.

Another limitation of the study at hand is the use of self-reported questionnaires, which could result in an overestimation of healthy lifestyles and participants being given the recommendation to start with fewer intervention modules. Self-reported questionnaires with regard to education, income, and working situation might also result in an underestimation of people with a low SES (eg, false information about income).

Finally, we did not analyze whether following the intervention recommendation resulted in behavior change or not. However, in one of our studies [2], it was investigated that a longer visiting time and greater number of visits in the intervention resulted in higher likelihood of behavior change. To our knowledge, this is one of the first studies focusing on socioeconomic and personal characteristics as a predictor of recommended intervention use. Because we found that older participants, females, unemployed people, ill participants, and people in a relationship more often use the intervention as recommended, we should ask ourselves what strategies can be taken to make interventions more attractive to use for those who did not use the intervention as recommended. Our intervention included several strategies that are known to increase intervention use, such as giving tailored feedback, using goal-setting strategies, action planning, self-efficacy monitoring, and the use of reminders [100,101]. On the other hand, the intervention might be more attractive if interactive elements and communication tools to facilitate social support were used, when involving the social environment, or if entertaining elements (eg, additional games, quizzes) were added [100,101]. Furthermore, the health modules were very similar in terms of structure and type of feedback, and it might be that participants disliked the repetition (which also increases participants’ burden). Future research is warranted to investigate whether improving the flow experiences of participants by using strategies to attract participants’ attention and make interventions more entertaining increases recommended intervention use [102,103].

 Conclusion

Our findings indicate that different subgroups use eHealth interventions in different ways. The more frequent as-recommended intervention use by unemployed, older, and ill participants may be an indication that these eHealth interventions are attractive to people with a greater need for health care information. Therefore, computer-tailored eHealth interventions might be a promising tool to increase health status and maintain healthy lifestyles.

 Multimedia Appendix 1

Number of started intervention lifestyle modules separated personal characteristics.

 Abbreviations FFQ

Food Frequency Questionnaire

 RHA

Regional Health Authorities

 SES

socioeconomic status

 SF-12

Short Form Health Survey questionnaire

 SQUASH

Short Questionnaire to Assess Health-Enhancing Physical Activity

 QFV

Quantity-Frequency-Variability

Hein de Vries is the scientific director of Vision2Health, a company aimed at implementing evidence-based eHealth programs. All other authors have no conflicts of interests.

 Noar SM Haarington NG Noar SM Harrington NG eHealth Applications: Promising Strategies for Behavior Change (Routledge Communication Series) 2010 03 10 London Routledge 336 Schulz DN Kremers SP Vandelanotte C van Adrichem MJ Schneider F Candel MJ de Vries H Effects of a web-based tailored multiple-lifestyle intervention for adults: a two-year randomized controlled trial comparing sequential and simultaneous delivery modes J Med Internet Res 2014 16 1 e26 10.2196/jmir.3094 24472854 v16i1e26 PMC3936298 Brug J Steenhuis I van Assema P de Vries H The impact of a computer-tailored nutrition intervention Prev Med 1996 25 3 236 42 10.1006/pmed.1996.0052 8781000 S0091-7435(96)90052-3 Kreuter MW Farrell D Olevitch L Brennan L Tailoring Health Messages: Customizing Communication with Computer Technology 2000 Mahwah, NJ Erlabum Krebs P Prochaska JO Rossi JS A meta-analysis of computer-tailored interventions for health behavior change Prev Med 2010 51 3-4 214 21 10.1016/j.ypmed.2010.06.004 20558196 S0091-7435(10)00231-8 PMC2939185 de Vries H Brug J Computer-tailored interventions motivating people to adopt health promoting behaviours: introduction to a new approach Patient Educ Couns 1999 02 36 2 99 105 10223015 S073839919800127X Hawkins RP Kreuter M Resnicow K Fishbein M Dijkstra A Understanding tailoring in communicating about health Health Educ Res 2008 06 23 3 454 66 10.1093/her/cyn004 18349033 cyn004 PMC3171505 Noar SM Benac CN Harris MS Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions Psychol Bull 2007 07 133 4 673 93 10.1037/0033-2909.133.4.673 17592961 2007-09203-006 Lustria MLA Noar SM Cortese J Van Stee SK Glueckauf RL Lee J A meta-analysis of web-delivered tailored health behavior change interventions J Health Commun 2013 18 9 1039 69 10.1080/10810730.2013.768727 23750972 Brug J Oenema A Campbell M Past, present, and future of computer-tailored nutrition education Am J Clin Nutr 2003 04 77 4 Suppl 1028S 1034S 12663313 Bennett GG Glasgow RE The delivery of public health interventions via the Internet: actualizing their potential Annu Rev Public Health 2009 30 273 92 10.1146/annurev.publhealth.031308.100235 19296777 Griffiths F Lindenmeyer A Powell J Lowe P Thorogood M Why are health care interventions delivered over the internet? A systematic review of the published literature J Med Internet Res 2006 8 2 e10 10.2196/jmir.8.2.e10 16867965 v8i2e10 PMC1550698 Noar SM Grant Harrington N Van Stee SK Shemanski Aldrich R Tailored health communication to change lifestyle behaviors American Journal of Lifestyle Medicine 2010 11 19 5 2 112 122 10.1177/1559827610387255 Krebs P Prochaska JO Rossi JS A meta-analysis of computer-tailored interventions for health behavior change Prev Med 2010 51 3-4 214 21 10.1016/j.ypmed.2010.06.004 20558196 S0091-7435(10)00231-8 PMC2939185 Kohl LFM Crutzen R de Vries NK Online prevention aimed at lifestyle behaviors: a systematic review of reviews J Med Internet Res 2013 15 7 e146 10.2196/jmir.2665 23859884 v15i7e146 PMC3714003 Ammann R Vandelanotte C de Vries H Mummery WK Can a website-delivered computer-tailored physical activity intervention be acceptable, usable, and effective for older people? Health Educ Behav 2013 04 40 2 160 70 10.1177/1090198112461791 23077157 1090198112461791 van 't Riet J Ruiter RAC Werrij MQ de Vries H Investigating message-framing effects in the context of a tailored intervention promoting physical activity Health Educ Res 2010 04 25 2 343 54 10.1093/her/cyp061 19841041 cyp061 van Stralen MM de Vries H Mudde AN Bolman C Lechner L The long-term efficacy of two computer-tailored physical activity interventions for older adults: main effects and mediators Health Psychol 2011 07 30 4 442 52 10.1037/a0023579 21639638 2011-10886-001 Norman GJ Zabinski MF Adams MA Rosenberg DE Yaroch AL Atienza AA A review of eHealth interventions for physical activity and dietary behavior change Am J Prev Med 2007 10 33 4 336 345 10.1016/j.amepre.2007.05.007 17888860 S0749-3797(07)00363-7 PMC2180189 Broekhuizen K Kroeze W van Poppel MNM Oenema A Brug J A systematic review of randomized controlled trials on the effectiveness of computer-tailored physical activity and dietary behavior promotion programs: an update Ann Behav Med 2012 10 44 2 259 86 10.1007/s12160-012-9384-3 22767052 PMC3442159 Brug J Campbell M van Assema P The application and impact of computer-generated personalized nutrition education: a review of the literature Patient Educ Couns 1999 02 36 2 145 56 10223019 S0738-3991(98)00131-1 Hamel LM Robbins LB Computer- and web-based interventions to promote healthy eating among children and adolescents: a systematic review J Adv Nurs 2013 01 69 1 16 30 10.1111/j.1365-2648.2012.06086.x 22757605 Broekhuizen K Kroeze W van Poppel MNM Oenema A Brug J A systematic review of randomized controlled trials on the effectiveness of computer-tailored physical activity and dietary behavior promotion programs: an update Ann Behav Med 2012 10 44 2 259 86 10.1007/s12160-012-9384-3 22767052 PMC3442159 Neville LM O'Hara B Milat AJ Computer-tailored dietary behaviour change interventions: a systematic review Health Educ Res 2009 08 24 4 699 720 10.1093/her/cyp006 19286893 cyp006 PMC2706490 Elfeddali I Bolman C Candel MJ Wiers RW de Vries H Preventing smoking relapse via Web-based computer-tailored feedback: a randomized controlled trial J Med Internet Res 2012 14 4 e109 10.2196/jmir.2057 22903145 v14i4e109 PMC3510689 Smit ES de Vries H Hoving C Effectiveness of a Web-based multiple tailored smoking cessation program: a randomized controlled trial among Dutch adult smokers J Med Internet Res 2012 14 3 e82 10.2196/jmir.1812 22687887 v14i3e82 PMC3414857 McClure JB Shortreed SM Bogart A Derry H Riggs K St John J Nair V An L The effect of program design on engagement with an internet-based smoking intervention: randomized factorial trial) J Med Internet Res 2013 15 3 e69 10.2196/jmir.2508 23529377 v15i3e69 PMC3636802 de Josselin de Jong S Candel M Segaar D Cremers H de Vries H Efficacy of a Web-based computer-tailored smoking prevention intervention for Dutch adolescents: randomized controlled trial J Med Internet Res 2014 16 3 e82 10.2196/jmir.2469 24657434 v16i3e82 PMC3978560 Riper H Kramer J Keuken M Smit F Schippers G Cuijpers P Predicting successful treatment outcome of web-based self-help for problem drinkers: secondary analysis from a randomized controlled trial J Med Internet Res 2008 10 4 e46 10.2196/jmir.1102 19033150 v10i4e46 PMC2629366 Riper H Spek V Boon B Conijn B Kramer J Martin-Abello K Smit F Effectiveness of E-self-help interventions for curbing adult problem drinking: a meta-analysis J Med Internet Res 2011 13 2 e42 10.2196/jmir.1691 21719411 v13i2e42 PMC3221381 Rooke S Thorsteinsson E Karpin A Copeland J Allsop D Computer-delivered interventions for alcohol and tobacco use: a meta-analysis Addiction 2010 08 105 8 1381 90 10.1111/j.1360-0443.2010.02975.x 20528806 ADD2975 van Keulen HM Mesters I Ausems M van Breukelen G Campbell M Resnicow K Brug J de Vries H Tailored print communication and telephone motivational interviewing are equally successful in improving multiple lifestyle behaviors in a randomized controlled trial Ann Behav Med 2011 02 41 1 104 18 10.1007/s12160-010-9231-3 20878293 PMC3030742 Smeets T Kremers SP Brug J de Vries H Effects of tailored feedback on multiple health behaviors Ann Behav Med 2007 04 33 2 117 23 10.1080/08836610701307801 17447863 Curioni CC Lourenço PM Long-term weight loss after diet and exercise: a systematic review Int J Obes (Lond) 2005 10 29 10 1168 74 10.1038/sj.ijo.0803015 15925949 0803015 Vandelanotte C Reeves MM Brug J De Bourdeaudhuij I A randomized trial of sequential and simultaneous multiple behavior change interventions for physical activity and fat intake Prev Med 2008 03 46 3 232 7 10.1016/j.ypmed.2007.07.008 17707079 S0091-7435(07)00299-X van Keulen HM Bosmans JE van Tulder MW Severens JL de Vries H Brug J Mesters I Cost-effectiveness of tailored print communication, telephone motivational interviewing, and a combination of the two: results of an economic evaluation alongside the Vitalum randomized controlled trial Int J Behav Nutr Phys Act 2010 7 64 10.1186/1479-5868-7-64 20815869 1479-5868-7-64 PMC2940922 Smit ES Evers SM de Vries H Hoving C Cost-effectiveness and cost-utility of Internet-based computer tailoring for smoking cessation J Med Internet Res 2013 15 3 e57 10.2196/jmir.2059 23491820 v15i3e57 PMC3636293 Schulz DN Smit ES Stanczyk NE Kremers SP de Vries H Evers SM Economic evaluation of a web-based tailored lifestyle intervention for adults: findings regarding cost-effectiveness and cost-utility from a randomized controlled trial J Med Internet Res 2014 16 3 e91 10.2196/jmir.3159 24650860 v16i3e91 PMC3978559 Mackenbach JP Stirbu I Roskam AR Schaap MM Menvielle G Leinsalu M Kunst AE European Union Working Group on Socioeconomic Inequalities in Health Socioeconomic inequalities in health in 22 European countries N Engl J Med 2008 06 5 358 23 2468 81 10.1056/NEJMsa0707519 18525043 358/23/2468 Lampert T Kroll L E von der Lippe E Müters S Stolzenberg H [Socioeconomic status and health: results of the German Health Interview and Examination Survey for Adults (DEGS1)] Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2013 05 56 5-6 814 21 10.1007/s00103-013-1695-4 23703503 Hulshof KF Brussaard JH Kruizinga AG Telman J Löwik MR Socio-economic status, dietary intake and 10 y trends: the Dutch National Food Consumption Survey Eur J Clin Nutr 2003 01 57 1 128 37 10.1038/sj.ejcn.1601503 12548307 1601503 Gidlow C Johnston LH Crone D Ellis N James D A systematic review of the relationship between socio-economic position and physical activity Health Education Journal 2006 12 01 65 4 338 367 10.1177/0017896906069378 Naimi TS Brewer RD Mokdad A Denny C Serdula MK Marks JS Binge drinking among US adults JAMA 2003 01 1 289 1 70 5 12503979 joc21082 Hiscock R Bauld L Amos A Fidler JA Munafò M Socioeconomic status and smoking: a review Ann N Y Acad Sci 2012 02 1248 107 23 10.1111/j.1749-6632.2011.06202.x 22092035 Eysenbach G The law of attrition J Med Internet Res 2005 7 1 e11 10.2196/jmir.7.1.e11 15829473 v7e11 PMC1550631 Cuijpers P van Straten A Andersson G Internet-administered cognitive behavior therapy for health problems: a systematic review J Behav Med 2008 04 31 2 169 77 10.1007/s10865-007-9144-1 18165893 PMC2346512 Moroshko I Brennan L O'Brien P Predictors of dropout in weight loss interventions: a systematic review of the literature Obes Rev 2011 11 12 11 912 34 10.1111/j.1467-789X.2011.00915.x 21815990 Schulz DN Schneider F de Vries H van Osch LA van Nierop PW Kremers SP Program completion of a web-based tailored lifestyle intervention for adults: differences between a sequential and a simultaneous approach J Med Internet Res 2012 14 2 e26 10.2196/jmir.1968 22403770 v14i2e26 PMC3376514 Wei K Teo H Chan HC Tan BCY Conceptualizing and testing a social cognitive model of the digital divide Information Systems Research 2011 03 22 1 170 187 10.1287/isre.1090.0273 van Deursen A van Dijk J Internet skills and the digital divide New Media & Society 2010 12 06 13 6 893 911 10.1177/1461444810386774 Brandtweiner R Donat E Kerschbaum J How to become a sophisticated user: a two-dimensional approach to e-literacy New Media & Society 2010 01 28 12 5 813 833 10.1177/1461444809349577 Warschauer M Technology and Social Inclusion: Rethinking the Digital Divide 2003 Cambridge, MA MIT Press Aerschot LV Rodousakis N The link between socio-economic background and Internet use: barriers faced by low socio-economic status groups and possible solutions Innovation: The European Journal of Social Science Research 2008 12 21 4 317 351 10.1080/13511610802576927 Akkermans MMP Cleassen PJGM Donselaar P Heerschap NM de Heij HN Kuijpers AD ICT, kennis en economie 2012 2015-05-04 Den Haag Centraal Bureau voor de Statistiek http://www.cbs.nl/NR/rdonlyres/130F8419-05C1-43AE-B5ED-4C373F34EC82/0/2012i78pub.pdf 6YHWDb4kN Statistics Netherlands 2014 05 07 2015-05-04 Dutch internet skills above EU averagehttp://www.cbs.nl/en-GB/menu/themas/vrije-tijd-cultuur/publicaties/artikelen/archief/2014/2014-4066-wm.htm 6YHWK2WfT Hargittai E Digital na(t)ives? Variation in Internet skills and uses among members of the “net generation” Sociological Inquiry 2010 01 10 80 1 92 113 10.1111/j.1475-682X.2009.00317.x van Deursen AJ van Dijk JA The digital divide shifts to differences in usage New Media & Society 2013 06 07 16 3 507 526 10.1177/1461444813487959 Dijk JA The Deepening Divide: Inequality in the Information Society 2005 Thousand Oaks, CA Sage Publications, Inc Costa-Font J Mossialos E Rudisill C When is the Internet a valued communication device for health information in Europe? Economics of Innovation and New Technology 2009 07 18 5 429 445 10.1080/10438590802547159 Howard P Rainie L Jones S Days and nights on the Internet: the impact of a diffusing technology American Behavioral Scientist 2001 11 01 45 3 383 404 10.1177/0002764201045003003 Hargittai E Hinnant A Digital inequality: differences in young adults' use of the Internet Communication Research 2008 08 04 35 5 602 621 10.1177/0093650208321782 Wangberg SC Andreassen HK Prokosch H Santana SM Sørensen T Chronaki CE Relations between Internet use, socio-economic status (SES), social support and subjective health Health Promot Int 2008 03 23 1 70 7 10.1093/heapro/dam039 18083686 dam039 Zarcadoolas C Blanco M Boyer JF Pleasant A Unweaving the Web: an exploratory study of low-literate adults' navigation skills on the World Wide Web J Health Commun 2002 07 7 4 309 24 10.1080/10810730290088157 12356289 Brodie M Flournoy RE Altman DE Blendon RJ Benson JM Rosenbaum MD Health information, the Internet, and the digital divide Health Affairs 2000 11 01 19 6 255 265 10.1377/hlthaff.19.6.255 Močnik D Širec K The determinants of Internet use controlling for income level: Cross-country empirical evidence Information Economics and Policy 2010 7 22 3 243 256 10.1016/j.infoecopol.2010.01.002 Martin SP Robinson JP The income digital divide: trends and predictions for levels of Internet use Social Problems 2007 02 54 1 1 22 10.1525/sp.2007.54.1.1 Statistics Netherlands 2013 12 13 2015-05-04 Rapid increase internet usage among older peoplehttp://www.cbs.nl/nl-NL/menu/themas/vrije-tijd-cultuur/publicaties/artikelen/archief/2013/2013-4005-wm.htm 6YHWU9bky Olphert W Damodaran L Older people and digital disengagement: a fourth digital divide? Gerontology 2013 59 6 564 70 10.1159/000353630 23969758 000353630 Jones S Fox S Generations online in 2009 2009 01 28 2015-05-04 Washington, DC Pew Internet & American Life Project http://www.pewinternet.org/2009/01/28/generations-online-in-2009/ 6YHWbiEZA Dutton WH Blank G The emergence of next generation internet users Int Econ Econ Policy 2013 8 29 11 1-2 29 47 10.1007/s10368-013-0245-8 Cooper J The digital divide: the special case of gender Journal of Computer Assisted Learning 2006 10 22 5 320 34 10.1111/j.1365-2729.2006.00185.x Hargittai E Shafer S Differences in actual and perceived online skills: the role of gender Social Science Q 2006 06 87 2 432 448 10.1111/j.1540-6237.2006.00389.x Leung L Lee PS Multiple determinants of life quality: the roles of Internet activities, use of new media, social support, and leisure activities Telematics and Informatics 2005 8 22 3 161 180 10.1016/j.tele.2004.04.003 Mitchell M Lebow J Uribe R Grathouse H Shoger W Internet use, happiness, social support and introversion: A more fine grained analysis of person variables and internet activity Computers in Human Behavior 2011 9 27 5 1857 1861 10.1016/j.chb.2011.04.008 Nie NH Hillygus DS Erbring L Internet Uuse, interpersonal relations, and sociability: a time diary study The Internet in Everyday 2002 12 05 Oxford, UK Blackwell Publishers Ltd Elder JP Ayala GX Harris S Theories and intervention approaches to health-behavior change in primary care Am J Prev Med 1999 11 17 4 275 84 10606196 S074937979900094X Vlismas K Stavrinos V Panagiotakos DB Socio-economic status, dietary habits and health-related outcomes in various parts of the world: a review Cent Eur J Public Health 2009 06 17 2 55 63 19662821 Cowell AJ The relationship between education and health behavior: some empirical evidence Health Econ 2006 02 15 2 125 46 10.1002/hec.1019 16167321 Margolis R Educational differences in healthy behavior changes and adherence among middle-aged Americans J Health Soc Behav 2013 54 3 353 68 10.1177/0022146513489312 23988727 0022146513489312 PMC3998203 Schulz DN Kremers SP van Osch LA Schneider F van Adrichem MJ de Vries H Testing a Dutch web-based tailored lifestyle programme among adults: a study protocol BMC Public Health 2011 11 108 10.1186/1471-2458-11-108 21324181 1471-2458-11-108 PMC3050747 de Vries H Mesters I van de Steeg H Honing C The general public's information needs and perceptions regarding hereditary cancer: an application of the Integrated Change Model Patient Educ Couns 2005 02 56 2 154 65 10.1016/j.pec.2004.01.002 15653244 S0738399104000758 Verweij A RIVM 2008 09 22 2015-05-04 Sociaaleconomische status Beschrijving opleidingscategorieën Onderwijsdeelname: Indeling opleidingsniveauhttp://www.nationaalkompas.nl/bevolking/scholing-en-opleiding/indeling-opleidingsniveau/ 6YHX5fvpI Ware J Kosinski M Keller S A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity Med Care 1996 03 34 3 220 33 8628042 Gandek B Ware JE Aaronson NK Apolone G Bjorner JB Brazier JE Bullinger M Kaasa S Leplege A Prieto L Sullivan M Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries: results from the IQOLA Project. International Quality of Life Assessment J Clin Epidemiol 1998 11 51 11 1171 8 9817135 S0895435698001097 Ware J Kosinski M Keller SD A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity Med Care 1996 03 34 3 220 33 8628042 Wendel-Vos GC Schuit AJ Saris WJ Kromhout D Reproducibility and relative validity of the short questionnaire to assess health-enhancing physical activity J Clin Epidemiol 2003 12 56 12 1163 9 14680666 S0895435603002208 RIVM Lokale en Nationale Monitor Volksgezondheid 2005 2015-05-04 Wat is de lokale en nationale Monitor gezondheid?https://www.monitorgezondheid.nl/LNMV.xml 6YHXMguZ4 Lemmens P Tan ES Knibbe RA Measuring quantity and frequency of drinking in a general population survey: a comparison of five indices J Stud Alcohol 1992 09 53 5 476 86 1405641 Mudde A Willemsen M Kremers S de Vries H The Hague: STIVORO 2006 2015-05-04 Den Haag STIVORO Measuring instruments for research regarding smoking and smoking cessation Meetinstrumenten voor onderzoek naar roken en stoppen met rokenhttp://stivoro.nl/wp-content/uploads/2012/docs/rapporten/diversen/meetinstrumenten_3.pdf 6YHXTfW20 Cotten SR Gupta SS Characteristics of online and offline health information seekers and factors that discriminate between them Soc Sci Med 2004 11 59 9 1795 806 10.1016/j.socscimed.2004.02.020 15312915 S0277953604000796 Fox S Fallows D Internet Health Resources 2003 07 16 2015-05-04 Washington, DC Pew Internet & American Life Project http://www.pewinternet.org/2003/07/16/internet-health-resources/ 6YHXYvxSq Cline RJ Haynes KM Consumer health information seeking on the Internet: the state of the art Health Educ Res 2001 12 16 6 671 92 11780707 Bell AV "I think about Oprah": social class differences in sources of health information Qual Health Res 2014 04 24 4 506 16 10.1177/1049732314524637 24623661 1049732314524637 Umberson D Montez JK Social relationships and health: a flashpoint for health policy J Health Soc Behav 2010 51 Suppl S54 66 10.1177/0022146510383501 20943583 51/1_suppl/S54 PMC3150158 Jensen JD King AJ Davis LA Guntzviller LM Utilization of internet technology by low-income adults: the role of health literacy, health numeracy, and computer assistance J Aging Health 2010 09 22 6 804 26 10.1177/0898264310366161 20495159 0898264310366161 Bundorf MK Baker L Singer S Wagner T Consumer demand for health information on the Internet: National Bureau of Economic Research 2004 03 2015-05-04 National Bureau of Economic Research http://www.nber.org/papers/w10386.pdf 6YHXm4ly7 Brashers DE Goldsmith DJ Hsieh E Information seeking and avoiding in health contexts Human Comm Res 2002 04 28 2 258 271 10.1111/j.1468-2958.2002.tb00807.x Baumeister RF Heatherton TF Tice D Losing control: how and why people fail at self-regulation 1994 San Diego Academic Press 307 Baumeister RF Muraven M Tice DM Ego depletion: a resource model of volition, self-regulation, and controlled processing Social Cognition 2000 06 18 2 130 150 10.1521/soco.2000.18.2.130 Crutzen R de Nooijer J Brouwer W Oenema A Brug J de Vries NK Strategies to facilitate exposure to internet-delivered health behavior change interventions aimed at adolescents or young adults: a systematic review Health Educ Behav 2011 02 38 1 49 62 10.1177/1090198110372878 21189422 1090198110372878 Brouwer W Kroeze W Crutzen R de Nooijer J de Vries NK Brug J Oenema A Which intervention characteristics are related to more exposure to internet-delivered healthy lifestyle promotion interventions? A systematic review J Med Internet Res 2011 13 1 e2 10.2196/jmir.1639 21212045 v13i1e2 PMC3221341 Csikszentmihalyi M Finding Flow: The Psychology of Engagement with Everyday Life 1997 New York Basic Books Webster J Trevino LK Ryan L The dimensionality and correlates of flow in human-computer interactions Computers in Human Behavior 1993 12 9 4 411 426 10.1016/0747-5632(93)90032-N