Epidemiological evidence shows that physical inactivity is an independent risk factor for developing coronary heart disease, cardiovascular disease and all-cause mortality [1–4]. Further, research also demonstrated that physical inactivity increases the risk for other chronic diseases as diabetes mellitus type 2 [5, 6], anxiety and depression [7, 8], different types of cancer , and osteoporosis [10, 11]. Since the majority of the adults in Western countries does not participate in regular physical activities [12, 13] many are at risk and therefore, effective interventions promoting an active lifestyle that can reach large population groups at low costs are of great importance .
Face-to-face counseling, as used in primary care, seemed to be an effective method for delivering physical activity interventions [15, 16], but it is very time-consuming, and only a small proportion of the population can be reached . An alternative strategy is computer-tailoring, which seems to be effective in changing smoking, diet or physical activity behaviour [14, 18–23]. Computer-tailored interventions can be used to mimic face-to-face counselling  by giving participants immediate personally adapted advice after completing an electronic diagnostic questionnaire. With the rapid development of the Internet, a new opportunity is created to distribute computer-tailored interventions in a cost-effective manner.
Today there are more than one billion Internet users worldwide  and this number is still increasing due to a drop in the cost of Internet connections and improved high speed access. The biggest penetration rate is found in North America (69.7%), followed by Oceania (53.5%) and Europe (38.9%) . At the moment the increase of Internet users is the largest in underserved populations, such as the elderly, lower educated persons and women; and consequently the gaps among age, educational attainment and gender are narrowing . Consequently Internet interventions could reach a wide variety of people at once, at any time and location. In the last years, more and more health professionals have started to use the Internet to deliver behavioral change interventions on various topics such as smoking [26, 27], diet  and physical activity [29, 30]. The former studies mainly focused on the acceptability, feasibility and efficacy of health promotion programs via the Internet, but little is known about the actual reach of these programs, more specific: who participates in interactive Internet interventions and who does not?
A critique often heard with regard to health promoting interventions and also to computer-tailored interventions is that it only makes the healthiest people healthier. It seems that participants in these interventions are employed, high educated and have more positive health behaviors compared to the general population [23, 31, 32]. In physical activity interventions, participants are often more physically active than the general population [23, 32]. Further, some studies showed that more women than men participate in health promoting interventions through the Internet [33–36].
A study by McClure  showed that characteristics of visitors to a smoking cessation website (Project Quit) differed by the recruitment strategy that was used. Participants recruited by newsletters were more likely to be female, Caucasian and older compared to participants who were recruited by a proactive invitation letter. However, the total sample was similar to participants who enrolled in phone counseling smoking cessation programs: participants were middle-aged and moderate-to heavy smokers with a history of numerous quit attempts. Further, Cobb and Graham  determined the characteristics of adults who search the Internet for smoking cessation information. They reported that the majority of visitors of the leading smoking cessation website (Quitnet) were female smokers between the ages of 26–44; intended to quit in the next 30 days, and made 5.1 quit attempts during the past year. In a study by Feil et al.  characteristics of participants and non-participants in an Internet-based diabetes self-management support program were compared. There were no differences found in gender or computer familiarity between participants and non-participants but younger patients and those who had diabetes a fewer number of years were more likely to participate. To our knowledge, no studies examined who participates in website-delivered physical activity promotion programs and who does not.
From a public health perspective it is essential to reach a whole population, as many are at risk. Therefore it is important to study characteristics of participants in online physical activity programs and characteristics of non-participants for whom other dissemination strategies to improve physical activity levels are needed. Special attention should be given on whether a physical activity program reaches those people who are at risk, namely adults with low physically activity levels.
In this paper we examined who participated, and conversely who did not participate in a computer-tailored physical activity intervention delivered through the Internet. The first aim of the current study was to investigate if showing interest (i.e. positive answer on reply card) and actually participation (i.e. completing online assessment at least once) could be predicted by means of individual characteristics such as age, gender, socioeconomic status (SES), employment and physical activity level in particular. The second aim was to report the most common reasons for non-participation.
Based on the literature [23, 31–39] we hypothesized that younger people, women, employees, those with higher SES, and those who are already regular physically active would be more likely to show interest and participate in the computer-tailored intervention.