An exercise prescribed intervention used in Denmark in the County of Funen and the Municipality of Frederiksberg called EoP is used to initiate physical activity among sedentary individuals with or in risk of lifestyle diseases. EoP is divided into two areas.

1) A Treatment Perspective (TP) directed towards individuals with specific medically controlled lifestyle diseases. The intervention design for TP was organised as described by Sørensen and colleagues 13. The participants in TP were referred by their GP and followed a supervised group-based training intervention along with 8–12 other TP participants (including participants not taking part in the study) (Figure 1). The group-based supervised training was carried out by physiotherapists or exercise specialists. During the first two months, two weekly 1-hour training sessions were completed. During the final two months, one weekly training session was completed, supplemented by one weekly unassisted training session. The group-based training sessions involved elements of aerobic exercise (e.g. Nordic Walking, Aerobic), strength training, stretching and games. Furthermore, the participants were introduced to activities in the local area during the 4 months. The physiotherapist or the exercise specialist focused primarily on training improving aerobic capacity (Figure 1) 13. In addition, the participants received motivational counselling at baseline and after four months. Subsequently they received voluntary phone based and/or personal motivational counselling after ten and sixteen months. The motivational counselling was carried out by physiotherapists or exercise specialists and used for making a physical activity schedule 36. The motivational counselling aimed at increasing daily physical activity and concerned barriers towards being physically active and physical activity initiatives in the local area of the participant. The participant was responsible for carrying out the schedule.

2) A Preventive Perspective (PP) directed towards healthy citizens in risk of developing lifestyle diseases due to physical inactivity was conducted. The participants in PP entered the intervention by their own initiative. In PP, only the motivational counselling existed as a structured part of the intervention. It was carried out by physiotherapists or exercise specialists. Training was carried out unassisted or in existing local sports clubs or initiatives. Participants in PP received personal motivational counselling at baseline and after four months. Subsequently they received voluntary phone based and/or personal motivational counselling after ten and sixteen months. Further needs for counselling outside set time schedule were accommodated. The motivational counselling was used for making a physical activity schedule in co-operation between the physiotherapist or the exercise specialist and the participant 36. The motivational counselling aimed at increasing daily physical activity and concerned barriers towards being physically active and physical activity initiatives in the local area of the participant. The participant was responsible for carrying out the schedule (Figure 1).

a) The participants' precondition in motivation, self-efficacy and readiness to change before the intervention is important for compliance. A psychological precondition to a greater extend directed towards behaviour change leads to a higher level of persistency in physical activity after intervention.

b) Participation in EoP leads to a positive change in motivation, self-efficacy and readiness to change. This positive change will lead to better adherence.

c) The participants' motivation, self-efficacy and readiness to change are different for TP and PP. Because PP participants enter the intervention by their own initiative, their psychological precondition is to a higher degree than the TP group expected to be directed towards behaviour change. This difference could be central for the participants' persistency in physical activity.

The aim of this study is to assess differences between TP and PP concerning motivation, self-efficacy, readiness to change and self-reported physical activity. Moreover, to assess psychological differences within the groups in short (4 months) and long term (10 and 16 months) motivation, self-efficacy, readiness to change. Furthermore, the aim is to assess self-reported physical activity at the given time-points.

GP's referred participants with the following characteristics to TP: 1) medically controlled lifestyle diseases (e.g. Type-2 diabetes and high blood pressure) 2) motivated to change lifestyle (this was estimated through personal conversation), 3) believed to be able to improve health from an increased level of physical activity, 4) willing to pay 750 Dkr. (100 €) for the intervention. At the PP intervention the participant could be advised to join PP by their general practitioner if they were physically inactive and/or did not meet any medically controlled condition to be enrolled in the TP. Furthermore, the participants could be enrolled by their own initiative by directly contacting the physiotherapist or exercise specialist. Information about PP was available at pharmacies, local media and health organizations (e.g. Diabetes Society, Heart Society, and The Danish Cancer Society).

In addition, GP's giving advice about TP and PP informed the participant about benefits of a physically active lifestyle in general.

After referral to TP, about half of the patients receiving a prescription contacted the physiotherapist and/or the exercise specialist to make an initial appointment. All patients who made this initial appointment were entitled to be a part of the intervention.

All individuals contacting the physiotherapist and/or the exercise specialist in PP and complying with the inclusion criteria were permitted to be a part of the intervention.

Recruitment in both interventions took place over 21 months in 2005 to 2007. Informed consent was obtained.

The analyses in the study are performed as a combination of a quantitative and a qualitative method. The quantitative data are collected through questionnaires. The qualitative data are collected through semi structured qualitative interviews with key informants 39.

Information about psychological values and self-reported measures (BMI, physical activity, health-related quality of life, compliance with the national guidelines for physical activity, self-reported physical condition and psychological conditions towards behaviour change) are assessed at baseline and at the time of motivational counselling (Figure 1). Elaborating information concerning psychological concepts is gathered by interviews at the same time.

All self-reported measures were assessed by self-administrated questionnaires distributed by the physiotherapist and/or the exercise specialist at baseline and after four months. After ten and sixteen months the questionnaires are distributed by the author. All questionnaires are returned to the author by mail in pre-paid envelopes. Information is collected in the following four categories:

To analyse the hypothesis of the study, the questionnaire consist of different scoring instruments used in other studies. Therefore, an overall power calculation of the sample size needed to analyse the full questionnaire was not possible. Instead, power calculations were done for key scoring instruments to indicate the number of participants needed to detect valuable changes. To detect a 2 point difference in the SF-12 Health Survey PCS and MCS scales over time within one group it requires a sample size of 140. To detect a 5 point difference a sample size of 46 is needed 40. To detect a small to moderate effect (0.4) of VAS and EQ-5D index, a sample size of 100 is needed 48. Power calculations were performed to estimate the sample size needed to detect a difference for Physical Activity. Estimated sample size for two-sample comparison of means done by the "sampsi" command in Stata 9.0 with means of 38 and 40, power as 0.8 and standard deviation as sd1(4) sd2(4) was 63. Estimated sample size for one sample with repeated measures done by the "sampsi" command in Stata 9.0 with the same values was 43 49. Power calculations were done to estimate the sample size needed to detect a difference for Self-efficacy. Estimated sample size for one sample with repeated measures done by the "sampsi" command in Stata 9.0 with a mean of 57, alternative mean as 61, power as 0.9 and standard deviation as sd(17) was 190 49. Using these assumptions for the overall study it was decided to include a sample size of 190 in each group.

All participants from TP and PP who volunteered and signed a written consent were included in the questionnaire study.

To strengthen and elaborate on the results from the questionnaire, psychological issues were assessed in semi-structured in depth interviews as well. The participants were strategically selected as key informants 50. All interviews were performed in the participants' home by the author. Interviews are transcribed verbatim based on predetermined rules of transcription decided between the author and the transcriber 39. Coding and analysis are primarily deductive and based upon the theoretical framework and the hypotheses presented earlier in the protocol 50. Agreement on coding categories and interpretation is reached from having the author and a colleague analyzing the same interviews followed by discussion and adjustment differences. All coding and analysis were done by the author.

Strategically selected key informants were asked to take part in the interview study. Selection criteria (age, gender, project, and educational background), were to assure that information gathered was adequate to cover the broad range of participants in the two projects. Those who signed a written consent were included in the study. To gather in depth information from the participants the sample size in the qualitative study was for TP 4 and for PP 3. Each participant was interviewed at baseline and after 4, 10, and 16 months.

The changes from 0–4, 0–10, and 0–16 months within the two groups separately will be assessed using paired t-test or Wilcoxon rank sum test depending on distribution 51.

Comparisons between the TP and PP groups at baseline, four, ten and sixteen months will be assessed by Mann-Whitney test analyses 51. Furthermore comparisons between TP and PP will be made with changes from 0–4, 0–10, and 0–16 months.

Drop-out analyses will be carried out for 0–4, 0–10, and 0–16 months. Compliant and non-compliant participants will be compared in regards to different baseline values. The effect of drop-out will be compared with differences between the TP and PP group.

Statistical significance was set at p < 0.05.

The Danish Data protection Agency registration number is 2005-41-5248. Due to the non-biological and non-treating perspective of the study no registration to the local ethics committee was needed. ClinicalTrials.gov registration number: NCT00594360.

Validation of the self-administrated questionnaire was done in a similar EoP intervention in Denmark using the same questionnaire. The test-retest reliability of the self-administrated questionnaire proved reliable in terms of agreement percent 18.