Participants and Design of the study
The participants (n = 597) included here are part of a study group consisting of 1410 men and women aged 57-78 years who are participating in a randomized controlled intervention trial on the effects of regular physical exercise and diet (DR's EXTRA study: Dose-Responses to Exercise Training. A randomized controlled trial on the effects of regular physical exercise and diet on endothelial function, atherosclerosis and cognition). In 2002, a representative 15% sample (n = 3000) of all 55- to 74-year-old men and women living in the city of Kuopio (Finland) was invited to participate in an exercise and diet intervention study. Of the subjects invited initially, 2062 expressed an interest in participating and 1410 subjects participated in all four baseline examinations conducted between April 2005 and November 2006. The main exclusion criteria at entrance were conditions that would inhibit safe engagement in the prescribed exercise training, malignant diseases as well as other conditions preventing potential participants from co-operating, as judged by the research physicians.
The participants were randomized into six intervention groups: 1) Reference, 2) Aerobic Exercise, 3) Resistance Exercise, 4) Diet, 5) Aerobic Exercise and Diet, 6) Resistance Exercise and Diet. The design of the DR's EXTRA study is presented in detail in Figure . Both the exercise and diet intervention study and the 'Informed Consent' study were approved by the Research Ethics Committee of the Hospital District of Northern Savo and written informed consent was obtained.
Study design at DR's EXTRA study.
The aerobic exercise group was provided with an individualized training program performed at an intensity of 55-65% of maximal oxygen uptake. Training frequency, duration and intensity were gradually increased during the first six months. Thereafter, the subjects were allocated into two subgroups of different weekly exercise doses: 1000-1500 kcal/wk (60 min, 5 times/wk) or > 1500 kcal/wk (90 min, 5 times/wk).
In the resistance exercise group, the subjects participated in supervised, individually prescribed strength training programs. Training frequency, duration and intensity were gradually increased during the first six months. Thereafter, the subjects were randomized into two subgroups receiving different weekly exercise doses: 1000-1500 kcal/wk (2 session/wk) or > 1500 kcal/wk (3 sessions/wk). In each session, 10-14 muscle groups were trained at an intensity of 60% of maximum, 2 sets, 15 repetitions/set. In addition, the subjects were advised to undertake aerobic exercise twice a week.
In the diet group, the participants received counselling by nutritionists. During the first six months, all participants were provided with instructions based on Finnish Nutrition Recommendations (FNR). The main goals were to substitute unsaturated for saturated fat and to increase the intake of fibre and antioxidants. After six months, the subjects were randomly allocated into two subgroups: FNR or Special Nutrition (SN) group. The FNR group continued with the same recommendations. The SN group was given additional instructions: to increase their use of vegetables, fruits, berries, chicken, nuts and almonds as well as decreasing their consumption of red meat.
First, data on the participant's opinion about the informed consent process were collected by questionnaire over a 23-month period in 2005-2007. The questionnaire was self-administered and tested in a pilot study [9
]. The key elements of informed consent were defined as follows: information, understanding, competence, voluntariness, and decision-making. During the three-month intervention visit at Research Institute of Exercise Medicine, the questionnaire was given to all willing participants (n = 1324) who were still involved in the study. In the 'informed consent' study, 1200 participants returned the questionnaire, the response rate being 91%.
Second, the data on implementation and success in exercise and diet intervention study after 12 months of intervention were measured by the trial personnel. They evaluated the following two questions: 1) How has the intervention been implemented in the exercise and diet intervention study after 12 months of intervention? 2) How successful have the participants been in the study after 12 months' intervention, as measured by changes in their activities with respect to exercise or diet? The design of this study is presented in detail in Figure .
In the data analysis, we took into account three intervention groups (aerobic exercise, resistance exercise, and diet) and omitted the other three groups. The reason for this selection was that in the reference group the participants received no intervention, and in the combined exercise and diet groups, two members of the intervention-group personnel evaluated the participant's success in each intervention separately. In addition, the data analysis was restricted to those participants who completed the 'informed consent' questionnaire. As a consequence, the number of participants totalled 597. From the 'informed consent' questionnaire, we selected those questions which were relevant to the research questions addressed in this paper. The list of variables examined in all analyses and their classifications are presented in Table .
List of variables used in data analyses and their codes and classifications
Of these 597 participants 10 dropped out before the 12 months' intervention. These being four from the aerobic exercise group, three from the resistance exercise group, and three from the diet group. The drop out reasons were as follows: death (n = 2), disease (n = 2), moved elsewhere (n = 1), loss of motivation (n = 1), personal reasons (n = 3), unknown reason (n = 1). The details of the drop outs have been compared with those of the participants still in the study. This comparison (cross-tabulations and Fisher's Exact Test) did not detect any statistically significant difference between baseline characteristics (background variables or intervention group). In addition, the differences that were checked between the drop outs and those who did not drop out were considered not to be clinically important.
Evaluations of implementation and success in the exercise and diet interventions were based on the participants' records and supplemented by intervention-group personnel's subjective estimation. For example, in the aerobic exercise group, the participants recorded the amount, duration and type of their training. In the resistance exercise group, the training record was stored in a personal smart card kept at the fitness facility of the institute, and in the diet group, the participants filled in a food diary. The bookkeeping was scored by a predetermined grading system. In addition, the intervention-group personnel met the participants regularly and interviewed and in this way, they ensured that the subjective evaluation was reliable.
The data were analyzed using the SPSS for Windows software (v. 14.0, 2005, SPSS Inc., Chicago, IL). First, frequencies and percentages were calculated to describe the data. Second, univariate analyses were calculated for every variable examined. (See the Additional file 1
: Univariate data analyses.) In the multivariable analyses, those variables were chosen which had a p-value of ≤ 0.1 in the univariate analyses. Third, backward directed stepwise multivariable analyses were used to identify correlates of variables concerning implementation and success. In the multivariable analyses, regression models were used taking p < 0.05 to identify significant predictors of outcomes. P-values from all steps, and p-values, estimates, and standard errors are reported from the last steps of the multivariable regression models. Only the models including main effects were fitted. Interactions (effect modifications) were not examined.
The main purpose of the analysis procedure performed in this study was to identify and examine the potential correlates for the chosen dependent variables and to generate future hypotheses for testing and confirming the independent determinants for implementation and success. The nature of the analysis protocol is exploratory at this stage.