The study's participants were drawn from a cohort of university students from a city in northern Spain. After a selection process beginning with preliminary interviews of 1,975 college students (1,179 women and 796 men), the final sample was made up of 106 subjects (68 females and 38 males) aged between 18 and 30 years, each of whom had scored between 29 and 34 on the adapted self-report depression scale established by Kandel and Davies (1982).

Participants with a score of 29 or higher on this scale were classified as having significant depressive symptoms (Choi et al., 1997). Other inclusion criteria were: (a) being sedentary (defined as vigorous exercise less than three times per week for at least 20minutes a day); (b) not receiving any treatment for depression; and (c) giving their written informed consent. The exclusion criteria were the following: (a) consumption of more than 21 alcoholic beverages per week, (b) hospitalization for a psychiatric disorder in the past 5 years, (c) current participation in other clinical trials, (d) drug use, (e) inability to participate due to medical prescription, and (f) for the females, planned or current pregnancy.

Participants were selected by means of a prior interview and two screening visits (SV1, SV2). The preliminary interviews were administered by the authors, with the aim of recruiting a wide-enough sample size for the study. Prospective participants were informed of the types of activities to be carried out upon being accepted to take part in the study, the exclusion criteria, and the potential benefits with respect to participation (including credits granting). Screening visits were intended to assess depressive symptoms and sedentariness (SV1), and to verify that they could perform physical exercise safely (SV2). After SV2, 108 participants were selected assessment of their capacity to adhere to regular physical exercise (1-week), and also to assess their self-determined motivation (Figure 1). This process lasted for six months. A restricted randomization design was used, which was based on three blocks (gender, age, and BMI) aimed at reaching a balance between group-size and other physical characteristics. The participants gave their written informed consent, and the study was conducted in accordance with The Helsinki Declaration of the World Medical Association (2008), with final approval for the study obtained from the Ethical Committee of the University.

Figure 1.

Participant flow from enrollment to prescreen interview.

Note. Age, p=.47; Gender, p=.98; BMI, p=.71.

(0.37MB).

Depressive symptoms. An adaptation of the six-item self-report scale defined by Kandel and Davies (1982) was used in the present study. This scale assesses six depressive symptoms experienced during the past 12 months (Choi et al., 1997). The responses to the items are added, providing a depressive symptomatology score ranging from 6 to 24 points. These scores were adapted by 10 points to produce a new range between 16 and 34 points (Choi et al., 1997). Response to treatment is defined as a reduction of depressive symptoms in general by at least 50% of the initial scores on the depressive mood scale (Brunoni, Lopes, Kaptchuk, & Fregni, 2009). The Cronbach alpha obtained was .72. While there are more powerful assessment tools to evaluate this component, given its psychometric properties and the fact that the depression scale administered offers a brief and general detection measure, this instrument was chosen for two key reasons: a) it allows identification of participants with depression symptoms easily; b) it has been used in previous studies which were also focused on the relationship between physical activity and depressive symptoms, thus making it possible to carry out additional broad-scoped comparisons (e.g., Sabiston et al., 2013).

Self-determined motivation. The Situational Motivation Scale (SIMS), as defined by Guay, Vallerand and Blanchard (2000), is a measure of situational (or state) motivation towards a chosen activity. This self-report inventory consists of four items for each subscale and was designed to measure intrinsic motivation, identified regulation, external regulation, and amotivation. Participants respond to the stem, “Why are you currently involved in this activity?” Each item is rated on a 7-point Likert scale, ranging from 1 (not at all in agreement) to 7 (completely in agreement). As the subscales of the SIMS assess types of motivation on a self-determination continuum, a self-determination index (SDI) can be calculated. In this study, SDI is computed by multiplying the score of each SIMS subscale by a factor that represents its position on the construct of the self-determination continuum: SDI=2 * intrinsic motivation+identified regulation - external regulation - 2 * amotivation) (Guay et al., 2000). The following Cronbach alphas were obtained: intrinsic motivation=.94; identified regulation=.88; external regulation=.78; and amotivation=.78.

Physical activity. The International Physical Activity Questionnaire (IPAQ) was developed as an instrument for cross-national control of PA and physical inactivity. The IPAQ has reasonable measurement properties for screening the population levels of PA between ages 18-65 years in diverse settings (Craig et al., 2003). In the present study, we used the short version, which provides information about time spent on moderate to vigorous physical activity (MVPA). The MVPA indicator is expressed in metabolic equivalents (METs) - minutes/week of MVPA (i.e. METs-MVPA). The IPAQ has reasonable properties to control the levels of PA. The IPAQ in Spanish was obtained from the website: www.ipaq.ki.se.

We used an experimental design in which participants were randomly divided into four groups (ET, NET, AE, & Control). The authors of the present manuscript selected participants and assigned the sample randomization sequence, which was implemented by way of sequentially numbered opaque sealed envelopes. The four groups were independent, and each individual group was prevented from knowing the existence of the remaining groups. For this purpose, intervention was carried out on different days. All groups were equivalent, thus they did not show statistically significant differences in age, gender or BMI (Figure 1).

For 8 weeks (24 one-hour sessions), Group ET engaged in an intervention program based on Epstein's TARGET strategies taught by a physical education teacher trained to that effect. During the same time interval, Group NET carried out their usual programmed training sessions without implementing Epstein's TARGET strategies. Also during the above time frame, Group AE carried out the same scheduled PA sessions, but without the intervention of a physical education teacher. Lastly, the control group carried out placebo physical exercise, defined as 3 days a week of stretching and flexibility exercises for 60minutes per session. The participants were randomly assigned to the experimental groups or control placebo exercise group. In this sense, the absence of attrition rates is explained by a study-credit awarded to students for their participation, and adherence of the participants motivated by benefits program.

The intervention program designed for the ET experimental group was adapted from the model developed by successive authors (Cecchini et al., 2001; Duda & Balaguer, 2007) to generate a task-involving motivational climate within the six TARGET dimensions described below:

Task. The instructor used open or predominantly perceptive tasks that were presented globally, in the form of problem-situations. At all times, teachers attempted to provoke interest in PA, using varied, amusing, and motivating proposals.

Authority. The instructor took into account the university students’ interests, and the students participated actively in the decisions. This type of learning is largely determined by the student's perception of causative locus, choice, and free will.

Reward. Individual progress was rewarded or recognized by the instructor. They made sure that all the subjects were treated equally, regardless of their initial level, and that they all had the same expectations in relation to personal success.

Groupings. Small, cooperative, flexible, and heterogeneous groups were established. Learning, personal improvement, teamwork, and a class climate promoting positive social interactions were encouraged.

Assessment. The achievement of individual and collective goals was evaluated as a function of personal improvement criteria.

Time. Lastly, the students were allowed to participate in the decision-making process related to learning rhythm and the time assigned to each task.

Two teachers supervised the ET and NET groups in this study, however, the AE and placebo groups did not require teacher-participation. Each teacher was randomly assigned to either the ET or NET experimental group (Montero & León, 2007), and each one remained with their assigned group. The participant instructors had at least 5 years of prior teaching experience and also agreed to attend a specific seminar to train in TARGET strategies before starting the implementation phase. The seminar consisted of 20hours of theory and 10hours of practical training, and included pre-designed lessons, which were subsequently analyzed and discussed. While carrying out the intervention program, the instructor had to attend a weekly meeting with the research group. One teaching session from each participant group was recorded each week in order to check the fidelity of the process. For this purpose, a list of interventions coded for each group was established. Researchers and teachers did the coding independently of each other. Recordings were checked by the research group in order to guarantee that the sessions proceeded according to the established program. Ultimately, a combination of training seminars, pre-designed sessions, and analysis of videos to provide adequate training was used (Braithwaite et al., 2011). In addition, the progression of the experiment was closely monitored by the study's chief investigators.

All the data were analyzed using the SPSS 19.0 program (IBM, Chicago, IL). In all three experimental groups, PA was of moderate-to-vigorous intensity and was monitored by the research group in each and every session so that it would be adequately balanced (e.g. schedules and development of the sessions). For this purpose, the activities were classified as a function of their intensity, and this was multiplied by the duration of the activity. Student's t-tests and paired t-tests were used. Hedges’ g was selected as a measure of the effect size to provide an estimation of the effect due to the size of the samples. The calculations of the power of the sample size were based on G * Power, 3.1. In the analysis of clinical efficacy, the assessment of the effects of the treatment on depressive symptom scores and SDI was based on generalized estimation equations (GEEs) for repeated measures. The model was adjusted to allow for participants’ age, gender, and body mass index (BMI). In the six months between post-test 1 and post-test 2, the participants were allowed to practice PA in their spare time if they wished to. Participants were classified in two groups: (A) those who performed PA, at least moderately, following the criteria established in the International Physical Activity Questionnaire (IPAQ); and (B) those who had not met the minimal criteria of moderate PA. Assessment of the effects on the depressive symptom scores (post-2) and SDI was based on GEEs for repeated measures.

None of the questionnaires had more than 8% missing data, so none were eliminated; 10 questionnaires had less than 8% missing data and were therefore randomly imputed with values derived from a multiple regression in which the scores of three items from the same series of indicators (that is, the items that measure the same construct) were used as predictor variables (Byrne, 2008).

Of the 1975 preselected participants, 5.36% were randomly assigned to the treatment. All those who began treatment finished it and attended at least 22 sessions (91.7% of the total treatment). Most of the randomly assigned participants were female (64.16%). Before the intervention, no significant pretest group differences were observed in depressive symptoms (p=.98) or in the SDI (p=.16).

Significant improvements in depressive symptoms were observed between pre-test and post-test 1 in all the groups (ET, NET, and AE, p<.001; control, p<.01) and between pre and post-2 in the experimental groups (ET, NET, and AE, p<.001; control, p>0.1). In the ET group, significant changes were also observed in the SDI between pre-post-1 and pre-post-2 (p<.001). In addition, no significant differences between post-1 and post-2 were observed in the ET group, but they were found in the remaining groups (Table 1).

Likewise, statistically significant differences were observed in depressive symptoms at post-1 and post-2 between the ET group and the rest of the groups (p<.001). All groups showed reductions in symptoms during the 8 weeks of treatment. Six months later (at Post-2), only the ET group remained stable (Figure 2). No significant differences were observed between the NET and AE groups either at post-1 (p=.78) or at post-2 (p=.52), but differences were found between these two groups and the control group, both at post-1 (NET, p<.01; AE, p<.01), and at post-2 (NET, p<.01; AE, p<.05). Regarding the SDI, significant differences were observed at post-1 and post-2 between the ET groups and the rest of the groups (p<.001). No significant differences were observed between the rest of the groups either at post-1 or at post-2 (Table 1).

Figure 2.

Depressive symptom by treatments.

Note: ET=TARGET-based intervention program, NET=intervention program without implementation of TARGET strategies, AE=program under conditions of active exercise.

(0.06MB).

Table 2 shows the GEE for repeated measures at post-1 and post-2 in depressive symptoms. More than 59% of the ET group showed at least a 50% reduction of their initial score at the beginning of the study, and their response to treatment had improved by a further 10% six months later. The responses of the NET and AE groups were similar to each other. The SDI predicted depressive symptoms (post-1 and post-2). Age (post-1, p=.99, post-2, p=.74), gender (post-1, p=.94, post-2, p=.10), and BMI (post-1, p=.26, post-2, p=.12) were not significant.

As shown in Table 3, participants who practiced moderate to vigorous PA (post-2) were less prone to have depressive symptoms than those who did not. Most of these participants belonged to the ET group (n=17), followed by the NET group (n=13), AE (n=12), and control (n=9), p<.001. The SDI also predicted lower levels of depressive symptoms. Age (p=.65), gender (p=.57) and BMI (p=.06) were not significant.