The present study was conducted between May and July 2021 as a pilot project to introduce the Leicht Bewegt intervention for office workers at the German Cancer Research Center in Heidelberg (DKFZ), Germany. Ethical approval was obtained by the ethics commission of the Faculty of Behavioral and Cultural Studies of the University of Heidelberg (protocol number: AZ Schm 2021 1/1). The corporate health management of the DKFZ sent an email to all employees (around 3,200) with information on the study to recruit study participants. Study participation was voluntary and independent of any work-related benefits. Furthermore, participation in all study-related content such as workshops or questionnaires was allowed to occur within working hours. As inclusion criteria, we defined fluency in German, employment with the DKFZ, and a regular DKFZ-workplace in Heidelberg, Germany. Since the program is designed as a group intervention, we asked employees to enroll directly as a group of about ten individuals including one designated active champion. Based on the availability of 30 activity trackers and the results of our power analysis (see below), we decided to form two cohorts (each with 30 participants) that began the study only one week apart.

Initially, six groups with ten participants each registered. After receiving the first six complete group registrations, we ended the registration phase. Five individuals dropped out due to illness or (spontaneous) absence before the start of the study, so that the six groups consisted of eight to ten members. A total of n = 55 employees, working in the administration and research sections of the DKFZ, were finally registered in the study (83.6% female). Most participants indicated to be full-time workers (60%) and were between 46 and 65 years old (45.5%). A majority of participants (62.3%) reported to alternate between home office and in-office presence work.

The assignment to the two cohorts was intended to be completely at random. However, based on participants planned absences (that they had to indicate during registration), this would have led to a considerable number of data gaps. Therefore, we decided to perform cohort assignment manually to allow as many participants as possible to fully participate in the study. Informed consent was obtained from all participants included in the study.

Our single arm pre-post-follow-up design consisted of an initial four-day objective measurement period of sedentary behavior that was immediately followed by a baseline questionnaire on the fifth day (T0). The Leicht Bewegt intervention (see below) was implemented in week 2 and 3. In week 3 also the second four-day objective measurement of sedentary behavior was conducted, immediately followed by the evaluation questionnaire (T1). The reflection workshop was organized in week 5 and in week 7 participants were asked to complete the follow-up questionnaire (T2). The intervention procedure is shown in Fig. 2.

Every participating group designated a voluntary active champion who was coached in an one-hour online training by a member of the study team to represent the group’s main organizer and motivator during the study period. The intervention started with a kick-off workshop guided by the active champion. The workshop lasted approximately 60 min and conveyed information on sedentary behavior and associated risks. In addition, group-specific action strategies for reducing sedentary behavior during working hours were elaborated and three of them determined in a voting process. The strategies could relate to the physical environment (e.g., moving the coffee machine to a more distant location), organizational processes (e.g., introducing meetings with standing as default) and/or individual behaviors (e.g., setting movement reminders on mobile phones; see Online Resource 1 for an overview of all elaborated strategies). Participants were also handed out table displays that included 50 pages of tips, reasons, and helpful links for sitting less and moving more at work as well as blank pages for personal notes. Four motivational emails were sent to participants during the two weeks of intervention. Those contained a picture of people or things that symbolize a dynamic lifestyle together with an attention-grabbing question or sentence (for example: “That little bit of standing up doesn't help, does it?!”) and a short explanatory text to go with it (see Online Resource 2 for an example).

The reflection workshop one week after the initial intervention, again led by the active champion, lasted approximately 20 min and was intended to review the implementation phase. In this regard, feasible and effective strategies to reduce sedentary behavior were strengthened and ineffective ones were optimized to implement less sedentary working habits in the long term.

To check the planned procedure and surveys regarding feasibility and comprehensibility, we conducted a small pretest of the intervention with 11 participants of one DKFZ-department prior to the regular project. This led to small adaptations regarding item formulations and the intervention procedure.

Unless otherwise specified, items and scales were derived from the HAPA (Schwarzer 2007), which was applied in several studies (Schwarzer 2008), and partly adapted to the target behavior and context of this study.

HAPA stage detection To capture the current HAPA stage of the participants, namely pre-intender, intender, or actor, the participants were asked whether they consciously reduced sitting time lately. For the baseline questionnaire, the response options were “No, and I don't plan to” (pre-intender), “No, but I am thinking about it”, (pre-intender) “No, but I have the firm intention to do so” (intender), “Yes, but only recently” (actor), and “Yes, for a long time already” (actor). In the evaluation questionnaire, the last two response options were changed into “Yes, but only as a result of the Leicht Bewegt program” and “Yes, even before the Leicht Bewegt program” to identify a potential stage change as an intervention effect.

Intention Intention was measured with three items asking for the intention (a) to sit a smaller proportion of the workday in the upcoming weeks, (b) to regularly interrupt sitting phases during work in the upcoming weeks and (c) to integrate more activity into their daily work routine in the upcoming weeks. These had to be answered on a seven-point Likert scale from 0 (“don’t agree at all”) to 6 (“fully agree”). An additional item asked for the subjective probability to turn their intentions into action in the upcoming weeks on a percentage scale from 0 to 100%. A mean score was calculated by multiplying the fourth item by six, dividing it by 100 and then summing all four items and dividing them by four (Sieverding et al. 2010). Cronbach’s α was 0.78 (T0) and 0.65 (T1).

Task self-efficacy Task self-efficacy was assessed with three items that asked to rate how confident participants were in their ability to implement above mentioned actions that were queried for the assessment of intention (e.g., “I am sure that I can interrupt sitting phases during work after 30 min at the latest.“). Response options ranged on a four-point Likert scale from “clearly disagree” (= 0) to “clearly agree” (= 3). Cronbach’s α was 0.57 (T0) and 0.67 (T1).

Risk perception Risk perception was assessed with three items asking for risks from prolonged sitting with regard to (a) becoming chronically ill, (b) to suffer from acute or chronic pain, or (c) to develop cardiovascular disease. Each risk had to be rated on a five-point Likert scale from “much below average (= 0)” to “much above average (= 4)”. Cronbach’s α was 0.84 (T0) and 0.88 (T1).

Outcome expectancies Participants were asked to specify whether they expect positive outcomes (i.e., having more energy, feeling more at ease, and suffering from less back pain) or negative outcomes (i.e., completing fewer tasks, losing focus on work, and losing time) to happen when significantly reducing and frequently interrupting sitting phases. Answers had to be given on a four-point Likert scale from “clearly disagree” (= 0) to “clearly agree” (= 3). Cronbach’s α was 0.64 (T0) and 0.56 (T1).

Maintenance self-efficacy To measure maintenance self-efficacy, we asked the participants to rate their confidence to deal with six possible barriers for long-term reduction of sedentary behavior during work (i.e., lack of visible changes, people in the work environment who are indifferent to sitting times, a long period of habituation, situations that trigger old sitting habits, a stronger desire to sit, and a high degree of effort to change habits). Response options ranged on a four-point Likert scale from “clearly disagree” (= 0) to “clearly agree” (= 3). Maintenance self-efficacy was not included in the baseline questionnaire. Cronbach’s α was 0.86.

Action/coping planning To measure action planning, participants were asked whether they planned in the last two weeks how, when, how often, and with which strategies to reduce sitting times. For coping planning, the participants were questioned whether they had planned how to continue to reduce sitting times during the last two weeks despite feeling restricted in terms of health, feeling tired or listless, having an unusually high amount of job tasks, or failing in reducing sitting times for a few days. Response options ranged again on a four-point Likert scale from “clearly disagree” (= 0) to “clearly agree” (= 3). Action/coping planning was not included in the baseline questionnaire. Their Cronbach’s α was 0.74 and 0.86, respectively.

Subjective measurement To assess participants’ subjective perception of their sedentary behavior during work, participants were asked to indicate how periods of sitting, standing, and walking were distributed as a percentage of a typical workday over the previous seven days. In the follow-up questionnaire, the previous three weeks were considered. In addition, participants were asked to indicate how often they interrupt their sitting time within one hour of a typical workday (from 0 to 5 times or more).

Objective measurement For objective measures of sedentary behavior, activPAL 3 (Pal Technologies Ltd., Glasgow)—inclinometers with high validity (O'Brien et al. 2022)—were used. ActivPals are continuously attached to the center of the front of the thigh using the transparent and hypoallergenic 3 M Tegaderm film patch. They register the inclination of the thigh and can distinguish between sitting/lying, standing, walking, sit-to-stand and stand-to-sit transitions and step counts (Aminian and Hinckson 2012). During the phases of objective measurement (see procedure), participants were instructed to wear it during four consecutive days. The results of the measurements were not visible to the participants. After the measurement phases, recorded files were downloaded and analyzed using the device-specific software PALanalysis.

For each measurement day, the sum of the proportion of working hours spent walking, standing, sitting, and interruptions of sitting times were calculated. To exclude non-working hours, individual working hours were previously requested in the baseline questionnaire and all tracker records outside working hours were ignored in the analysis. In addition, the first measurement day in both measurement periods was excluded from the analysis to minimize a bias of possible habituation effects due to the unfamiliar tracker wearing.

Satisfaction rating To capture participants’ experiences and opinions regarding the intervention, they were asked whether they had changed their sitting behavior, whether they would like to implement the elaborated strategies in their everyday work in the future, and whether they would recommend future participation in the intervention to others. Response options were "no,” “rather no,” “rather yes,” and “yes.”

Other measures included participants’ age (range of 18–29 years, 30–45 years, or 46–65 years), sex, their work circumstances (i.e., part-time, full-time, or freelance as well as working from home, in presence in their office, or alternating between those two options), and their general health based on the Short-Form-Health-Survey-12 (SF-12; Ware et al. 1996).

We performed further data processing and analysis via the statistical software IBM SPSS Statistics 27. A p < 0.05 was considered as statistically significant. Descriptive statistics were used to describe sociodemographic characteristics of the study population, their activity behavior and their satisfaction with the intervention. To check the significance in changes of subjective (T0 to T1; T0 to T2) and objective activity measures (T0 to T1) as well as in motivational HAPA variables (T0 to T1), we used paired t -tests. We calculated Cohen’s d (Cohen 1988) to indicate effect sizes for t -tests (small effect: d ≥ 0.2, medium: d ≥ 0.5, and large: d ≥ 0.8). Pearson correlation coefficients were used to analyze bivariate correlations between motivational HAPA variables (T1) and intention (T1) as well as between volitional HAPA variables (T1) and parameters of subjective activity (T2). In order to estimate the required sample size for the planned pre-to-post comparisons in our within-subjects design, we used G*Power (Faul et al. 2009) for a power analysis. Assuming a 5% α-level and a power of 90%, the required minimum sample size was n = 36 to detect at least medium-sized effects (Cohen’s d = 0.50) –which is based on findings of the effect on sitting times of the Australian predecessor version of our Leicht Bewegt intervention (Healy et al. 2018).