The decline in physical activity (PA) and the rise of sedentary work are strongly related to desk-based occupations as many employees spend most work hours sitting (1–4). Low activity levels contribute to decreased productivity, higher absenteeism, and increased healthcare costs (4, 5). Hence, the World Health Organization (WHO) recommends reducing sedentary behavior and increasing PA (6) to significantly counteract the risk for cardiovascular disease, diabetes, obesity, and mental health disorders (2, 7–11).

To address these challenges, workplace interventions have gained attention. Organizations increasingly adopt workplace health initiatives to promote PA and thereby aim to increase productivity, job satisfaction, and reduce absenteeism (12–16).

In this context step challenges, particularly those incorporating feedback and social comparison, have shown high potential (17–24). Team incentives may further enhance engagement (14, 25–29). Previous research on step challenges has shown that both financial incentives (26, 30) as well as non-monetary rewards increase PA in short-term but long-term sustainability is uncertain (31, 32). Outbalancing incentive-removal post-intervention on the one hand and increased awareness of PA benefits and habit formation on the other seem to be decisive for sustained improvements. Further, randomized controlled trials indicate that intervention success strongly depends on design and participant characteristics (33–35). Within the specific setting of higher education, a “healthy university concept” (36) includes PA promotion. Setting-specific team-based competitions with incentives have shown very promising increases of up to 4,800 steps per day (30). Besides, the ARK project in Norway demonstrated the effectiveness of holistic health promotion programs in higher education (36), and a systematic review of 17 studies found positive health outcomes in PA, weight management, and nutrition (37). However, diverse employee demographics and organizational structures present unique challenges for workplace interventions (38, 39) with time constraints, workplace cultures, and the complexity of integrating such programs into daily routines representing typical barriers (40). Besides, gender and age differences may influence changes in PA behavior. Men tend to be more motivated by competition, whereas women prioritize fitness and appearance (41, 42), and gender specific differences in motivation for PA diminish with age (43). Younger individuals are motivated by health and goalsetting (44). Also, factors like lifestyle, education and job characteristics influence activity levels (40, 45). Additionally, in the context of workplace PA interventions, leadership may influence behavior changes (16, 46, 47). Leaders shape organizational culture and often serve as role models (15, 48–50). Notably, participation in health interventions may also contribute to improvements of leadership (15, 51).

However, few studies have explored the role of leadership in the context of workplace PA interventions involving self-formed teams without designated intervention leaders. This gap is especially relevant in higher education institutions, where leadership structures differ significantly from those in the private sector. At universities, leadership is often distributed and exercised through indirect forms such as mentorship, coordination, and project management rather than through formal, hierarchical authority (52, 53). Understanding these unique leadership dynamics is critical for evaluating PA promotion interventions in academic environments. Additionally, recognizing personal and work-related factors seems crucial for designing effective workplace interventions for PA promotion. Therefore, this study examines the effectiveness of a six-week team step challenge using a gamified mobile app in increasing daily step counts among employees at a German university. Further this study explores individual, work-related, and team-related influencing factors. A particular focus is placed on the role of leadership, examining whether individuals in leadership roles show different outcomes, and whether their presence influences team performance.

We hypothesize that a non-monetary, team-based step contest will increase daily step counts compared to baseline levels, but we expect step counts to decline post-intervention. In addition, we examine how intervention effects relate to individual characteristics, and workplace dynamics. In addition, we expect individuals with leadership responsibilities to demonstrate greater engagement, and explore whether their presence affects overall team outcomes.

Figure 1 visualizes dynamics of daily step counts throughout the study. Average daily step counts increased from 8,955 steps/day at baseline to 10,653 during the intervention, declining to 8,514 in follow-up (see Table 1). A one-way ANOVA using the logarithm of the step variable identified a significant main effect of the intervention [F (2, 660) = 42.15, p < 0.01, η 2 = 0.11]. Post hoc pairwise comparisons revealed that step counts were significantly higher during the intervention compared to baseline (p < 0.01) and follow-up (p < 0.01). The difference between baseline and follow-up was not statistically significant. A post-hoc one-way ANOVA showed a significant main effect of treatment weeks on step counts (F (5, 1,650) = 17.44, p < 0.01, η 2 = 0.05). Post hoc comparisons indicated a significant decline in step counts from the first to the later weeks of the intervention.

The mean age was 35.07 (±11.04) years. ANOVA results indicated no significant main effect of age on step counts (see Table 2), but older participants (55–65 years) exhibited the largest relative increase in daily step counts. Among participants, 62% were female (see Table 3). Baseline step counts were comparable between males and females, though males exhibited a greater increase during the intervention (see Tables 1, 2).

More than 72% of participants reported engaging in PA multiple times per week during their free time (see Table 3). Descriptive statistics indicated that active participants walked more across all periods (see Table 1). A significant main effect of PA in free time on step counts was observed, though no significant interaction effect with the intervention was found (see Table 2).

A key focus of this study was the role of leadership. Data from the university’s HR system indicated that in May 2025, 423 individuals in scientific positions held officially designated leadership roles, collectively responsible for 3,846 employees (excluding student assistants). In scientific support, 62 leadership positions were identified, overseeing 871 employees. Leadership roles cover a wide range of functions. In the scientific domain, these include professorial leadership, research group leaders, spokespersons of major research projects, deans, and central facility directors. In scientific-support, leadership roles encompass department and unit heads, project managers, heads of technical and administrative services, and coordinators of specialized services such as family support or equal opportunity offices. Among study participants, approximately 20% reported having personnel responsibilities at work (22.79% among scientific staff, 19.13% among science support staff) (see Table 3). While this factor did not exhibit a significant main effect on mean daily step counts, a significant interaction effect with the intervention was identified (see Table 2). Individuals in leadership roles demonstrated a significantly greater increase in daily step counts. 75% of participants were in teams with at least one member holding personnel responsibilities (see Table 3), yet the presence of such individuals within a team did not significantly influence the challenge’s effectiveness in increasing daily step counts (see Table 2). 37% of participants were in teams with one leader, 23% with two, 8% with three, and 7% with four leaders. The maximum proportion of team members with leadership roles was 50%.

Results suggested a trend, with participants living closer to the university walking more across all three periods (see Tables 1, 2). However, the ANOVA results revealed no significant interaction effect of commuting distance with the intervention (see Table 2). Similarly, means of transport for commuting had a significant main effect on step counts, but we found no significant interaction effect with the intervention (see Table 2). Bonferroni-adjusted pairwise comparisons revealed significantly lower step counts for car users compared to public transport users and those walking.

PA at work did not significantly affect steps, nor was there a significant interaction effect with the intervention (see Table 2). However, descriptive statistics suggested that individuals with higher PA tended to have higher step counts (see Table 1).

Participants were nearly evenly split between science staff and science support roles (see Table 3). While both groups had similar baseline step counts, science support staff showed a greater increase in daily steps during the intervention (see Tables 1, 2). In contrast, simple main effects analysis did not indicate a significant impact of work area on step counts.

First-time step challenge participants, who comprised the majority (76%) of the sample (see Table 3), walked significantly less in all periods compared to others, and the intervention was more effective among those with prior challenge experience (see Table 2). Additionally, prior participation was positively correlated with working in science support roles (p < 0.01) and negatively correlated with joining the challenge due to persuasion by colleagues (p < 0.01). There is a trend suggesting that those in leadership positions may be slightly more likely to have participated before (p = 0.09) (see Table 2).

The reason for signing up and team composition criteria did not significantly influence steps, nor did it impact intervention effectiveness (see Table 2). However, descriptive statistics indicated that those who reported to form teams due to exercise habits showed higher step counts across all periods (see Table 1).

Team size ranged from 3 to 15 members. ANOVA results did not show a significant relationship of team size to step counts (see Table 2). Descriptive statistics indicated that smaller teams showed the smallest increase in daily steps while teams with 8–10 members showed the largest increase in daily steps (see Table 1). However, all groups experienced a decline in steps during the follow-up period.

Our study provides interesting insights into step count trends during a step challenge intervention. Participants walked more than double the average German population (57), reaching approximately 10,700 steps/day during the intervention. Even pre-intervention, participants were already quite active, averaging around 9,000 steps/day, consistent with research showing that workplace interventions often attract physically active individuals (34). Indeed, the majority of participants reported engaging in sports multiple times per week. While research suggests that less active individuals benefit more from PA promotion programs (34, 35), such interventions often fail to engage this group. Our findings highlight this participation bias and underscore the need for targeted strategies to attract less active employees.

The challenge led to an increase of about 1,700 steps/day, demonstrating its effectiveness in promoting activity. This rise aligns with previous workplace PA interventions (18, 21, 24) including team-based interventions (14, 19, 20, 25, 26, 29), though it is lower than the 4,799-step increase observed in a team competition with financial incentives (30). This smaller effect may be due to a higher baseline (8,955 vs. 5,959 steps/day), the absence of a daily step goal, or the lack of monetary rewards. Nonetheless, our findings confirm that team-based challenges can still drive meaningful behavior change. Notably, our step increase surpasses the modest effects reported in some studies (18–20).

Participants showed strong initial engagement, likely driven by novelty, followed by a gradual decline, suggesting potential fatigue or being incompatible with daily duties over longer terms. This pattern underscores the importance of monitoring and addressing factors that may contribute to decreasing activity over time. After the challenge, step counts dropped, indicating that while short-term interventions effectively boost activity, maintaining long-term habits may require additional strategies. Step counts were even lower than before the challenge, though not statistically significant, possibly due to recovery after an intense final push or a return to typical activity levels if participants had increased their steps during the baseline period in preparation for the challenge.

By utilizing the multi-campus higher education environment, marked by a diverse range of stakeholder roles and operational goals (38, 39), our study captures the influence of team-based competitions on PA within a heterogeneous organizational context. This context provides additional insights into variations in activity levels and the effectiveness of workplace interventions.

Sociodemographic differences were evident. Our results suggest that female work forces are more likely to participate in team-based step contests. However, males exhibited a greater step increase, likely due to their higher competitiveness (41–43, 58, 59). Age influenced the effectiveness of the intervention with older participants showing the highest relative step increase. This aligns with research showing that older adults are more driven by extrinsic motivation in PA (27, 41), while younger individuals may respond better to goal setting (44). Our study did not include a daily step goal, which has been a key motivator in other studies (34). Further, our findings do not indicate that family responsibilities hinder PA aligning with prior research (19, 60). In general, higher self-reported PA in free time positively correlated with daily step counts aligning with prior research (19, 45). However, while previous research suggests that less active individuals benefit more from such interventions (24, 34, 35, 61), our study does not support this for the university setting.

In addition, work-related characteristics played a substantial role. Our analysis confirms results of previous research showing a link between commuting and PA (42, 60), with car users walking less. Besides, working from home showed a trend of lower steps, possibly due to reduced incidental movement (3) but the results were not significant. As suggested by prior research, we found differences between academics and science-supportive employees (40). Participants in academic or scientific roles exhibited a lower step count increase compared to science support staff. Science support staff were also more likely to have participated in a previous step challenge, as indicated by the positive correlation with prior participation, suggesting a higher initial engagement with PA initiatives.

Additionally, leadership may play an important role in shaping behavior within workplace PA interventions (16, 46, 47). In the university context, personnel responsibility typically involves supervisory duties over staff or students. This includes tasks such as delegating work, monitoring performance, approving leave, and supporting professional development. Individuals with personnel responsibilities are often accountable for team outcomes and contribute to the functioning of academic or administrative structures. Leadership at universities is more complex than formal supervisory roles alone. It is shaped by relational dynamics, trust, and informal negotiations of influence and responsibility (52, 53). The diversity of leadership roles reflects the distributed and non-hierarchical nature of leadership in higher education. This complexity extends across both scientific and scientific-support domains, including positions such as deans, research group leaders, department heads, technical managers, and coordinators of institutional services. Such a nuanced leadership structure is particularly relevant when evaluating participation in university-wide initiatives like team-based health promotion programs. In these contexts, the presence and engagement of individuals in leadership roles may influence participation and outcomes through role modeling, team motivation, or cultural signaling (15, 16, 46, 48–50).

Data from the university’s HR system shows that approximately 9% of employees hold formal leadership roles—9.91% in scientific positions and 6.65% in scientific-support roles. In contrast, 20% of study participants reported having personnel responsibilities, suggesting that individuals in leadership roles may be more inclined to engage in such health initiatives. Among scientific staff in the study, the figure was even higher (22.79%), aligning with institutional trends. This overrepresentation suggests that workplace PA interventions may appeal especially to individuals with leadership responsibilities.

It is also possible that the discrepancy between reported and formal leadership roles reflects the unique leadership culture in universities. Given the distributed and often informal nature of academic leadership (52, 53), individuals may perceive themselves as having personnel responsibility even in the absence of formal supervisory roles. Such perceptions can still influence engagement in workplace health initiatives, as informal leadership may foster a sense of responsibility, role modeling, or team cohesion.

Taken together, these findings suggest that workplace PA interventions may particularly appeal to individuals who see themselves in leadership roles—whether formally recognized or not. Survey results support this trend, showing that participants in leadership roles were more likely to have taken part in previous step challenges.

During the intervention, these individuals also showed greater increases in step count—possibly reflecting a desire to serve as role models or to support team cohesion (48, 49). Leaders may also be more accustomed to setting and pursuing goals, making them particularly responsive to structured, gamified health programs (62, 63). Another possible explanation is that individuals in leadership roles may have experienced increased pressure to perform due to their visibility within the team. This perceived accountability could have driven them to increase their activity levels, independent of intrinsic motivation or role modeling intentions.

Despite these findings, our results indicate that the mere presence of a leader in a team did not significantly affect other team members’ outcomes. This could be because the step challenge primarily relied on individual motivation rather than leadership-driven encouragement (46). Leaders may have focused on their own performance rather than actively fostering participation within their teams. Lack of leadership support may hinder PA promotion (16, 46, 47). Research highlights that there is a need to equip leaders with knowledge to foster participation and engagement (64). To leverage leadership more effectively in workplace PA interventions, organizations should implement strategies that actively involve leaders in promoting engagement within their teams. These insights highlight the importance of understanding how leaders’ PA behaviors may influence the effectiveness of workplace PA promotion interventions as pointed out in prior research (15, 38, 47).

Although some teams included multiple individuals with leadership responsibilities, there were no teams composed exclusively of leaders. The maximum proportion of leadership roles within a team was 50%, suggesting that leaders did not form separate, leader-only teams but were instead integrated into mixed-role groups. It is also plausible that some teams included multiple leadership levels (e.g., institute heads, professors, and research group leaders).

Further, participants with prior experience in step challenges had higher baseline step counts and demonstrated a greater increase in activity levels, highlighting a potential predisposition toward competition or intervention-driven PA – an observation that aligns with previous research (25).

Team composition might also play a role. Participants that joined teams based on shared exercise habits showed on average higher step counts across all three study periods. Team size ranged from 3 to 15 members, similar to earlier studies (19, 26). Teams with moderate size exhibited higher step counts.

Despite these insights, limitations must be acknowledged. Self-selection may have introduced bias and the reliance on participants’ step-tracking devices and self-reported data could affect the accuracy of the obtained data. Additionally, other forms of PA (e.g., cycling, swimming, yoga, aerobics) were not captured, as the app only recorded walking—i.e., step-based movement detected by smart devices. The predominantly female, office-based sample and short follow-up period also limit generalizability and long-term conclusions. Furthermore, it is unclear whether the teams in the study were strictly work teams, and whether the individuals reporting to have personnel responsibilities were formal, work-related leaders of their team colleagues. Finally, while we excluded extreme step values (<1,000 or >35,000 steps/day) to reduce potential measurement error, this decision may have led to the omission of valid but rare activity patterns. However, a robustness check including these values yielded comparable results, suggesting limited impact on the overall findings. A key limitation of this study is the absence of a control group, which restricts our ability to attribute the observed increase in step count solely to the intervention. The challenge took place in autumn 2023, with the baseline phase occurring during the semester break and the follow-up during the academic semester. External factors such as seasonal weather changes, academic workload fluctuations, and working-from-home patterns may have influenced participants’ activity levels regardless of the intervention. Future studies should consider including a control group to better isolate the effects of the intervention.

This study highlights the potential of a six-week team-based step challenge on university employees’ PA with an average increase of almost 20% in daily step count. However, the effectively boosted short-term activity was not maintained post-challenge. Further, step count increase was modulated by personal, work-related and team composition factors. Most beneficial outcomes might be achieved for male and older individuals, for leaders and science support staff, for those with prior experience and already active subjects, while several other groups in the higher educational setting may not be attracted in a similar way.

This strengthens the notion that both diverse employee demographics and organizational structures as well as the complexity of integrating PA enhancing interventions into daily routines represent typical barriers for the higher educational setting (38, 40).

A particularly notable insight from our study is the role of leadership on PA changes. Leaders demonstrated higher increases in step counts, likely due to strong goal orientation or a desire to lead by example. However, their presence alone did not significantly trigger team-wide activity levels, suggesting that leadership alone is not sufficient to drive collective behavior change. Actively involving and equipping leaders to support and motivate their teams might enhance the effectiveness of future programs.

Future research should therefore explore how different leadership styles influence participation and engagement in workplace PA programs. Interventions that actively involve and equip leaders to support and motivate their teams may enhance the overall impact of such programs. Additionally, strategies are needed to reach the less active individuals and more sedentary groups, as well as ways to keep them engaged and motivated throughout the challenge. A more diverse participant pool would additionally enlighten our understanding of how different demographic groups engage in PA challenges. Additionally, given the rising prevalence of remote work, developing tailored interventions for this population is increasingly important. Research is also needed to identify ways to sustain PA gains.

For institutions, these results suggest that short-term, low-cost digital interventions can be effective in initiating PA improvements. However, for long-term impact, such programs should be embedded into broader workplace health promotion strategies. This includes leadership training to support team motivation, personalized approaches for different employee groups, and support for integrating PA into daily routines.

Overall, our findings highlight the complexity of influencing factors on PA behavior and emphasize the need for personalized, sustainable interventions to promote long-term activity across diverse populations and how this is related to leadership support.