Introduction

Navigating through our environment, we are frequently exposed to situations that can be evaluated as stressful, and therefore have a significant impact on our health. With regard to the health effects of stress, a distinction needs to be made between chronic and acute stress. Chronic stress is defined as long-term experiences of adverse psychosocial conditions, such as unemployment, occupational stress, difficult interpersonal relationships, or having to care for a family member with a chronic disease [1]. Chronic stress exposure has been shown to be a predictor of depression, cardiovascular disease, and other non-communicable diseases [2]. During acute stress, an experience is interpreted as a threat by the central nervous system (CNS). Consequently, both physiological responses and changes in behavior are initiated, leading to allostasis and adaptation. Orchestrated by the CNS, the body mounts a short-term stress response to provide energy, enabling individuals to cope with challenges [3]. Over time, repeated activation of stress systems can lead to accumulation of allostatic load, in which excessive exposure to mediators of endocrine, neural, and immune stress can have negative effects on various organ systems and thereby contribute to diseases [3].

Many stressors in modern societies are characterized by their repeated, acute nature, such as interruptions by digital technologies [4]. Therefore, in addition to investigating chronic stress, it is important to understand the individual response of human beings to acute stress, which is typically examined in laboratory settings using standardized stress protocols [1].

The considerable personal and financial resources required for these protocols limit their widespread use and impede a more comprehensive understanding of the health effects of acute stress. To address these disadvantages, new digitized and automated protocols are being developed, potentially allowing for more cost-effective and less time-consuming implementation. However, these new protocols have not yet been systematically investigated and compared. Consequently, the aim of this systematic review is to provide an overview of existing digital stress induction protocols, to compare them in terms of their potential to activate the two major physiological stress systems—the sympatho-adrenal-medullary (SAM) axis and the hypothalamic–pituitary–adrenal (HPA) axis—and to evaluate the psychological and physiological stress responses they elicit, based on reported effect sizes.

A number of laboratory protocols have been developed that all have the aim to induce acute stress in human research participants and under laboratory conditions. Among them, the Trier Social Stress Test (TSST; [5]) is considered the gold standard. It reliably activates the HPA axis, the sympathetic nervous system (SNS), and its SAM components [6,7]. It also affects further downstream systems, such as the cardiovascular, immune, and metabolic systems [8]. Several adaptations exist, including the group version (TSST-G; [9]), the child version (TSST-C; [10]), and non-stress control variants such as the friendly TSST (f-TSST; [11]) and the placebo TSST [12]. These modified versions differ mainly in their social-evaluative and uncontrollability components, resulting in weaker HPA axis activation while still stimulating the SNS. The TSST contributed significantly to developing a more comprehensive understanding of psychological and physiological acute stress responses and was also shown to be the most successful protocol for HPA axis activation [6].

Besides the TSST, a number of other established stress induction protocols have been developed, including the Trier Mental Challenge Test (TMCT; [13]), the Montreal Imaging Stress Task (MIST; [14]), the Stroop Color-Word Interference Test [15], the Cold Pressor Test (CPT; [16]), the Socially Evaluated Cold-Pressor Test (SECPT; [17]), the Maastricht Acute Stress Test (MAST; [18]), the Paced Auditory Serial Addition Task (PASAT; [19]), the challenge test [20], noise stress [21], and the Mannheim Multicomponent Stress Test (MMST; [22]). These protocols provide a comprehensive framework for examining the physiological, psychological, behavioral, emotional, and cognitive consequences of acute stress in humans [23].

While most of these protocols have been successfully used to elicit psychological and physiological stress responses, they also come with a number of disadvantages. One of the main limitations is their high demand for resources, including the availability of testing facilities, personnel capacity, and participant time, as well as the requirement for participants to travel to a laboratory setting instead of completing the test at home or in a work environment. Hence, most of these protocols are uneconomical and additionally require considerable personal and financial resources. Furthermore, studies that are carried out with the help of human research assistants could be burdened by experimenter effects [24]. Therefore, it appears to be essential in the context of future research to minimize the resource requirements. A promising solution to these problems could be electronic, virtual or otherwise automated stress protocols.

For better generalizability of results, it could also be beneficial to induce acute stress in the natural environment of the tested individuals in addition to the laboratory setting. This dual approach would enable more effective exploration of how surrounding environments influence both physiological and psychological stress response. It would also facilitate a clearer understanding of the causal relationships between environmental components and the physiological and psychological stress response.

While some automated stress protocols have been proposed and tested, it needs to be formally established whether automated stress protocols reliably elicit a psychological and/or physiological stress response. Questions also arise regarding whether these responses are comparable in intensity and duration to those induced by conventional stress protocols. Additionally, there is a need to identify which individual components of digital stress tests are particularly effective in eliciting stress responses in the tested participants.

A number of different automated and/or digitized versions of the TSST are already being used for a wide variety of research purposes. The virtual TSST (V-TSST) versions are completely digitized adaptations of the regular TSST by Kirschbaum et al. [5] using virtual reality (VR) settings such as 3D projections or immersive headgear.

In 2019, Helminen et al. [25] conducted a meta-analysis of 13 studies which utilized the V-TSST for stress induction. The authors measured cortisol levels before test administration and at peak stress. They found a medium average effect size (ES = 0.65) for the increase in cortisol from baseline to peak value. However, despite effectively inducing psychosocial stress according to this review, the cortisol response elicited by the V-TSST was lower compared to the conventional TSST [25]. In a systematic review with meta-analysis, Helminen et al. [26] conducted a comparison between studies utilizing V-TSST protocols and those employing the traditional TSST. According to the authors, studies using the V-TSST demonstrated stress (responses in cortisol, heart rate, self-reported stress) that were comparable to those observed in traditional TSST studies.

The TSST is not the only stress protocol which has been digitized. The goal of the Digital Stress Test (DST; [27]), a web-based protocol, was to provide a scalable digital tool which allows standardized induction and recording of acute stress responses outside of the laboratory without contact to an experimenter. The DST was developed on the basis of well-researched protocols for stress induction. Participants who completed the DST showed significantly higher perceived stress indices, measured by visual analogue scales [28] as well as by the international Positive and Negative Affect Schedule Short Form [29], than the participants who were tested with a control condition of the DST on all post-baseline measurements. Similar effect sizes for negative affect were measurable in the participants who underwent the DST as in the participants who were tested in conventional TSST studies [27].

Materials and methods

Discussion

In this study protocol, we propose a systematic review that will analyze existing digital stress tests—defined as fully automated and standardized procedures utilizing digital technologies to elicit stress and/or similar psychological states—by examining their effectiveness in triggering acute stress responses, assessed through physiological and/or psychological parameters. Specifically, we will focus on the potential of these digital protocols to activate the two main physiological stress systems, the SAM axis and the HPA axis, as well as to induce self-reported stress or related psychological responses such as negative affect.

As a first step, we will screen major web databases, including PubMed/MEDLINE, PsycINFO, Web of Science, Scopus and CENTRAL, for studies proposing or examining digital stress tests, as well as digitized versions of established conventional stress tests. While previous meta-analyses have addressed specific digital protocols such as the V-TSST and its variations [25,26], to our knowledge, no comprehensive systematic review or meta-analysis has yet compared all types of digital stress tests designed to elicit stress and/or similar psychological states autonomously in digital environments without experimenter intervention.

For the purpose of this review, digital stress tests will be defined as interventions conducted automatically according to standardized, predetermined protocols in either controlled or real-world settings using digital technologies such as virtual reality environments, computers, smartphones, or similar devices. Although experimenters may engage with participants before and after the test, the stress induction itself must proceed entirely autonomously, without experimenter interference.

Our systematic review aims to provide a detailed overview of which conventional stress protocols have been successfully digitized and which have been developed as fully digital from the outset. Moreover, we will examine to what extent these digital protocols effectively elicit stress responses, specifically regarding activation of the SAM and/or HPA axis and self-reported psychological stress. By synthesizing effect sizes across studies, we intend to evaluate and compare the relative effectiveness of these digital stressors on both physiological and psychological levels.

Findings from this systematic review will be disseminated through publication in a peer-reviewed open-access journal and presentation at relevant academic conferences. A preprint will be posted to ensure early accessibility. All supplementary materials—including data extraction templates and any analysis scripts—will be publicly available via our preregistration project on the Open Science Framework (OSF). This repository will be updated with data and additional documentation at the time of publication.

Supporting information