Based on the aforementioned theoretical concepts, we define design primes as any formal aesthetic design feature that activates mental concepts that in turn influence situational experiences, emotions or behaviours (a.) by its mere aesthetic appearance (“style”) or (b.) by interacting with it (e.g., through principles of embodiment) (see Figure 1). In the retail sector this is commonly used for instance when fruits and vegetable are presented in wooden boxes instead of the plastics boxes they were actually transported in, to prime concepts of small-scale farms and nature as opposed to industrial farming. In paediatric hospitals, one might refer to the colourful and playful design features commonly used as design primes, since they aim at activating mental concepts of joy and playfulness as opposed to medical procedures.

While, as presented above, this concept has been implemented for decades (e.g., Packard, 1957), this paper illustrates approaches of utilising design primes in the context of health-promoting design. Therapeutic built environments such as hospitals and rehabilitation clinics are usually filled with aesthetic features that communicate to some extent a specific medical regime (Rehn-Groenendijk 2022). Whereas certain elements such as, e.g., flooring material or lighting are still mainly designed based on specific technical or functional needs (e.g., hygiene, safety) and an overall aesthetic concept, their additional subconscious effects on parameters such as clinical outcomes and health behaviour is often underestimated. Although prominent examples of the effect of design (Hernandez-Garcia et al., 2021) and in particular of aesthetic features on bio markers such as blood pressure are well established (e.g., “white coat hypertension,” Pickering, 1996), their potential use to support therapeutic processes is rarely embraced. At the same time, there is an ongoing discourse on the effect of the built environment related to by constructs and paradigms such as psychosocially-supportive design (Ulrich, 1991), healing environments (Dijkstra, 2009), evidence-based design (Malkin, 2008; Stichler and Hamilton, 2008; Halawa et al., 2020) and approaches linked to salutogenesis (e.g., Dilani, 2008). This can serve as a strong conceptual foundation to make use of these aesthetic features as subtle tools for promoting health. The multiple pathways through which these mechanisms might be effective are for instance illustrated by Salovey et al.’s (2000) work on the interrelation between emotional states and physical health and have been further elaborated by the work done in fields such as psychoneuroimmunology (e.g., Moraes et al., 2018).

We therefore propose, that formal-aesthetic design can not only contribute to technical goals (e.g., ergonomics) as well as affective responses (e.g., aesthetic appeal) but can directly address more complex mental concepts that in turn affect a number of subsequent thought patterns, emotions and behaviours. This applies in particular to the emotionally charged context of healthcare settings. In some medical fields, insights from this line of research might be difficult to implement and require small-scale interventions (e.g., Rehn-Groenendijk 2022). With regards to the specific field of orthopaedic rehabilitation, an overarching theme can be identified and addressed by design. In the German health system, which is the context of the experimental study described here, orthopaedic rehabilitation comprises a variety of pathologies. Most prevalent are postsurgical care after hip and knee joint replacement and treatments of rheumatic diseases. The hallmark of orthopaedic medicine can be seen in its focus on movement and physical activity. While most patients suffer from illnesses or pathologies that to some extent compromise physical activity (e.g., range of motion), one of the most profound improvements after therapeutic measures such as hip replacement is the almost immediate enhancement of movement ability. In line with this biomechanics approach are core perspectives of orthopaedic rehabilitation (e.g., Kulig and Burnfield, 2008). As a result, patients in orthopaedic rehabilitation can be expected to be highly sensitised towards physical activity. At the same time, sufficient motivation, commitment and health behavioural changes (e.g., nutrition and physical activity) usually are key components to achieve lasting positive health outcomes. In line with this, orthopaedic rehabilitation takes usually 3–6 weeks and comprises, in addition to specific medical treatments, educative elements to facilitate health behaviour change (for an overview see Ip, 2007). With regards to this, the built environment can have a significant complementary effect on both physiological and mental processes that can support therapeutic processes and outcomes (e.g., Ulrich et al., 2008; Ziegler, 2015; Ryan and Browning, 2020).

Starting point of the design process was the creation of eight different mood boards representing formal aesthetic styles such as “playful” or “natural” (Figure 2). These boards were presented to 125 students, who evaluated the boards using semantic differentiation (Osgood, Charles 1979), a common practise to evaluate design variations (e.g., Li, 2022). The semantic differential was based on 34 items clustered around the seven categories “performance accomplishment,” “emotional arousal,” “sense of control,” “self-regulation,” “creativity,” “attractiveness,” “competence” and “therapeutic.” This structure and selection of items were inspired by the User Experience Questionnaire (Laugwitz et al., 2008) but adapted to fit the overall research question of this project.

Based on this survey, three mood boards (3, 7 and 8) could be allocated to be labeled as “positive” (having higher scores in terms of the aboved mentioned categories), whereas three other boards (1, 5 and 6) were scored markedly lower and could thus be labeled as “negative.” Two boards were evaluated ambiguously and were therefore excluded from the further process.

In addition to these six mood boards, a number of explorative design concepts and ideas were created in an iterative design process using sketches, CAD tools and structured creativity workshops (Rehn, 2019, p. 379). In this workshop, four participants without specific prior knowledge on this topic were instructed to come up with associations, specific ideas and eventually design concepts around stimulus words such as “self care.” Usual creativity techniques (e.g., bisociation, brainstorming) were applied using sticky notes, markers and paper (cf. Braem, 1989; Noack, 2008). The ideas derived from the workshop were consolidated with prior concepts developed by author 1 and elaborated into explorative designs.

These explorative designs formed the basis of another empirical pre-test focussing on the notion of embodiment while interacting with the lectern. Therefore, the four formal-aesthetic parameters (1) basic shape, (2) handle form, (2) material and colour of the lectern and (4) haptic sensation while writing were investigated (Figures 3–7). The pre-test was conducted on 2 days during a lecture at Darmstadt University of Applied Sciences in January 2015. The sample therefore comprised students between the ages of 18 and 36. Since participants could voluntarily decide how many stations they wanted to participate in, station sample sizes varied between 62 and 76 per station (n₁ = 75, n₂ = 62, n₃ = 66, n₄ = 76), of which 24–29 were female and 31–46 were male participants. Participants evaluated all four parameters using a semantic differential based on 12 item-pairs on a 7-point-scale (e.g., “When writing on the writing pads A, B, C and D, I feel…” coded from “relaxed” to “nervous”).

Based on these pre-tests, an iterative design process continued that comprised several design concepts developed as scetches and CAD models before being built as miniature models and eventually manufactured as real size prototypes from (A) wood, hard leather and acrylite or (B) alumium profiles and acrylite (see Figure 8). The “positive” design prime resembles a robust and massive wooden lectern providing excessive space for writing on top of it. Its material properties, comfort and overall aesthetic appeal are chosen to represent valorisation for the task and the participant. The “negative” design prime is based on a rather functionalist, clean and technical setup made from alumium profiles and a red top part. Its writing surface is almost too small to write comfortably on it and slightly gives way when pressure is applied, which is supposed to emphasise qualities of weakness and uncertainty. The colour red was chosen to additionally exacerbate cognitive task solving (Mehta and Zhu, 2009). Adding to the clinical and sterile aesthetics of the overall design, the legs of the lecture build a squarelike shape embracing the participant and by this aiming to limit their subjectively experienced freedom of movement while using the lectern. All in all, in contrast to the “positive” design prime, most of the formal-aesthetic features of this design communicate less valorisation and hinder users’ performance potentially compromising their situational sense of competence and self-efficacy.

To confirm whether participants actually perceived the design prime the way they were intended to, participants were asked at the end of the actual randomised control trial (see 3.2) to rate the lectern they had used on a 6-item semantic differential. As expected, the positive design scored significantly higher on all items confirming its overall design “style” as being more encouraging, usable and attractive (see Figure 9).

To test the hypothesised effects of the designed primes on mental concepts of the people using them, an experiment was administered. The experiment took place independently in two different locations, (a) a University in Darmstadt, Germany, and (b) a Clinic for orthopaedic rehabilitation in Bad Kreuznach, Germany. In the University setting, n = 83 students (35 male, 48 female; median age between 22 and 25) took part in the study. In the clinical setting, n = 38 patients of the orthopaedic rehabilitation clinic took part (12 male, 25 female, 1 unknown); median age between 51 and 60 years). In both settings, participants were allocated randomly to one of two groups that were presented with the two lecterns described above. Group 1 (university: n = 41; clinic: n = 19) was presented with the stable, aesthetically favourable lectern (“positive condition,” see Figure 8, left). Group 2 (university: n = 42; clinic: n = 19) was presented with the unstable, aesthetically less favourable lectern (“negative condition,” see Figure 8, right).

To test the effect of both primes, participants were asked to evaluate an unrelated medical product (a pulse oximeter). Participants were recruited via advertisements, e.g., on posters, a week prior to the experiment. On the day of the experiment, participants approached the research assistant, received an envelope and were randomly assigned to one of two rooms (positive vs. negative condition). The actual location of the primes was alternated after each day, decreasing the likelihood of other environmental factors in the room to potentially affect the outcomes. The research assistant assigned participants alternating between both rooms or depending on which room was free. For each room, a set of papers was prepared in an enveloped label with a code relating to one of the two primes that was only known to the examiner. The assistant was unaware of which prime was placed in which room and of the actual goal of the experiment (double blind condition). In the room, the pulse oximeter was placed on the lectern while the lectern was the only useful piece of furniture to write on. After filling out the questionnaire, participants were asked to put the papers back into the envelope and go to the examiner in a different room to collect their incentives. Once participants had handed in the envelope, the actual background of the study was revealed to them behind closed doors (see Figure 10).

To evaluate the effect of the design primes a six pages questionnaire was developed comprising 46 items and an open-ended mentioning task. The first six items referred to the product participants thought to evaluate (e.g., “The product is easy to use.”). Furthermore, the questionnaire included a 10-item scale on general self-efficacy (WIRKALL_r; Schwarzer and Jerusalem, 1999), an 11-item scale on resiliency (RS11, Schumacher et al., 2005), a 9-item scale on sense of coherence (SOC-L9; Schumacher et al., 2000) and a 4-item short scale on sense of control (KÜ4, Kovaleva et al., 2012). Participants rated their agreement on a 4-point Likert-type scale ranging from 1 = “not true” to 5 = “totally true” (WIRKALL_r; α = 0.89), on a 7-point scale with varying codings (e.g., “very often” to “hardly or never”) (SOC-L9; α = 0.85), on a 7-point scale ranging from 1 = “do not agree” to 7 = “fully agree” (RS11; α = 0.88) and on a 4-point Likert-type scale ranging from 1 = “does not apply” to 5 = “fully applies” (KÜ4; α = 0.74). For this article, the most relevant element was a mentioning task at the end of the questionnaire asking participants to list all health-promoting activities for daily life that come to their mind. In addition to that, they were asked to indicate which of these activities they (a.) are already doing, (b.) assume they could do or (c.) do not believe to be able to do. Apart from these items, sociodemographic items were included (i.e., age, gender).

The items mentioned in the mentioning task were clustered into six categories derived from the data (diet & nutrition, sport, social activity, relaxation, joy, self-control & autonomy). The absolute mentioning frequency comprises all aspects independently from whether participants commented that they could implement them or not. In the clinical setting, from each group, several participants did not provide data on this task (group 1: n = 6; group 2: n = 8) and were thus excluded from the analysis. Similarly, in the university setting, n = 5 (group 3: n = 2; group 4: n = 3) participants did not provide data on this task and were excluded. Remaining sample sizes are n = 13 (group 1), n = 11 (group 2), n = 39 (group 3) and n = 39 (group 4). As hypothesised earlier, mentioning frequency of words that are related to physical activity pose the central variable of interest.

Group differences were assessed for two variables regarding mentioning frequency: absolute mentioning frequency and relative mentioning frequency. To account for verbal fluency, absolute mentioning frequencies of aspects related to physical activity were tallied and standardised by the overall sum of health-promoting activities mentioned. Resulting scores range from 0 to 1, with higher scores indicating stronger focus on aspects related to physical activity in all mentions. Before examining group differences in mentioning frequency of physical activity aspects, for the clinical setting, a Kolmogorov–Smirnov test indicated normal distribution of the relative mentioning frequency variable (p > 0.05) but not of the absolute mentioning frequency variable (p < 0.001). For the university setting, a Kolmogorov–Smirnov test indicated no normal distribution of the relative mentioning frequency variable and of the absolute mentioning frequency variable (p < 0.001). Hence, group differences in variables without normal distribution were assessed with a Mann–Whitney-U test as a non-parametric alternative to the t-test.