In this study, we examined, using multiple sources of data, how different objective and subjective job demands were associated with day-time recovery among home care staff. The results indicate that lower number of different clients and their lower care needs (CMI) are associated with higher amount of day-time recovery. In addition, nurse’s workday going nearly as planned (as opposed to going as planned) was associated with higher amount of recovery. Other subjective job demands were not associated with day-time recovery. Based on the findings of our study, it further seemed that care work in home care is not particularly physically demanding, which was also determined in a study by Mänttäri and co-authors (Mänttäri et al. 2023).

Firstly, it is important to note that the overall level of day-time recovery had significant variance (0–380 min). According to the 24-h HRV measurements, some participants had no day-time recovery, while others had multiple hours. Different job descriptions, daily task lists, and quality of spare time might explain some of the variance. A workday with densely scheduled client visits, limited break time, followed by a busy evening could potentially lead to a valid measurement with 0 min of recovery. On the other hand, a registered nurse working mostly in the office, having relatively few client visits and small number of stressful events, followed by a quiet evening, could feasibly have over 6 h of day-time recovery. However, while recovery-focused physiological measurements using the Firstbeat Bodyguard2-devices have been determined as reliable and valid (Palmer et al. 2021; Parak & Korhonen 2013), it is possible that, for example, accuracy of the recovery analysis and inter-individual differences might have affected the results.

Our findings suggest that home care nurses with more direct care time with clients, which may indicate more intense work, might have less day-time recovery, since the results indicated a weak association. The results further suggested that clients’ care needs affect the recovery of nurses, as nurses with clients that required more resources had lower amount of day-time recovery. This might be partly explained by the fact that clients with higher care needs require additional care time and more complex care tasks, which might for instance hasten subsequent client visits, leaving only little or no time for a break. These results are in accordance with the findings from previous studies, which have highlighted the association between job demands and poor recovery (Gifkins et al. 2020; Sonnentag et al. 2022). Due to the lack of care workforce, more efforts should be placed on strategies to arrange nurses’ time for recovery. In addition, methods and skills that promote day-time recovery from work should be utilised and developed. While the population ages throughout Europe, the demand for nurses will further increase. Consequently, it is crucial to at minimum maintain the work ability of nurses working in older people care. Deteriorating working conditions can lead to high turnover of nurses, resulting in lower care availability and quality (Antwi & Bowblis 2018), which would be a major public health concern.

The finding regarding the association between clients’ higher care needs and nurses’ lower recovery is especially relevant, as to our knowledge no studies have been able to account for the case-mix in this setting. However, similar results have been found in other nursing populations, for instance regarding patient-related stressful situations in emergency care (De Wijn & Van Der Doef 2020). The care needs of the clients should be considered when planning daily visits and coordinating care, because it might be associated with client safety. Sufficient care resources should be ensured for clients with high care needs, requiring for instance assistance from more than one nurse. The objective demands may reflect the findings related to subjective demands as well. An association between time pressure and lower recovery was found, however, the association did not remain significant after including the objective demands into the model. This might be explained by the correlation between direct care time and time pressure, which are both indicators of workday intensity. Still, the subjective stress cannot be ignored, as the decisions for example relating to turnover, are made based on subjective experiences. A former study showed that objective demands (staffing level) was a strong instrument for job demand and strain (Elovainio et al. 2015). Although the lack of staff is a major cause of job demands, the perceived demands also seem to have their effect on strain, or in this case, on day-time recovery.

Regarding care continuity and day-time recovery, we found that nurses with higher number of different clients per week had lower amount of recovery. On one hand, this can simply relate to the intensity of the workday, as the number of different clients was correlated with the number of total clients through the day. On the other hand, the discovered lower recovery might relate to a lack of job control, where a nurse is unable to plan their day and might have to visit more unfamiliar clients. The results regarding job control are in line with previous research on nurses and other occupations, which indicates that increasing the level of job control might promote better recovery from work (Gifkins et al. 2020; Kinnunen & Feldt 2013). If the number of different clients per week is low, the clients are more familiar to the nurses, which means that the continuity of care is high. This enables understanding the specific needs and wants of each client, their treatments, and where to find different things at their homes. In terms of recovery, higher continuity of care might refer to more ‘routine-like’ work, which may enhance recovery during workday via familiarity of environments and clients. In addition, better continuity of care in home care has been associated with better patient outcomes, which can help save care resources and improve quality of care (D. Russell et al. 2011).

One unexpected finding was that the amount of breaktime did not contribute to higher recovery in our study. Previously, the positive association between within-day work breaks and recovery has been found both among nurses and among the general population (Sagherian et al. 2023; Sianoja et al. 2016). Nurses in our study had on average only 30 min of breaks during their workday, which could be considered relatively low. It is possible that in home care work securing sufficient and uninterrupted breaks during the day can be difficult, which might have affected the results. If the breaks were often fragmented or disturbed by for example phone calls or consultations, their effect on recovery can diminish. Securing adequate break time during workday is important, since studies have shown that fragmented or too short breaks are associated with burnout (K. Russell 2016), and break disturbances with leaving intentions (Wendsche et al. 2022). Lastly, as the level of work autonomy has been shown to moderate recovery occurring during lunch breaks (Trougakos et al. 2014), it is possible that breaktime not being associated with day-time recovery can be partly attributed to low job control of home care nurses.

Another surprising finding was that smaller disruptions during the workday can potentially increase the amount of measured day-time recovery. It is possible that some disruptions might be relatively small in their effect and may instead lead to additional break-time, for instance in form of waiting for clients or other nurses. The potential of within-working day recovery in preventing stress and maintaining performance is in the observation that recovery is not equal to rest (Kinnunen et al. 2011). Therefore, by including familiar routine-like work tasks to the workday, it is possible to influence employees’ day-time recovery. As demonstrated in a study from Bono and co-authors (Bono et al. 2013), positive events during workday were associated with lower stress. In our study, we were not able to distinguish whether the disruptions were positive or negative. The negative aspects of brief disruptions, such as a busier schedule for rest of the day, seemed to not have a significant effect on day-time recovery, yet they might contribute negatively to perceived job demands and work-related wellbeing.

Age and physical fitness (VO2max) were used as adjusting variables in the multivariate models. Individual factors, such as age, sex, physical fitness, body composition, and health status have been found to affect heart rate variability (Shaffer & Ginsberg 2017; Teisala et al. 2014). However, in this study, physical fitness did not have an effect on measured day-time recovery, whereas the results indicated a tendency for lower day-time recovery in older nurses. This might be due to the finding that work in home care is not especially physically straining, and therefore psychosocial risk factors instead might be more prevalent. In comparison, work in hospitals and assisted living facilities might involve more physical demands, such as moving patients (Poole Wilson et al. 2015). Mänttäri and co-authors (Mänttäri et al. 2023) found that physical workload in relation to maximal capacity was higher for older nurses in Finnish home care, which was attributed to age-related decline in physical fitness. Higher relative physical workload could also explain the lower day-time recovery in older nurses in the current study. The VO2max values used as an indicator of physical fitness were calculated based on self-reported background variables and are, therefore, only estimates, which might also explain the lack of association. Nevertheless, literature indicates that maintaining adequate physical fitness is important for reducing relative occupational workload and ensuring sufficient recovery from occupational stress (De Bloom et al. 2018; Sliter et al. 2014).

Our study had a few distinct advantages. First, the multi-source study design with objective and subjective job demands as explanatory variables, in addition to physiological outcome data, offered a novel way of exploring the effects of different job demands on recovery among home care staff. Second, heart rate and heart rate variability monitoring using ambulatory method and performed during authentic work time provides reliable and valid results on physical workload (Smolander et al. 2008, 2011), and recovery (Palmer et al. 2021; Parak & Korhonen 2013). The objective measurement further diminishes the biases related to self-estimation.

Next, the inclusion of the workday characteristics, obtained from a comprehensive time measurement survey and clients’ RAI data, offered insight into the direct and objective job demands home care nurses experience. Additionally, we were able to take into account the clients’ need for care, which is seldom done. This is not a minor aspect, as in this study the mean of clients’ case mix varied substantially, indicating that clients of some nurses may require significantly more care time than others.

Last, as opposed to the number of participants often found in physiological measurement studies, in our study an adequate number of participants was recruited. Regardless, the statistical power of the analyses was slightly limited, which caused some of the probable associations not to be statistically significant. In order to control for missingness in our data and analyses, we used pairwise deletion to make the most out of the dataset.

However, our study had some limitations. First, while the workday characteristics consisted of objective measurements of job demands, the data were collected using relatively complex self-reported surveys. According to Lopetegui and co-authors (Lopetegui et al. 2014), the use of self-reporting in time motion studies can result in overestimating care time and in discrepancies on how nurses fill the surveys. These factors were mitigated with video tutorials and training sessions organised for the participating care units. In addition, the nurses reported the start and end times for each task, which may diminish the risk of overestimating care time. This same time code in regard to care time is used also in the official nurse documentation systems.

Some factors exist which can affect the validity and the reliability of the results. First, there exists a possibility for selection bias, as the care units were selected on a volunteer basis. This means that units with no excess time for research activities, and possibly worse workforce or client situation, might not be included in the study. As such, the units included in the study might be in a better situation in terms of care workforce and overall job strain. Next, in terms of statistical validity, some of the independent variables (especially the survey questions), were positively skewed. Sensitivity analysis was conducted with differently categorized variables: The direction of the results remained, but the effect of something slightly disrupting the day often weakened to statistical non-significance.

It is unclear how the results will transfer to other care settings, as the work in home care is considerably more mobile and independent compared to for example work in nursing homes. In addition, countries with significantly differently organised care for older people might have other factors that affect care workers job demands and recovery from it. Last, due to the cross-sectional study design, causality cannot be inferred from the statistical analyses.