This field-study focuses on physical workload in relation to cardiopulmonary capacity, specifically among bulk waste collectors. Heart rate measurements and their correlation with individual oxygen consumption \((\mathrmO}_2)\) were used to assess the strain. The average oxygen uptake during the shift \((\mathrm O}_})\) was found to be 1267 ml/min, confirming the heavy physical workload of this occupation [16].

Our results show that most of the recommendations for RHR and RAS from the literature (Additional file 2: Acceptable limits of cardiovascular strain) for an acceptable workload were exceeded by the subjects. The limits based on \(\mathrmO}_2\) at the ventilatory threshold 1 (VT1) of the individual subject were not exceeded (Tables 3, 4, 5, and Additional file 2: Acceptable limits of cardiovascular strain). This disparity could be explained by the athletic endurance capacity of the study participants with a high oxygen consumption at VT1 (\(\mathrm O}_\%\mathrm O}_}\) of 61.2% (SD 14.1)) (Table 2) [32]. These findings support VT1 being a better choice or the upper limit for heavy occupational work compared to relative heart rate (RHR) or relative aerobic strain (RAS), because \(\mathrm O}_1}\) takes the individual endurance capacity into account. RHR or RAS consider only the maximum work capacity (%HRR, \(\%\mathrm O}_\)). For these parameters, it is assumed that the endurance capacity is 33% RHR [8], 40% RAS [10] or 40–45% RAS [38] as fixed percentages of the maximum work capacity, which should not be exceeded. The increased endurance capacity of subjects in hard-working occupations is not considered here. Therefore, the authors recommend using the HR or \(\mathrmO}_2\) at the ventilatory threshold 1 (VT1) of the individual, which is the upper physiological limit to sustain prolonged physical work.

In the field measurements, the subjects were able to perform a total of 3.17 h of heavy occupational work, distributed over intervals (Tint,C) of 16.4 min (SD 4) in the shift. During these intervals of collection, the oxygen consumption \(\mathrm O}_\) showed no differences from \(\mathrm O}_1}\) (p = 0.152) (Tables 4 and 5). This demonstrates, that a workload beyond VT1 for an extended duration is not feasible [24, 32].

Rapid slowing down of the HR was observed while driving (Tint,driving) to the next customer, which took on average 14.8 min (SD 4.1) (Additional file 4: Heartrate measurements in the field). Therefore, these driving intervals can be counted as recovery phases. Resting for at least 5 min is described as sufficient recovery for the cardiopulmonary system [39,40,41]. These interruptions were sufficient to reduce the total strain during the shift below VT1. Ilmarinen's recommendation (1992) that intensive work over an 8-h workday is acceptable if breaks are available can be confirmed with these observations [14]. This is further confirmed by the findings of Moser et al. (2015), who indicated that there are no discernible differences in physiological reactions between continuous and high-intensity interval workloads [42].

Compared to previous studies conducted in waste management and occupations with high physical exertion, we found no significant differences in average heart rate during 8-h shifts (HRshift) in our collective (Additional file 3: Other studies vs. our results) [3,4,5,6, 13, 20]. We found lower cardiopulmonary strain compared to other studies in Brazil, the Netherlands, Iran, and Japan regarding HR, RHR, and RAS during the actual waste collection activity [4,5,6, 13]. The different work organisation, with smaller individual weights and other shift lengths, could be causal for these observed differences.

However, other studies measuring oxygen consumption \((\mathrmO}_2)\) showed significantly lower values during uninterrupted work compared to \(\mathrm O}_\) in our study, i.e., oxygen consumption only during the time with physical stress (TC) [3, 20]. These results suggest that bulk waste collection involves either higher peak workloads or higher total workloads compared to the aforementioned studies.

Comparing the oxygen consumption \((\mathrmO}_2)\) of garbage collectors [3] with our collective bulk waste collection shows a higher \(\mathrmO}_2\) during the work intervals. This refers to the total \(\mathrmO}_2\) compared to \(\mathrmO}_2\) and also to \(\mathrmO}_2\) relative to \(\mathrm O}_1}\) (p =  < 0.0001, p = 0.0001, resp.; Additional file 3: Other studies vs. our results). The work of bulk waste collectors seems to be more stressful, at least during the intervals of bulk waste collection.

Another subgroup of waste collectors included in an investigation of a variety of occupations [20] showed a similar result with lower values for total \(\mathrmO}_2\) and lower \(\mathrmO}_2\) in relation to VT1 during continuous work (-599 ml/min, -27%, resp.; p =  < 0.0001 and p < 0.001, resp.). The authors, however, performed this measurement of \(\mathrmO}_2\) only over 20 min to obtain a comparative value. To assess the physical strain for the whole shift (8 h), they used the average HR relative to the HR at lactate turning point 1 (HR8h%HRLTP1). LTP1 is physiologically equivalent to VT1; consistently, the corresponding parameter to HR8h%HRLTP1 is HRshift relative to HR at VT1 (HRshift%HRVT1). We found no significant difference between these parameters based on LTP1 and VT1 (-5.8%, p = 0.0828; Additional file 3: Other studies vs. our results). In summary, bulk waste collectors show higher peak loads but similar overall strain compared to other occupations with high physical workloads.

Despite being exposed to high cardiovascular strain on a daily basis, the subjects occasionally reached heart rates close to their maximum heart rate (HRmax) during the field measurements and they showed average performance in terms of maximum oxygen consumption \((\mathrm O}_)\) and maximum power output (Pmax [W]) during CPX (Table 2). These results confirm the previously mentioned findings that physically demanding work has no training effect on maximum aerobic capacity [19, 32, 43, 44]. However, our results indicate that it might have an effect on endurance performance (\(\mathrm O}_1}\%\mathrm O}_}\) of 61.2% (SD 14.1)) [32].

The difference of 13.5% between \(\mathrm O}_\%\mathrm O}_}\) (90.5 SD 13) and Pmax%Pmax,pred (104% SD 14.4) was expected because the predicted values for Pmax are older and lower than the predicted values for \(\mathrm O}_\). Apart from this, the relationship between power output and \(\mathrmO}_2\) is linear [32].

The cumulative volume of compacted bulky waste examined in this study was 3.9 metric tons per shift, resembling the company's historical data from 2005 to 2018, with an average of 3.7 tons. Thus, the selected tours can be considered representative.

Field-studies with individuals who move unpredictably (from a research perspective) and perform heavy physical labour are complex; they require a high level of commitment from the individuals involved, as measurement devices may interfere with their habitual movements and pace of work, which pose a safety risk. Therefore, direct \(\mathrmO}_2\) measurement with the mobile CPX device was not possible. The assessment of \(\mathrmO}_2\) via heart rate is a good alternative but may over- or underestimate the actual oxygen consumption during work because HR is influenced by several factors [45]. The most important factor possibly resulting in an elevated HR could be temperature, especially heat. The study was conducted with an average dry bulb temperature of 19.4°Celsius (SD 5.86; 12–30.9°Celsius) and a mean relative humidity of 47.5% (SD 16.6; 25–82. For heavy occupational work, the optimal temperature and relative humidity are considered 17 °C (15 °C-21°C) and 50% (30%-70%) [46]. Considering these recommendations, temperature and relative humidity were within the ranges of optimal climate conditions, where a low influence on HR can be expected. However, during peak temperatures, such as the highest temperature of 30.9°Celsius in our study, it was shown that \(\mathrmO}_2\) may be overestimated up to 64% [40].

Medications that influence the HR were an exclusion criterion. No cardiovascular disease, metabolic disease, or psychiatric disease that could influence the HR were found in PME. Of the study population, 35.7% were overweight (BMI ≥ 25) and 42.9% were obese (BMI ≥ 30), which could result in a higher HR. In this study, five subjects smoked, which may have raised HR. Mental tension and noise due to the working conditions could not be ruled out and may have influenced HR. The influence of circadian rhythm is not relevant because we investigated all subjects at the same time of day. Even considering the possible overestimation of \(\mathrmO}_2\), the feasibility of using HR measurement outweighs the health and safety risks of using devices for direct \(\mathrmO}_2\) measurement.

The 14 subjects finally included in the analysis were on average younger than the baseline population (mean 42.6 years (25.6–57.1) (n = 14) vs. 47.3 (23–63) years (n = 105)) (Additional file 5: Characteristics of baseline population). This could lead to an overestimation of the mean \(\mathrm O}_\), as it tends to decrease with advancing age [38].

The study was conducted only once with each subject; therefore, intraindividual variations in performance may go unnoticed. Since the subjects were accompanied the whole time during the field measurements, it is possible that they may have changed their natural behaviour. Participating in a study and being under observation are possibly visible in the outliers of \(\mathrm O}_\%\mathrm O}_1}\) with the lowest working at 62.8% and the highest at 146%.