Effect of task factors on the heart rate recovery time in a lifting task

2.1Participants

Twenty-four male students from Louisiana State University (LSU) between the ages 20–37 participated in this study. Participants were selected on a volunteer basis. Due to the use of human subjects, a permission from LSU institutional review board (IRB) was obtained and each participant had to read and sign the consent form before performing the experiment. The full demographic data for participants is presented in Table 1.

2.2Apparatus

A wooden platform with adjustable shelves was used for the lifting task. The bottom shelf was adjusted to the average knuckle height (77.8 cm) and the top shelf was adjusted to the average shoulder height (147.6 cm) of test participants.

A wooden crate was used to hold various weights. The crate measures 45.5 cm × 30.5 cm × 20.5 cm and has two handles installed on each side of the crate, cushioned by racquet grip tape.

A variety of weights were used to fill the crate as part of the experiment for each participant to perform a lifting task with a certain weight (10 or 20 kg).

In order to alert a participant of the cycles (frequencies) of a lifting task, an iPhone application named Gymboss Interval Timer was used.

A Bluetooth heart rate monitor (Polar H7, USA) was worn by each participant throughout the experiment (attached to the chest). This device transmits data to a smartphone interface (Polar app) and the collected data (participant’s heart beats) can be transferred to Microsoft Excel.

2.3Experimental design

A 23 factorial design was used for this research. Aghazadeh [21] used a similar design in an MMH study. In his work, three independent variables of height, frequency, and container type, each at two levels, were used to predict the maximum acceptable weight of lift (MAWL). In the current study three independent variables of frequency (lifts/min), duration (min), and weight (kg) of the lift, each at two levels, were used to determine the HRR. Each independent variable had a high level and a low level (frequencies of 6 and 9 lifts/minute, durations of 10 and 20 minutes, and weights of 10 and 20 kg). Giving the possible treatment combinations, eight (23) treatments were obtained in the experiment (Table 2). Each participant was randomly assigned to a unique treatment (between subject design). The randomization was done by an online tool named “Random.org”. The experiment was conducted over three days, and on each day a full set of eight treatments was used (by 8 participants). Therefore, the design structure of this experiment was a randomized complete block design (RCBD). Blocking was accounted for by using the variable day as a random effect in the model. The plan was to replicate each treatment with three observations, giving twenty-four observations.

The main response variable of this study was the HRR which was the duration following a lifting task that the heart needs to reach a steady state.

The secondary response variable was the perceived level of exertion. A common method in rating the perceived difficulty of a manual task is the use of Borg scale [22]. The original Borg Scale ranges from 6 to 20 points. A modified version of Borg Scale called Borg CR10 was used in this study, which employs a response format that ranges from a value of 0 (no exertion at all) to 10 (maximal exertion). After each lifting task participants were asked to rate the difficulty of the task from 0 to 10. A copy of the Borg CR10 is presented in Appendix A.

2.4The lifting task

The task consisted of repetitive freestyle lifting of a box from knuckle to the shoulder height, while no lowering was required. A helper lowered the box from the other side of the lifting apparatus throughout the experiment. Lifting was chosen due to the fact that is pervasively used in material handling despite advancements in work mechanization [23]. Each person performed the lifting for a predetermined amount of weight, with a specific frequency, and over a certain amount of time (for example, “lift 10 kg for 10 minutes at 6 lifts per minute”). The participant began the lifting with the activation of the interval timer. Two responses were captured as a result of the lifting task: HRR and Borg scale. Upon completion of a lifting task, the participant was asked to sit and relax with no distraction (cell phone use, eating, taking, etc.) for 10 minutes. The heart rate activity of each participant was monitored before, during, and after a lifting task by the heart rate monitor.

2.5Heart rate data processing

Pulat [24] explains in a sub-maximal physical activity, the heart rate is at a stable range before the activity and will reach to a steady high level while the activity is in progress. Upon finishing the activity, the heart rate will reach to some steady state (Fig. 1). However, based on the nature of the task and individual characteristics, the heart rate might not return to the same steady state as in the pre-activity phase and remains at a range above the resting level for some time [25]. Discovering the duration reaching a steady state after the completion of a lifting task was of the interest of this experiment. The steady state was defined as a phase on the heartbeat graph that the heartbeats form a semi-horizontal line while fluctuating withina 10% range. The minimum length (time) for a phase to be considered “steady”, was decided to be 2 minutes. The heart rate monitor used in this study captured a reading (data point) for every second. In order to find the HRR, the graph of heart rate against time was drawn in Excel and a smoothing algorithm with moving average of 30 seconds was applied to the graph. Then, the recovery duration was determined using the Excel gridlines.

Figure 1.

Heart rate in a sub-maximal physical activity (adopted from Pulat, 1997).

3.1HRR

First, the average recovery time among the two levels of each factor along with their percentage change was calculated (Table 4). Based on the results of this table, by doubling the weight of the lift from 10 kg to 20 kg, the average HRR increased by 37.4%, and by increasing the frequency from 6 lifts/minute to 9 lifts/minute the average HRR increased by 34.8%. On the other hand, we observe a 2.7 decrease in the HRR when lift duration increased from 5 minutes to 10 minutes.

A mixed model ANOVA for HRR values was conducted (Table 5). At 10% significance level, weight (F-value = 8.11, P-value = 0.0129), frequency (F-value = 7.15, P-value = 0.0181), and the interaction between frequency and duration (F-value = 4.50, P-value = 0.0523) had a significant impact on the HRR.

To further analyze the main effect of each lifting factor, the grand average of the HRR with respect to each individual lifting parameter (at two levels) was calculated and three linear graphs were drawn (Figs 2–4). The best-fit equations for each graph were obtained as follows:

Figure 2.

Effect of weight on HRR.

The best-fit equation of Fig. 2 for the average HRR based on two levels of weight is as follows:

HRR for weight response:

The best-fit equation of Fig. 3 for the average HRR based on two levels of frequency is as follows:

Figure 3.

Effect of frequency on HRR.

HRR for frequency response:

Figure 4.

Effect of duration on HRR.

The best-fit equation of Fig. 4 for the average HRR based on two levels of duration is as follows:

HRR for duration response:

3.2Borg-rating

The average Borg-rating among different levels of each factor along with their percentage change was calculated (Table 6). By increasing the weight from 10 kg to 20 kg, a 20.1% increase in the average perceived level of exertion is observed. On the other hand, increasing the weight and duration of the lift did not result in a significant increase in the average Borg-rating among test participants.

A mixed model ANOVA for Borg-rating values was conducted (Table 7). Based on this result, at 10% significance level, just weight of the lift had a significant effect on the Borg-rating. (F-value = 3.33, P-value = 0. 0894).