• No results found

Theoretical background: absolute versus relative (specific) oxygen consumption It is important to first comment upon the indices used within this report to quantify the

The 15-point Borg scale 6

3. RESULTS AND DISCUSSION

3.1 Simulation one: Hazmat incident (in pairs) .1 Example experimental data

3.1.3 Theoretical background: absolute versus relative (specific) oxygen consumption It is important to first comment upon the indices used within this report to quantify the

13 This is the absolute rate at which oxygen must be consumed (oxygen flow: L.min-1) to liberate the energy required to successfully complete this work. Some individuals may be unable to satisfy this demand.

14 This generalisation assumes that the metabolic efficiency of cycling remains constant across individuals.

addition to that which is directly related to the work being performed, and the hazmat simulation provides an ideal illustration of this fact. For instance, the firefighters were wearing encapsulating, chemical and biological protective clothing (Figure 2). Due to using open-circuit respiratory gas analysis methods to collect oxygen consumption data, the head and upper torso were not wholly encapsulated. Nevertheless, this encapsulating clothing imparts two added stresses upon the wearer. Firstly, such ensembles increase the metabolic demand of locomotion due to elevated frictional forces that exist within layered clothing, and the restrictive nature of such garments across limb joints (Teitlebaum and Goldman, 1972; Nunneley, 1989; Dorman and Havenith, 2009). Secondly, multi-layered garments, and in particular encapsulating ensembles, trap heat (Nunneley, 1989; McLellan, 2008). As a consequence, skin blood flow increases to facilitate the dissipation of metabolic heat, thereby driving heart rate upwards, and out of proportion to the increase in metabolic rate.

This is still a stress that firefighters must face, and it cannot be ignored. Thus, whole-body physiological strain must be evaluated from both heart rate and oxygen consumption data, and this is why these variables were measured simultaneously within this project.

3.1.3 Theoretical background: absolute versus relative (specific) oxygen consumption

15 Normalising involves dividing the index of interest (e.g. oxygen consumption) by some variable that is tightly correlated with that index (e.g. surface area or mass). Thus, the absolute oxygen consumption (L.min-1) is converted to a relative (specific) oxygen consumption (mL.m-2.min-1 or mL.kg-1.min-1). In these examples, it is assumed that the relationship between the index of interest and the chosen divisor is always linear.

16 The word “specific” designates any quantity normalised to (divided by) body mass (Royal Society, 1975).

have a larger absolute oxygen consumption. In this circumstance, normalising15 these data for variations in mass will help to remove the influence of inter-individual variations in body mass, permitting one to compare the oxygen cost of locomotion across people with different body sizes. These normalised data are presented in two different forms (Table 15).

In the first instance, data are presented in a form that is most familiar to readers; the simple division of oxygen consumption by body mass (mL.kg-1.min-1). This linear format is also consistent with the current practise within many organisations that use fitness standards for recruiting purposes (e.g. maximal aerobic power of 45 mL.kg-1.min-1 [Gledhill and Jamnik, 1992a]). It is also consistent with the popular procedure by which variations in physical endurance or fitness are compared among individuals (maximal aerobic power). That is, normalisation is performed using a linear (arithmetic) assumption, such that across the entire range of body sizes, this simple ratio is assumed to permit a valid comparison of the relative impact of a task on different individuals, with the influence of body mass on oxygen

consumption being removed (body mass-independent). Data presented in this manner are described as the relative or specific16 oxygen cost of a task (Royal Society, 1975).

Notwithstanding its popular use, the difference between the absolute and specific oxygen consumption derived in this manner is often misunderstood. Indeed, this normalisation is frequently inappropriate (Appendix Two). This is so for several reasons: (1) a one-to-one relationship between oxygen consumption and body mass does not exist (Kleiber, 1932;

Tanner, 1949; Taylor et al., 1981; Schmidt-Nielsen, 1984; Åstrand and Rodahl, 1986;

Nevill et al., 1992); (2) linear normalisation fails to account for all of the inter-individual variability in oxygen consumption (Kleiber, 1947); (3) the coefficient of variation for oxygen consumption often exceeds that for body mass (Tanner, 1949); (4) for maximal aerobic power, there is a positive relationship between the peak absolute oxygen

consumption and body mass, but a negative relationship is evident between peak specific oxygen consumption (mL.kg-1.min-1) and body mass (Taylor et al., 1981; Schmidt-Nielsen, 1984; Åstrand and Rodahl, 1986; Nevill et al., 1992; Bilzon et al., 2001a); and (5) the affect of these artefacts increase as individuals approach the extremes of body size, so the extrapolation of such regression relationships beyond the range of the primary observations is fallacious (Tanner, 1949; Schmidt-Nielsen, 1984). Therefore, the injudicious division of body mass into oxygen consumption may be invalid in many circumstances.

Accordingly, a second form of data normalisation has been adopted for this report, and this relies upon the well-established power relationship between oxygen consumption and body mass that obtains across metabolic states from rest (Kleiber, 1932; Schmidt-Nielsen, 1984) through to maximal exercise (Taylor et al. 1981; Åstrand and Rodahl, 1986; Nevill et al., 1992). Thus, for one to compare the mass-independent oxygen consumption of individuals of different sizes, one must derive specific oxygen consumption as a power, and not as a linear function. Therefore, whilst it is well established that body size is important, it is

17 Simplification: mL.kg-1.h-1 / km.h-1 = mL.kg-1.km-1

absolutely critical to apply the correct scaling function (power), rather than the more convenient (linear) function. Accordingly, this convention has been adopted for this project (mL.kg-0.67.min-1), and these specific (relative) oxygen consumption data are presented in Table 15, and in the corresponding simulation Tables that follow.

When data are presented in this power-scaled format, one can describe the specific oxygen cost of any simulation (e.g. Figure 21 and Table 15) with a knowledge that the inter-individual variations in the body masses of the firefighters studied (e.g. Table 6) were not responsible for determining the outcome. This is an absolutely essential attribute for any occupational task assessment, if the corresponding physiological attributes of potential recruits are to be gender neutral and legally defensible. With respect to the current simulation, which was a loaded ambulatory task, firefighter mass varied by almost 50 kg from the heaviest (113.6 kg) to the lightest (65 kg) person tested. Thus, one would predict the absolute oxygen consumption of the larger firefighter to be considerably higher, since that individual was carrying approximately 50 kg more mass, even before donning the personal protective clothing and equipment, and before the loads associated with the activity were considered. This expectation was realised, with the mean absolute oxygen

consumption of the heaviest individual being 0.39 L.min-1 during seated rest, and 1.93 L.min-1 when averaged across the entire simulation. The corresponding data for the lightest firefighter were 0.20 L.min-1 and 1.17 L.min-1. It was noted above that, for some equipment manipulation tasks, the absolute oxygen consumption is an important consideration. Before returning to this, we will further develop our discussion concerning locomotion.

The metabolic cost (per unit mass) of ambulatory tasks performed on flat surfaces will increase as a linear function of movement speed in both animals (Taylor et al., 1970) and humans (Mayhew, 1977). The applicable units may be expressed as mL.kg-1.h-1 for the (dependent) metabolic cost, and as km.h-1 for speed. Thus, the oxygen cost derivative17 will have the units of mL.kg-1.km-1 (an analogue of specific fuel economy), and this is

independent of speed, but it is dependent upon body mass and the distance covered within the task (Schmidt-Nielsen, 1984). If the distance remains fixed, as it does in these fire-fighting simulations, then the metabolic cost of a level-surface ambulatory activity simplifies to a mass-dependent, power relationship. It has been established that the exponent of this relationship is a negative function of body mass across both quadrupedal mammalian species (mouse to the horse: Taylor et al., 1970) and bipedal species (birds and humans: Fedak et al., 1974). For bipeds, the exponent is -0.33 (Schmidt-Nielsen, 1984). Within the context of the current fire-fighting simulations, this implies that the oxygen cost of load carriage will be greater within smaller individuals. This is a well-known fact (Louhevaara et al., 1986;

Bilzon et al., 2001a).

Indeed, Taylor et al. (1980) demonstrated that when carried loads were normalised to body mass, the absolute oxygen cost changed in direct proportion with the change in the specific load, such that a 5% increase in relative load was accompanied by a 5% elevation in oxygen consumption. However, if the load carried is constant (e.g. a 35-kg ventilation fan; two-person carry: 17.5 kg), and it is carried by individuals having different body masses (e.g.

18 This helps to explain why small animals can ascend tress with ease: there is only a small difference between the oxygen cost of horizontal and vertical locomotion in these animals (Schmidt-Nielsen, 1984).

our heaviest [113.6 kg] and lightest firefighters [65 kg]), then this load will represent a greater metabolic demand for the lighter individual (27% versus 15%), relative to performing the same task in the unloaded state. This is precisely the scenario that

firefighters face, and it enables larger individuals to work with less strain and for longer durations without fatigue (Louhevaara et al., 1986; Bilzon et al., 2001a). Thus, when load carriage is of importance within an occupation, one must evaluate physiological function under loaded situations (Vanderburgh and Flanagan, 2000; Bilzon et al., 2001a;

Vanderburgh, 2008; Vanderburgh et al., 2011), as has occurred within the current project.

The discussion to this point relates only to activities performed on level ground. During activities that have a significant vertical component, it appears that the possession of a lighter body may be advantageous. The specific oxygen costs of moving 1 kg a vertical distance of 1 m is reasonably constant across species (1.36 mL.kg-1.m-1), even in those of widely different body masses (Taylor et al., 1972; Cohen et al., 1978). Since smaller animals have a much larger specific oxygen consumption at rest than bigger animals, then this constant demand represents a much smaller change for these animals18. Of course, the comparisons between mice and horses (Taylor et al., 1972) are extreme examples that prove the principle, which does transfer to humans. However, and of much greater importance, is the fact that larger individuals must first move a greater absolute body mass through the same vertical distance (Teh and Aziz, 2002). Therefore, smaller people will have an advantage on the stairs. This advantage may be increased marginally since the smaller individual (e.g. 55.3 kg [Table 7] versus 113.6 kg [Table 6]), like smaller animals, will have a greater specific oxygen consumption at rest. Thus, the constant specific oxygen cost of the ascent represents a slightly smaller change, relative to the resting state.

In some situations, it may be the absolute oxygen cost of an occupational task that is important. In cycling, where the body mass is fully supported, this is absolutely so.

However, within some occupations, the performance of some tasks will elicit a relatively fixed oxygen cost on the worker, and, regardless of the size of an individual, this demand must be met to successfully complete the task. In these situations, it becomes absolutely necessary to characterise occupational tasks using individuals of widely varying body masses. This criterion was comprehensively satisfied within the current observations, since, across the sixteen simulations investigated, firefighter mass ranged from 55.3 kg (Table 7) to 113.6 kg (Table 6).

Let us now consider an activity that would elicit, on average, an absolute oxygen

consumption of 3.0 L.min-1 above rest, across a wide range of individuals. For simplicity, let us think only about this oxygen cost as if it were independent of body mass. Now let us evaluate the capacity of a group of 21 different (hypothetical) individuals to perform this activity. We will consider people with body masses ranging from 40-100 kg (seven 10-kg categories), and, within each mass category, we will have three individuals, each of whom will possess one of the following levels of specific peak aerobic power: 40, 50 and 60

19 For this illustration, the validity of this linear normalisation is ignored in favour of simplicity.

20 The 90% threshold is deliberately liberal, chosen only for this illustration, and it creates a 10% margin for safety. However, in the workplace, this safety margin might need to be closer to 25%.

21 Physiological strain = (resting + task oxygen demand [L.min-1]) / (peak oxygen consumption [L.min-1])%

Successful task performance is possible when strain (metabolic scope) is <90%.

mL.kg-1.min-1 (considered as: “average”, “good”, “excellent”)19. The hypothetical question of interest is as follows: which of these individuals could complete this occupational task with an oxygen cost less than 90%20 of their peak aerobic power? Asked another way, who would be capable of consuming an additional 3.0 L.min-1 of oxygen above rest, whilst still working at an intensity below 90% of maximal? Through a series of elementary

calculations, one could determine that only seven people might be able to complete this task21. Indeed, it would appear to be impossible for all but one individual with a body mass below 80 kg. This demonstrates that lighter people, even those with an excellent specific peak aerobic power, may simply be unable to generate the required absolute oxygen consumption to perform some occupational activities. Accordingly, one may question the utility of setting some physiological employment standards on the basis of a specific peak aerobic power, particularly when a linear normalisation procedure has been applied.