Sports Medicine

, Volume 36, Issue 11, pp 911–928

Perceived Exertion

Influence of Age and Cognitive Development


    • Laboratory of Sport SciencesFEMTO UFR STAPS de Besançon
  • Anthony D. Mahon
    • Human Performance Laboratory, School of Physical Education, Sport, and Exercise ScienceBall State University
Leading Article

DOI: 10.2165/00007256-200636110-00001

Cite this article as:
Groslambert, A. & Mahon, A.D. Sports Med (2006) 36: 911. doi:10.2165/00007256-200636110-00001


Because little is known about the effects of aging on perceived exertion, the aim of this article is to review the key findings from the published literature concerning rating of perceived exertion (RPE) in relation to the developmental level of a subject. The use of RPE in the exercise setting has included both an estimation paradigm, which is the quantification of the effort sense at a given level of exercise, and a production paradigm, which involves producing a given physiological effort based on an RPE value.

The results of the review show that the cognitive developmental level of children aged 0–3 years does not allow them to rate their perceived exertion during a handgrip task. From 4 to 7 years of age, there is a critical period where children are able to progressively rate at first their peripheral sensory cues during handgrip tests, and then their cardiorespiratory cues during outdoor running in an accurate manner. Between 8 and 12 years of age, children are able to estimate and produce 2–4 cycling intensities guided by their effort sense and distinguish sensory cues from different parts of their body. However, most of the studies report that the exercise mode and the rating scale used could influence their perceptual responsiveness.

During adolescence, it seems that the RPE-heart rate (HR) relationship is less pronounced than in adults. Similar to observations made in younger children, RPE values are influenced by the exercise mode, test protocol and rating scale. Limited research has examined the ability of adolescents to produce a given exercise intensity based on perceived exertion. Little else is known about RPE in this age group.

In healthy middle-aged and elderly individuals, age-related differences in perceptual responsiveness may not be present as long as variations in cardiorespiratory fitness are taken into account. For this reason, RPE could be associated with HR as a useful tool for monitoring and prescribing exercise. In physically deconditioned elderly persons, a rehabilitation training programme may increase the subject’s ability to detect muscular sensations and the ability to utilise these sensory cues in the perception of effort.

RPE appears to be a cognitive function that involves a long and progressive developmental process from 4 years of age to adulthood. In healthy middle-aged and elderly individuals, RPE is not impaired by aging and can be associated with HR as a useful tool to control exercise intensity. While much is known about RPE responses in 8- to 12-year-old children, more research is needed to fully understand the influence of cognitive development on perceived exertion in children, adolescents and elderly individuals.

The perception of exertion can be considered as a configuration of sensations: strain, aches and fatigue involving the muscles and the cardiovascular and pulmonary systems during exercise. These sensations are generally classified as being derived from either cardiopulmonary or peripheral factors. Cardiopulmonary factors include variables such as heart rate (HR), oxygen uptake (V-dotO2), respiration rate and minute ventilation, while peripheral/metabolic factors include blood lactate concentration, blood pH, mechanical strain, skin and core temperature.[1] In Borg’s model,[2] it is observed that as exercise performance increases along an intensity-dependent continuum, there are corresponding and interdependent increases in response intensities along perceptual (i.e. perceived exertion) and physiological (e.g. HR, respiratory rate, V-dotO2) continua, demonstrating a positive relationship.

From this relationship, different rating scales have been validated in adults, such as Borg 6–20 rating of perceived exertion (RPE) scale,[3] which was constructed to provide perceptual data that are linear with HR and power output. Borg[2] also developed a category-ratio scale (CR-10) that is appropriate for assessing sensations that may arise from physiological variables that grow exponentially, such as blood lactate or pulmonary ventilation. However, these rating scales have been validated only in adults not in children. For this reason, there are linear rating scales for children created on the basis of common expressions and familiarity with a limited number range (e.g. 1–10) such as the Children’s Effort Rating Table (CERT),[4] and pictures and expressions such as the OMNI,[57] Cart and Load Effort Rating (CALER),[8] Pictorial CERT (PCERT),[9] Bug and Bag Effort (BABE)[10] and Rating of Perceived Exertion in Children (RPE-C) Scales.[11] A pictorial curvilinear scale also has been proposed recently by Eston and Parfitt.[12] All of these rating scales have been used with varying degrees of success as a means of assessing exercise effort.

At the present time, it is still not clear how the brain interprets afferent feedback to induce perceived exertion. It has been suggested that an integration of these different cues may indirectly and unconsciously influence perceived exertion during exercise.[13] Ulmer[14] and more recently Hampson et al.[15] have suggested that exercise performance may be controlled by central calculations and efferent commands that attempt to couple the metabolic and biomechanical limits of the body to the demands of the exercise task in a process described as teleoanticipation. However, little is known about the components of the cognitive functions involved in perceived exertion. It is likely that these functions are developmental, expanding in scope and maturating in both precision and efficiency as the individual’s movement-related experiences become more varied. One interesting question is to know how these functions evolve in relation to aging and particularly between birth and adult age and after 50 years of age.

The cognitive functions involved with perceived exertion are usually investigated by using two different paradigms: estimation (passive) and production (active). According to Eston and Parfitt,[12] these paradigms place different demands upon a three-effort continua (perceptual/psychological, physiological and performance/situational) with memory of exercise experience particularly relevant in the production paradigm. Following an exercise situation, memory will degrade and impact upon future active productions. In comparison, the estimation paradigm is based upon the interpretation of current stimulation. When using the estimation paradigm, the subject is first asked to describe or estimate his or her RPE at intervals dictated by the researcher.[1] The application of RPE in this manner is routinely used during graded exercise, but it is also employed during submaximal bouts of steady-state exercise. In the production protocol paradigm, the subject is asked to self-regulate exercise intensity by producing a pre-determined RPE, which is often anchored to a given exercise intensity determined from an estimation paradigm. During the production trial, the subject will self-adjust the exercise intensity to reproduce and maintain a given physiological effort. To avoid order effects, the subject is often asked by the researcher to produce different exercise bouts in an order selected randomly by an instructor.[1] Prescription congruence is tested by determining a physiological response (i.e. HR, V-dotO2, percentage of maximal HR, or percentage of maximal oxygen uptake [V-dotO2max]) equivalent to a target RPE. When the physiological response does not differ between estimation and production trials, prescription congruence is accepted.[16]

Van den Burg and Ceci[17] have tested the accuracy of RPE in exercise prescription to achieve a HR target. The authors reported that the HR response equivalent to 13 did not differ significantly between the estimation (145 beats/min) and the production trials (149 beats/min). In addition, Robertson et al.[18] examined also the validity of a perceptually based cross-modal prescription using treadmill and stationary cycling. They found no exercise mode effect when the exercise intensity was set equal to the relative V-dotO2max of the subjects tested. Moreover, Dunbar et al.[19] reported in adults no significant difference between estimation and production trials performed at 50% of V-dotO2max (mean error <10%) during treadmill and cycling exercise. In summary, most of the studies where exercise intensity is regulated by producing a previously estimated RPE support the validity of a perceptually based exercise prescription in apparently healthy adults.

This article discusses the key findings from the published literature concerning perceived exertion in relation to the cognitive developmental level of healthy subjects. More specifically, this relationship will be described during various exercise modes and experimental protocols. This article will present the different developmental periods reported by Piaget[20] in relation to the level of cognitive maturation in children. Piaget[20] observed that the cognitive development of children is not a linear continuum, but a step by step process characterised by four distinct periods. The author found that an adaptation mechanism constituted by assimilation and accommodation phases allows the transition between these periods, which are: (i) the sensory-motor period (from 0 to 3 years); (ii) the pre-operational period (from 4 to 7 years); (iii) the concrete operations period (from 8 to 12 years); and (iv) the formal intelligence period (from 13 years to adult).

Two additional periods will be determined for aged subjects (>50 years) according to Poortmans,[21] because as maximal HR decreases with age,[22,23] the RPE-HR relationship also might change. Furthermore, the loss of the sensibility of proprioceptors caused by sarcopenia, decreases the speed and quality of nervous propagation.[21] In addition, it has been reported that elderly persons could have degradation in cognitive performance, particularly for perception tasks.[24,25] All of these variables may affect perceived exertion in aged persons. Therefore, the aim of this article is to discuss how the cognitive functions involved in perceived exertion evolve consecutively during these six periods.

1. The Sensory-Motor Period: 0–3 Years

Not surprisingly, there is little research on the utilisation of RPE during exercise in children in this age range (table I). In a study involving handgrip force,[26] children with an average age of 3.3 years were not able to produce, in graded or in randomised order, four handgrip force levels. According to Piaget,[20] the upper end of this age group corresponds to the end of the sensory-motor period where young children are unable to perform a relationship between the symbolic representation suggested by a picture and the different grip-force levels. Furthermore, children of this age are not able to rate their perceptions; therefore, it is not surprising that they cannot accurately reproduce a given effort based on RPE.[26] Overall, these findings suggest that the cognitive developmental level in children 0–3 years of age does not allow for the application of perceived exertion during physical activity.
Table I

Publications involving rating of perceived exertion (RPE) in 3- to 7-year-old children

2. The Pre-Operational Period: 4–7 Years

Compared with the limited number of studies involving children in the 0- to 3-year age group, there are more studies that have examined perceived exertion responses during exercise in children whose ages fall within the pre-operational period (table I). Williams et al.,[4] using an estimation paradigm, examined young children (4–5 years of age) using the CERT, the first published scale to use a limited number range (1–10) in recognition of the limited cognitive ability of young children. They reported reasonably good correlations (r = 0.73) between perceived exertion and HR across increasing exercise intensity. Groslambert et al.[11] using the RPE-C also found a significant HR-RPE correlation (r = 0.78), although reliability coefficients within different levels of exercise ranged from r = 0.17 to 0.77. Williams et al.[4] reported that children aged 4.7 years did not understand the procedures to be followed for a perceived exertion production trial using the CERT and therefore were unable to assess the accuracy for children of this age to use perceived exertion to determine exercise intensity. However, other researchers have reported that children in this age range are able to accurately produce four handgrip force levels during an incremental handgrip test.[26] In contrast, children in this study were not able to accurately reproduce four handgrip force levels in random sequence. This result also has been reported by Piaget[20] who observed that the cognitive development level of children of this age does not allow to compare together more than two objects of different sizes.

As children advance through this developmental stage it appears that the ability to use RPE as an estimate of exercise exertion improves. Williams et al.[4] found that the correlation between perceived exertion using CERT and HR was r = 0.95 in children aged 6.7 years. However, when the children in this age group were asked to produce a given RPE (5 and 7 on the CERT), the HR assessed during an estimation trial was not significantly related to the HR observed during the production trial at both levels of effort. One hypothesis is that estimation and production protocols require two dissimilar psychological processes that are strongly related to the body experience and/or the cognitive development level of the children.[29] Another hypothesis is that these findings were related to the unique mode of exercise (stepping) and not necessarily to an inability in the children to use perceived exertion in this mode.

At the end of the pre-operational period (5–7 years), it has been reported that children can produce accurately in random order three to four handgrip force levels from their perceived exertion.[26,27] Likewise, when children of this age performed a more familiar activity that resembles as ‘real world’ play experiences (e.g. outdoor running), they were able to accurately estimate and produce three running intensities.[28] From these observations, it may be concluded that the pre-operational period is a developmental phase where children are able to progress from using peripheral sensory cues during handgrip tests to a reliance on their cardiorespiratory cues during outdoor running. This point is summarised in figure 1.
Fig. 1

Evolution of the perceived exertion capacities from 3 to 20 years.

3. The Period of Concrete Operations: 8–12 Years

It is generally recognised that at a given relative exercise intensity, measures of perceived exertion using the Borg 6–20 scale are typically lower in children versus adults either through direct comparison[30] or indirectly by comparing perceptual responses in children and adults at similar exercise intensities, but across different studies (table II).[31,32] In contrast, others have reported that perceived exertion responses using the Borg 6–20 scale are similar in children and adults when referenced to the ventilatory threshold.[33,34] In both of these studies, ventilatory threshold occurred at the same percentage of V-dotO2max in boys and men. Differences in the perceptual responsiveness between children and adults may be due to the validity of the rating scale used, particularly the Borg 6–20 RPE scale and the exercise intensity at which the comparison is being made. Indeed, the original notion that exercise HR could be determined by multiplying Borg 6–20 RPE value by 10 may be valid only for middle-aged and older individuals. The ratio does not appear to be accurate for children, adolescents or young adults.[30] In support of this, it has been reported that the RPE-HR correlation for the Borg 6–20 scale is low in children aged 9–11 years (r = 0.45–0.79,[29,35] increases during adolescence (r = 0.74–0.87)[3638] and is still higher (r = 0.89–0.95) in adults.[39,40] However, when perceived exertion is measured by a rating scale adapted for children (e.g. OMNI scale),[5] the perceived exertion-HR correlation values across increasing exercise intensity are quite similar between children aged 8–12 years (r = 0.87–0.94)[5] and adults (r = 0.81–0.90).[41] Likewise, Lamb[35,42] and Lamb and Eston[43] have highlighted the importance of using a rating scale adapted and validated for children. Their studies have shown the perceived exertion-HR relationship is more pronounced when the children used the CERT (r = 0.69–0.79)[35,44] compared with the Borg 6–20 scale (r = 0.45–0.79).[35] In contrast to these results are the findings of Utter et al.[6] who reported correlations between OMNI perceived exertion and various physiological measures of effort that were considerably lower than the correlations reported by Robertson et al.[5] Explanations for this discrepancy are not fully apparent but may be due to differences in exercise modality (treadmill[6] vs cycle ergometer[5]) and age range (6–13 years[6] vs 8–12 years[5]).
Table II

Publications involving rating of perceived exertion (RPE) in 8-to 12-year-old children

It is interesting to note that the children in this age group also can discriminate levels of exertion in different parts of their body (leg, chest and overall body) during both a graded exercise test[7,34,45] and steady-state submaximal exercise.[48] The studies by Robertson et al.[7,45] used the OMNI scale and the studies by Mahon et al.[34,48] utilised the Borg 6–20 scale. In these studies, the sensations arising from the legs, appears to have provided the dominant sensory signal in children of this age with the cardiorespiratory factors serving as a secondary cue. This suggestion also was made by Mahon et al.,[46] although they only assessed overall perceived exertion. According to Piaget,[20] this period corresponds to the developmental level where the children are progressively able to accurately distinguish feelings in the different parts of their body. However, it appears that exercise intensity must be high because it has been reported that at slow-to-moderate walking speeds neither the respiratory-metabolic nor peripheral ratings of perceived exertion appeared to dominate the whole-body sensory-integration process in children of this age.[54]

In this age group, it has been reported that V-dotO2 and HR did not differ significantly between estimation and production trials at two different levels (2 and 6) on the OMNI scale;[16] however, as should be expected, V-dotO2 and HR differed between the two perceived exertion levels. Williams et al.[49] reported that HR differed across three different Borg 6–20 RPE levels (9, 13 and 17) but not within each level in 11-year-old boys and girls; however, comparisons were restricted to a production trial only. Likewise, Eston et al.[44] using CERT levels of 5, 7 and 9 noted that the HR and power output measured during an estimation trial were significantly correlated to HR and power output during a production trial. However, HR and power output during the production trial were significantly lower than the corresponding measurements obtained in the estimation trial.

Ward et al.,[50] Lamb[42] and Eston et al.[8] also confirmed that children in this age range have some ability to perceptually discriminate various levels of exertion during production trials. Using a cycling protocol, Ward et al.[50] found that the percentage of peak aerobic power and percentage of predicted maximal HR increased significantly across four different Borg 6–20 RPE production trials (RPE 7, 10, 13 and 16). However, a similar response for velocity and percentage of predicted maximal HR were not observed in a walk/run trial.[50] There also were significant variations in the slope and intercept of the RPE-HR relationship between the estimation trial and both production trials.[50] Lamb[42] observed that power output during production trials differed significantly across four different CERT (3, 5, 7 and 9) and Borg 6–20 RPE (8, 12, 15 and 18) levels. HR also differed between perceived exertion levels with the exception of the HR between the two highest levels on each scale. Significant correlations between HR and power output were apparent for both scales, although the correlations involving the Borg 6–20 RPE scale tended to run higher than with the CERT,[42] a surprising finding considering that the CERT is designed for use with children.

Based on this information, it appears that children of this age can discriminate up to four intensities during cycle exercise.[16,42,49,50] This finding is in line with Piaget[20] who observed that most 8-year-old children are able to accurately rate together more than three objects of different sizes. The author also reported that at the end of the concrete operations period, the level of psychological development allows the children to understand the constancy of the sizes. Indeed, the children of this age group are able to determine a ‘standard perception’ (e.g. item ‘tired’ in the OMNI-scale of Robertson et al.[5]) that allows from this standard perception to compare and to rate different perceptions (e.g. not tired at all, a little tired or very, very tired). Thus, the type of rating scales and the verbal and pictorial descriptors influence the perceptual responsiveness.

4. The Formal Intelligence Period: 13–18 Years

The adolescent period corresponds to the beginning of the logical-mathematic meaning. Adolescents are progressively able to make hypotheses or to understand different mathematical concepts (table III).[20] Thus, children in this age range should have the cognitive ability to understand and accurately rate perceived exertion using the Borg 6–20 RPE scale.[3] However, the RPE-HR relationship found in this age group during a maximal incremental cycling exercise tends to be slightly lower (r = 0.74–0.87)[3638] than the relationship observed in adults (r = 0.89–0.95).[39,40] This result may be due, such as in children, to the RPE-HR ratio that is not adapted for adolescents. In addition, Pfeiffer et al.[55] reported in adolescent girls very significant OMNI-%HRmax (r = 0.86) and OMNI-%V-dotO2max (r = 0.89) correlations, compared with the Borg 6–20 RPE-%HRmax (r = 0.66) and Borg 6–20 RPE-%V-dotO2max (r = 0.70) correlations. The intraclass and single-trial reliability estimates were higher for the OMNI (r = 0.95 and 0.91, respectively) compared with the Borg scale (r = 0.78 and 0.64). The authors concluded that the OMNI scale is more valid and reliable than the Borg 6–20 RPE scale for use in adolescent girls during treadmill exercise. The comparison of the correlations assessed in the study by Pfeiffer et al.[55] in adolescents girls (r = 0.89 and 0.86) and by Utter et al.[6] in 10-year-old children (r = 0.41–0.60 for RPE-%V-dotO2max), suggest that cardiorespiratory factors involved in perceived exertion may increase in relation to aging. This finding has been recently reported by Yelling et al.[9] who observed that during estimation stepping trials using the PCERT, at each intensity level the RPE-HR relationships were higher in adolescents aged 15.3 years (r = 0.26–0.87) compared with children aged 12.4 years (r = 0.21–0.66).
Table III

Publications involving rating of perceived exertion (RPE) in 13- to 18-year-old adolescents

The RPE of adolescents also is affected by the protocol used. Marinov et al.[57] compared the RPE of adolescents at the end of a Balke and a Bruce protocol. The Balke protocol involves walking up progressively steeper grades while the Bruce protocol usually requires subjects to run. In this study, peak V-dotO2 was lower on the Balke (34.2 ± 1.8 mL/min/kg) versus the Bruce (48.6 ± 2.7 mL/min/kg) protocol. However RPE, using the CR-10 scale, was significantly lower in the Balke protocol (4.5 ± 0.8) compared with the Bruce test (6.5 ± 0.4). The Balke protocol also averaged nearly 7 minutes longer to complete. Although the RPE values from both tests seem low for maximal level of exertion, the variation in RPE between protocols might have been due to the differing modes of exercise. A similar difference between the walking and running tests was noted by Mahon and Ray[47] in slightly younger children. It is also possible that RPE differences were due to the fact that different groups of children performed the tests, so in essence the difference between the two protocols represents not only a protocol difference, but also a subject difference.

From these few observations, it seems that the cardiorespiratory factors involved in perceived exertion may increase in relation to aging. In addition, it appears that the Borg 6–20 RPE-HR relationship in adolescents is less pronounced than adults and RPE values may be influenced by the mode of protocol used. Alternatively, the OMNI scale seems to be more tightly coupled to physiological measures of strain than the Borg 6–20 scale in this age group. Surprisingly, little else is known about the RPE responses in adolescents and more research is needed to better understand the perceptual responses and the optimal rating scale to use in this age group. This is significant given that adolescence is usually a time when a child’s level of physical activity begins to decline.[58] Knowledge of a child’s perception of exercise and the physiological factors mediating perceived exertion in this age group and how it might change with further maturation may be important in promoting healthy physical activity and exercise recommendations.[59]

5. Middle-Aged (50–65 Years) and Elderly Persons (>65 Years)

Bar-Or et al.,[60] using the Borg 6–20 scale, reported a good linearity and significant correlation (r = 0.77–0.80) between RPE and HR (table IV). However, it has been found that the age-related decline in maximal HR[23] would imply that the RPE-HR relationship also might change with age.[61] Bar-Or et al.[60] reported in 41- to 61-year-old subjects that when comparisons are made at the absolute exercise intensity (i.e. the same work rate, V-dotO2 or power output), RPE is generally lower in young than in middle-aged persons. Nevertheless, when comparisons are made at the same relative exercise intensity (i.e. %V-dotO2max), no significant difference of perceived exertion was found between young and 50- to 65-year-old healthy persons.[61,62] In addition, Ceci and Hassmen[63] observed that healthy 33- to 65-year-old men were able to accurately discriminate three running exercise intensities confirming the use of RPE as a means of regulating exercise intensity in this population.
Table IV

Publications involving rating of perceived exertion (RPE) in 54- to 75-year-old adults

With aging, some authors have reported that cerebral flow will diminish, which may lead to a decrease in cognitive functions that may affect perceived exertion in elderly persons.[24,25] To our best knowledge, only four studies carried out on perceived exertion have been performed in elderly persons.

Borg and Linderholm[22] reported in 18- to 79-year-old healthy male subjects that HR at a given rating decreased with increasing age. More recently, Shigematsu et al.[67] reported a very strong RPE-HR relationship (r = 0.95) in elderly women (75.5 years) during a maximal graded cycle ergometer test. A recent study involving physically deconditioned persons 75.2 years of age showed no significant relationship between RPE and HR during the course of a graded arm test to maximal exertion.[68] However, following 6 weeks of arm training, a significant HR-RPE relationship was found in most of the subjects. This result suggests that training may have increased the subject’s ability to detect muscular sensations and the ability to utilise these sensory cues in the perception of effort. In addition, physical exercise may enhance cerebral perfusion and oxygen delivery, and also increase the level of essential neurotransmitter (serotonin, noradrenaline [norepinephrine] and dopamine) responsible for memory capacities.[69] Finally, Dunbar and Kalinski[66] reported that 70-year-old women can accurately use RPE to regulate exercise intensity during a 20-week training programme. However, at intensities >40% of V-dotO2max, an acclimation period is needed. Therefore, according to Hughes et al.,[70] it may be possible that in elderly persons, perceived exertion is more affected by the physical fitness and the health status of the subject than by aging alone.

Overall, these findings suggest that there is a good linearity between RPE and HR in middle-aged and elderly healthy persons. In addition, when comparisons are made at the same relative exercise intensity, no significant difference of perceived exertion was found between young and 50- to 75-year-old healthy persons. Furthermore, when the level of aerobic fitness is controlled, age differences in perceptual responsiveness may be not present. In summary, RPE is not impaired by aging in healthy middle-aged and elderly persons and could be associated with HR as a useful tool for the monitoring and prescription of exercise.

6. Conclusions

RPE appears to be a cognitive function that reflects a long and progressive developmental process from 4 years of age to adult. Before 4 years, children are not able to rate their perceived exertion with a high degree of accuracy. From 4 years, it seems that the peripheral cues provided from muscles and joints are at first involved in perceived exertion. After 5 years, the cardio-respiratory factors are progressively involved and allow 6-year-old children to estimate and produce accurately three levels of running intensity from their perceived exertion. From 8 to 12 years of age, children are able to discriminate up to four levels of exertion but are sensitive to the exercise mode and also the rating scale used. In adolescents, it appears that the RPE-HR correlation is less pronounced than adults and RPE values are influenced, like in children, by the mode of protocol used. In middle-aged and elderly healthy persons, RPE is not impaired by aging and could be associated with HR as a useful tool to control exercise intensity. While much is known about 8- to 12-year-old children, more research is needed to better understand the perceived exertion developmental steps in 4- to 7-year-old children, but also in adolescents and elderly persons. Further research involving these different populations is recommended.


No sources of funding were used to assist in the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this manuscript.

Copyright information

© Adis Data Information BV 2006