Anaerobic Threshold • 299
Anaerobic Threshold: The Concept and Methods
Krista Svedahl and Brian R. MacIntosh
Svedahl, K., and MacIntosh, B.R. (2003). Anaerobic threshold: The concept and methods
of measurement. Can. J. Appl. Physiol. 28(2): 299-323. © 2003 Canadian Society for
Key words: maximal lactate steady state, lactate threshold, ventilatory threshold, OBLA,
individual anaerobic threshold
Mots-clés: maximum de lactate en régime stable, seuil de lactate, seuil ventilatoire, SAS
(OBLA), seuil anaérobie individuel
The anaerobic threshold (AnT) is defined as the highest sustained intensity of exercise for
which measurement of oxygen uptake can account for the entire energy requirement. At the
AnT, the rate at which lactate appears in the blood will be equal to the rate of its disappear-
ance. Although inadequate oxygen delivery may facilitate lactic acid production, there is no
evidence that lactic acid production above the AnT results from inadequate oxygen deliv-
ery. There are many reasons for trying to quantify this intensity of exercise, including as-
sessment of cardiovascular or pulmonary health, evaluation of training programs, and cat-
egorization of the intensity of exercise as mild, moderate, or intense. Several tests have
been developed to determine the intensity of exercise associated with AnT: maximal lactate
steady state, lactate minimum test, lactate threshold, OBLA, individual anaerobic thresh-
old, and ventilatory threshold. Each approach permits an estimate of the intensity of exer-
cise associated with AnT, but also has consistent and predictable error depending on proto-
col and the criteria used to identify the appropriate intensity of exercise. These tests are
valuable, but when used to predict AnT, the term that describes the approach taken should
be used to refer to the intensity that has been identified, rather than to refer to this intensity
as the AnT.
The authors are with the Human Performance Laboratory, Faculty of Kinesiology,
University of Calgary, 2500 University Dr. NW, Calgary, AB, T2N 1N4.
300 • Svedahl and MacIntosh
Le seuil anaérobie (AnT) correspond au plus haut niveau d’intensité d’effort physique dont
l’énergie provient exclusivement de métabolisme aérobie. Au seuil anaérobie, la quantité
de lactate diffusant dans le sang est égale à la quantité en sortant. Bien qu’un transport
d’oxygène inadéquat puisse accroître la production d’acide lactique, il n’y a pas d’indication
solide voulant que la production d’acide lactique au-delà du seuil anaérobie soit due à un
transport d’oxygène déficient. Nombreux sont les arguments militant en faveur de la quan-
tification de l’intensité d’exercice au seuil anaérobie, notamment: évaluation de la santé
cardiovasculaire ou pulmonaire et des programmes d’entraînement ainsi que la
catégorisation de l’intensité de l’effort soit léger, modéré, et intense. Des tests ont été mis
au point pour indiquer l’intensité d’exercice au seuil anaérobie: maximum de lactate en
régime stable, minimum de lactate, seuil de lactate, SASL (OBLA), seuil anaérobie individuel,
et seuil ventilatoire. Chacune de ces approches donne une estimation de l’intensité d’exercice
au seuil anaérobie, mais l’erreur associée varie selon le protocole d’évaluation et les critères
d’identification de l’intensité d’exercice. Ces tests sont utiles, mais quand ils servent à
établir le seuil anaérobie, on devrait préciser le nom de l’approche utilisée pour identifier
l’intensité d’effort au seuil anaérobie plutôt que d’associer cette intensité au seuil anaérobie.
Few concepts in the field of exercise science have generated such debate as that of
anaerobic threshold. Disagreement among researchers stems not only from the
absence of methodological standardization but also from a lack of consensus on
the theoretical basis of the concept itself. Efforts to accurately describe a threshold
intensity have resulted in an immense pool of scientific data. Yet the issue remains
an unresolved controversy. One reason for the ongoing controversy is the lack of
consensus for the definition of anaerobic threshold and the persistent inappropri-
ate use of the term. It is important to recognize that anaerobic threshold is a con-
cept, and that the definition is a conceptual definition. In contrast, the various
ways to detect the intensity of exercise associated with the anaerobic threshold
have resulted in a proliferation of terms that are more appropriately given opera-
tional definitions. These measurements should not always be equated with anaero-
bic threshold, since there are clear differences between conceptual and operational
definitions. Considering the inconsistency with which these terms are used, read-
ers should interpret a term like anaerobic threshold or lactate threshold from the
context of its use.
The purpose of this review is to provide a conceptual definition of anaerobic
threshold and related terms, and to discuss the theoretical concept and methods of
measurement. An historical perspective on the meaning of anaerobic threshold is
presented, with a discussion of likely (and unlikely) mechanisms. This is followed
with a brief description of some of the tests that have been proposed as providing
an estimate of the anaerobic threshold.
These definitions are generalized and are intended to provide a framework for
subsequent discussion. The definitions are elaborated upon later in this paper.
Anaerobic threshold: The term “anaerobic threshold” is defined as an in-
tensity of exercise, involving a large muscle mass, above which measurement of
Anaerobic Threshold • 301
oxygen uptake cannot account for all of the required energy. Stated in other terms,
this is the exercise intensity above which there is a net contribution of energy
associated with lactate accumulation.
Maximal lactate steady state: Maximal lactate steady state (MLSS) is de-
fined as “the highest exercise intensity at which blood lactate concentration does
not increase beyond the initial transient during constant load exercise” (Tegtbur et
al., 1993, p. 620). In other words, the intensity at MLSS represents the highest
intensity for which there is an equilibrium between lactate transport into the blood
and lactate removal from the blood (Heck et al., 1985).
Lactate minimum speed: The lactate minimum speed is the speed of loco-
motion at which blood lactate reaches a minimal value during an incremental exer-
cise test (increments in speed of locomotion), which is initiated in the presence of
Lactate threshold: Lactate threshold is the exercise intensity that is associ-
ated with a substantial increase in blood lactate during an incremental exercise
test. Various specific criteria are used to identify this increase, and some of these
have their own special name.
Onset of blood lactate accumulation: Onset of blood lactate accumulation,
or OBLA, is defined as the intensity of exercise at which blood lactate concentra-
tion reaches 4 mM during an incremental exercise test (Sjodin et al., 1981).
Individual anaerobic threshold: The individual anaerobic threshold (IAT)
is a special version of a lactate threshold. IAT is defined as the intensity of exercise
identified by a line drawn from a recovery lactate concentration, tangent to the
lactate concentration observed during an incremental test (Stegmann et al., 1981).
Ventilatory threshold: Ventilatory threshold is defined as the exercise in-
tensity at which the increase in ventilation becomes disproportional to the increase
in power output or speed of locomotion during an incremental exercise test.
What Is the Anaerobic Threshold?
The definition of anaerobic threshold relates to exercise involving a large muscle
mass. It is recognized that within a single muscle, glycolysis can occur, resulting
in net output of lactate even at rest (Gladden, 2000; Stainsby et al., 1984; 1991).
Under these circumstances, measurement of oxygen uptake could not account for
all the energy use by the muscle. Therefore the concept of anaerobic threshold
must apply only to the intact whole body when a substantial portion of the muscle
mass is active. To understand the concept of anaerobic threshold, it is important to
understand the metabolic systems that provide energy during exercise.
Technically speaking, if “anaerobic metabolism” is defined as replenishment
of ATP without the use of oxygen, then substrate level phosphorylation would be
considered anaerobic. This would include reactions associated with creatine kinase,
glycolysis, and the Krebs cycle. Since measurement of oxygen uptake permits ac-
counting for some of these steps, the presence of glycolytic activity is not necessar-
ily evidence that the exercise intensity has exceeded the anaerobic threshold.
Typically, pyruvic acid resulting from glycolysis is either incorporated into
oxidative metabolism via the Krebs cycle or is converted to lactic acid. The con-
302 • Svedahl and MacIntosh
version of pyruvic acid to lactic acid is a valuable step in that cytoplasmic NADH
is oxidized. This ensures a continued supply of NAD
for glycolysis. Therefore,
instead of inhibiting glycolysis, lactic acid formation permits continued glycoly-
sis. Furthermore, it is clear that lactate can be oxidized either within the muscle
fiber in which it is produced (Brooks, 2000; Brooks et al., 1991) or in an adjacent
fiber or another muscle (Donovan and Pagliassotti, 2000). In this case, measure-
ment of oxygen uptake could account for this glycolytic production of ATP.
It is the accumulation of lactate or other glycolytic intermediates, not simply
evidence of lactate production, which should be considered to represent the meta-
bolic rate above anaerobic threshold. This accumulation could be in muscle tissue
and/or in the blood. Accumulation of lactate represents the situation whereby gly-
colytic production of pyruvic acid and lactic acid exceeds the rate of incorporation
of these molecules into the Krebs cycle. It seems reasonable to assume that if
lactate is accumulating in the blood while exercise intensity is constant, then the
intensity of exercise exceeds the anaerobic threshold, as defined above.
Is There an Anaerobic Threshold?
To address the question of whether or not there is an anaerobic threshold, it is
important to consider the fate of lactate in the body (see Donovan and Pagliassotti,
2000). As mentioned above, a single muscle can have a net lactic acid production
even at rest. However, it is known that lactate may be taken up and oxidized in
another organ or tissue in the body. In defining anaerobic threshold, the ultimate
(short-term) fate of the lactate that is released from a muscle must be considered in
order to determine whether that lactate represents accumulation. If the lactate (or
pyruvate) that makes its way to the blood is subsequently taken up by another
muscle or other organ, and oxidized, then it would not accumulate. If on the other
hand the lactate that is released from a muscle results in increasing blood lactate
concentration, then the measurement of oxygen uptake could not account for the
ATP replenishment associated with that lactate formation. Therefore, by defini-
tion, when blood lactate concentration increases over a prolonged duration at a
given intensity of exercise (power output or speed of locomotion), the intensity
would be considered as being above the anaerobic threshold.
When blood lactate concentration is not increasing, the rate of lactate re-
moval from the blood must equal or exceed the rate at which lactate is moving into
the blood. If all of the lactate removed from the blood is oxidized, the intensity of
exercise would be considered as being at or below the anaerobic threshold. How-
ever, lactate can have several pathways of metabolism. Lactate (or pyruvate) can
be taken up by the liver and the kidney, and undergo gluconeogenesis. It is pos-
sible that oxidative metabolism in the liver or kidney provides the energy needed
to transform the lactate and/or pyruvate back to glucose by oxidative metabolism,
so this pathway of disposal represents a means by which oxygen uptake can ac-
count for the glycolytic formation of ATP. If we accept this argument, then the
anaerobic threshold would occur at the highest intensity of exercise for which a
steady state for blood lactate can be sustained. This intensity of exercise has also
been referred to as the maximal lactate steady state (MLSS).
The only circumstance when MLSS would not be equal to anaerobic thresh-
Anaerobic Threshold • 303
old would be if blood lactate concentration could remain constant while lactate
accumulates in muscles. Under these circumstances the measurement of oxygen
uptake cannot account for glycolytic ATP formation, although the rate of lactate
entry into the blood was equal to the rate of lactate removal from the blood. This
would probably occur if the volume of active muscle was relatively small. Other-
wise it should be considered that MLSS actually represents the intensity of exer-
cise at the anaerobic threshold.
Since the anaerobic threshold refers to an intensity of exercise, it is impor-
tant to recognize that this intensity is presumably just one point on the intensity/
duration relationship. Figure 1 presents a typical intensity/duration relationship.
Intensity in this case is expressed as energy input. Alternatively, intensity could be
expressed as mechanical power output, or speed of locomotion. This particular depic-
tion of the intensity/duration relationship presents the duration of exercise when
intensity is at or below maximal oxygen uptake, and is based on the following: At
maximal oxygen uptake, exercise can be sustained for up to about 60 min (Billat et
al., 2000); at anaerobic threshold, which typically occurs at 60 to 80% of maximal
oxygen uptake, exercise can be sustained for up to about 60 min (Lajoie et al., 2000);
at less than the anaerobic threshold, exercise can be sustained for several hours.
Exercise intensity is best quantified by measuring the rate of metabolic en-
ergy input while performing a task. This can be done by measuring oxygen uptake
when the intensity is below the anaerobic threshold. However, the rate of oxygen
uptake is not constant while exercising at an intensity above anaerobic threshold,
and by definition, oxygen uptake does not account for all of the energy input above
this intensity. When the energy demand for the exercise is near maximal oxygen
uptake, there is a steady increase in oxygen uptake while the conditions of the
exercise (speed or power) remain constant (Gaesser and Poole, 1996). This steady
increase is called the slow component of oxygen uptake.
Figure 1. The exercise intensity/duration relationship, for intensities equal to or less
than maximal oxygen uptake. As exercise intensity decreases, time to fatigue increases.
Energy Input (J · s
Duration of Exercise (min)
304 • Svedahl and MacIntosh
It is thought that the slow component exists when exercise intensity exceeds
the anaerobic threshold (Jones et al., 1999), but the mechanism of this slow in-
crease in oxidative metabolism is not known (MacIntosh et al., 2000). Since oxy-
gen uptake is not constant when a slow component exists, it may be more appro-
priate to designate the intensity of exercise by power output or speed of locomo-
tion, but this depends on the reason for expressing the intensity of exercise. Heart
rate is often used, but this is not the most appropriate or precise means of express-
ing intensity, due to the occurrence of cardiac drift even below the apparent anaero-
bic threshold (Lajoie et al., 2000). Furthermore, there is daily variation in heart
rate response at a given intensity of exercise (MacIntosh et al., 2002).
A brief history of our understanding of the circumstances of lactate forma-
tion and its appearance in the blood will now be presented, with the intent of giv-
ing some perspective to the use of the term anaerobic threshold. It is not our pur-
pose here to provide an extensive review of this history but rather to point out
some key observations.
Historical Perspective on Lactic Acid Formation
Much of the information presented below on the early recognition of a role for
lactic acid formation in skeletal muscle metabolism has been obtained from a very
interesting book by Dorothy Needham (1971). The reader is directed to this source
for the specific references for this material.
It was recognized as early as 1807 that lactic acid was formed in muscle.
Needham (1971) indicates that Berzelius was the first to identify lactate in muscles,
and this was in the muscles of hunted stags. Another early scientist who observed
lactate in muscle was Claude Bernard, who reported that the amount of lactic acid
in muscle was proportional to previous exercise. In the early 1900s, an intensive
search to understand the biochemistry of energy metabolism resulted in consider-
able advances in the understanding of the role of lactic acid and its involvement in
providing energy for muscle contraction, but this was not without contradiction
and confusion. In 1907, Fletcher and Hopkins made a profound observation which
is as true today as it was then: “it is notorious that, quite apart from the question of
the oxidative removal of lactic acid—which has not previously we think been ex-
amined—there is hardly any important fact concerning the lactic acid formation in
muscle which, advanced by one observer, has not been contradicted by some other”
(as cited in Needham, 1971, p. 45).
It was reported by Pflüger in 1875 (as cited in Needham, 1971) that muscle
contraction could occur in an anaerobic (oxygen-free) environment. The meta-
bolic pathway that can provide energy under these circumstances came to be rec-
ognized as glycolysis. Therefore we refer to anaerobic glycolysis now, when lactic
acid is formed, whether or not oxygen is present.
This misuse of the term “anaerobic” may be an important factor in the per-
vasive misunderstanding of the circumstances in which lactic acid formation oc-
curs. The persistent use of the term anaerobic has led to the common belief that the
presence of lactic acid in muscle is evidence that oxygen delivery was insufficient
to satisfy the demand. A net production of lactic acid has often been interpreted as
a symptom of inadequate oxygen delivery (Hill and Lupton, 1923). This is not
necessarily the case.
Anaerobic Threshold • 305
The Cause of Increased Lactic Acid Formation
There is no doubt that when oxygen availability is limited, lactic acid will be formed
in muscle, making a net contribution to the provision of energy. However, this is
not sufficient rationale to conclude that the presence of lactic acid in muscle means
that limited oxygen availability was restricting oxidative metabolism. It is impor-
tant to consider whether lactic acid can be formed in muscle when adequate oxy-
gen seems to be present. The most striking evidence for this was presented by
Jobsis and Stainsby (1968), who 35 years ago showed that the mitochondrial re-
dox state becomes more oxidized when contractions are initiated in the dog gas-
trocnemius muscle. Graham and Saltin (1989) confirmed that the mitochondrial
redox state (NAD
/NADH) rose in humans during exercise at a time when lactic
acid formation was accelerated.
Lactic acid production is known to be accelerated when contractions are
initiated (Stainsby et al., 1991). Richardson et al. (1998) and Connett et al. (1984)
have shown that oxygen availability is sustained when lactic acid formation is
substantial. Recently Hogan (2001) has shown that it is not lack of oxygen that
stimulates the glycolysis which results in lactic acid formation at the onset of exer-
cise. In Hogan’s study it was observed that oxygen content of single skeletal muscle
fibers decreases with a time constant similar to the time constant for the increase in
oxygen uptake. This observation confirms that the relatively slow increase in oxy-
gen uptake at the start of exercise is not due to limitations in oxygen delivery.
Presumably, oxidative metabolism has a high inertia, and phosphocreatine and
glycolysis provide the ATP replenishment while oxidative metabolism is acceler-
ated. Glycolysis resulting in the formation of lactic acid should be interpreted as a
process occurring without the use of oxygen, not necessarily in the absence of
oxygen. It is now recognized that although hypoxia may result in increased forma-
tion of lactic acid, absence of oxygen is not a prerequisite for lactic acid formation
(see review by Gladden, 1996).
Several factors can promote lactic acid formation in muscle. One of these is
accelerated glycogenolysis and glycolysis (Febbraio et al., 1998; Richter et al.,
1982; Stainsby, 1986), resulting from increased sympathoadrenal activity. The
control mechanism for activation of phosphorylase-b was delineated by Rall et al.
(1957). This effect of sympathoadrenal enhancement of lactic acid formation could
very well be the primary mechanism for the marked elevation of blood (or plasma)
lactate during an incremental test. Mazzeo and Marshall (1989) reported a strong
correlation between plasma lactate and epinephrine concentration among runners
and cyclists during incremental exercise tests. They also observed no significant
difference between an inflection in plasma lactate concentration and an inflection
in plasma epinephrine concentration when expressed as a percent of maximal oxy-
gen uptake. This was the case for both cycling and running tests, although the
inflections occurred at different relative intensities (i.e., higher for running among
runners and higher for cycling among cyclists).
McMorris et al. (2000) also found a significant correlation between power
output at lactate threshold and power output at catecholamine thresholds, but they
argue that there is not a clear cause and effect (lactate threshold sometimes pre-
ceded the catecholamine threshold). In contrast, Dickhuth et al. (1999) found low
correlations between catecholamine and lactate thresholds. These discrepancies
306 • Svedahl and MacIntosh
between investigators probably relate to the different criteria for identification of
the thresholds. This is consistent with the observations of McMorris et al. (2000),
who evaluated different criteria for detection of the lactate threshold and found
varying correlations with a catecholamine threshold.
Other factors may contribute to accelerated lactic acid formation. Another
possible reason for increased lactic acid formation in a muscle is inadequate trans-
fer of reducing equivalents to the mitochondria (Holloszy and Coyle, 1984). Un-
der these circumstances, lactate formation can help to maintain the NAD
ratio in the cytoplasm. The lactate-to-pyruvate ratio would be expected to increase
under these circumstances, a symptom that Wasserman interprets as indicative of
oxygen limitation (Wasserman et al., 1999). See Graham (1991) and Gladden (1996)
for further discussion of these metabolic implications.
Blood lactate concentrations can be elevated at rest, even in the presence of
adequate oxygen delivery. It is not the presence of lactate in the blood, nor even
the presence of a concentration above resting that is important. Rather, the net
result of lactate transport into and out of the blood must be considered. At a certain
exercise intensity, the rate of lactate production and transport into the blood will
exceed the rate of removal from the blood. This could be due to redistribution of
blood flow away from lactate removal sites (nonexercising muscle, liver, kidney,
heart), or to transformation of some tissue from lactate removal sites to lactate
producing sites, as the intensity of exercise increases. This includes recruitment of
additional motor units within an active muscle, since some lactate is likely to dif-
fuse between active and inactive muscle cells within a muscle (Karlsson and Jacobs,
1982). As the pool of motor units becomes more active, there are fewer inactive
(or only mildly active) muscle fibers available to serve as lactate removal sites.
Under these circumstances, lactate will accumulate and measurement of oxygen
uptake cannot account for all the energy requirements of the exercise.
Origin of the Concept of a Threshold Intensity
The notion of a “threshold” or intensity of exercise, above which there is accumu-
lation of lactate, also has a long history of scientific investigation. Owles (1930)
wanted to quantify lactate in the blood during low intensity exercise and found
that when the exercise was mild, the blood lactate concentration did not rise above
resting values. However, at intensities well below maximal oxygen uptake, blood
lactate was above resting levels. Consistent with the point raised above, Owles
interpreted this to indicate that oxygen delivery became insufficient, leading to the
formation of lactic acid. This appears to be the first reference to a relevant thresh-
old intensity of exercise. However, it should be noted that Owles measured lactate
after 30 min of constant intensity exercise, and when lactate was elevated above
the resting level, he inferred that there had been an accumulation of lactate. There
was no attempt to determine whether blood lactate concentration was changing at
this intensity of exercise, so this accumulation cannot be related to the anaerobic
There is no question that as exercise intensity increases, there will be a higher
concentration of lactate in the blood regardless of the underlying cellular mecha-
nisms. In the 1950s and early 1960s Hollmann and colleagues (see review by
Hollmann, 1991) started using the measurement of blood lactate in submaximal
Anaerobic Threshold • 307
exercise tests to detect a critical intensity of exercise indicative of exercise intoler-
ance in cardiac and pulmonary patients. They assumed that if arterial blood lactate
could be maintained at a constant level, then the exercise was “purely aerobic.” This
could be considered the beginning of a concept of maximal lactate steady state.
The term “anaerobic threshold” was proposed by Wasserman and McIlroy
(1964). Similar to Hollmann, they wanted to identify an intensity of exercise that
provided a substantial, yet safe, amount of physical stress for patients suffering
from cardiovascular disease. Their rationale was that if a submaximal test could
reliably detect an objectively determined level of stress, then it would not be nec-
essary to expose these patients to maximal exercise testing. They saw some value
in identifying the intensity at which there appeared to be a limitation in the cardio-
vascular system’s ability to deliver oxygen to the working muscles. Wasserman
and McIlroy believed that when this occurred there would be a substantial increase
in blood lactate concentration, and proposed identifying this intensity of exercise
in several ways. They reported that the anaerobic threshold was associated with
decreased plasma bicarbonate and pH, as well as increased R (respiratory exchange
ratio) and increased ventilatory equivalent for CO
). Within the scope
of their initial work, little consideration was given to the notion that these three
events may not occur in synchrony. This problem was exacerbated by the selection
of the term “anaerobic threshold” to designate this intensity of exercise—a term
that has instigated much debate and controversy. Wasserman still maintains that
accumulation of lactate in the blood is a symptom of inadequate oxygen delivery
(Wasserman et al., 1999).
Wasserman and McIlroy used an incremental test to identify the anaerobic
threshold. With this type of test, a steady state of lactate transport into and out of
the blood would not be established. However, there were several compelling rea-
sons to use an incremental test, and the utility of this approach is not reduced by
the disagreements over a potential underlying mechanism for the increased lactic
acid formation. Indeed, although some symptoms associated with this threshold
intensity of exercise are commonly seen in healthy subjects and even in endurance
athletes, it is not certain that cardiac patients are not limited by oxygen delivery at
the intensity that is associated with an accelerated accumulation of lactate in the
blood. The disagreements over mechanism should not detract from the value of an
incremental test to identify an objectively determined intensity of exercise associ-
ated with metabolic stress.
Discussion of the cellular mechanisms associated with lactic acid formation
and accumulation of lactate in the blood will probably continue to be debated for
many years. However, it can be agreed that there is an intensity of exercise above
which lactic acid will accumulate in the blood, and several tests have been devel-
oped to detect this intensity. It can also be agreed that detection of this intensity is
an important predictor of capability for endurance exercise, a fact that was not
considered when the concept was first proposed.
It is too late to suggest changing the name of the anaerobic threshold. The
use of the term is pervasive, not only in the scientific and clinical literature but also
by coaches, athletes, and people who exercise regularly. However, the various
ways of detecting an intensity of exercise, above which measurement of oxygen
uptake cannot account for all of the energy use, provide disparate results. For this
reason, it should be acknowledged that these methods provide only an estimate or
308 • Svedahl and MacIntosh
approximation of the anaerobic threshold, and it is strongly recommended that
terms with appropriate operational definitions be used in place of the term anaero-
bic threshold. These alternatives are presented below, but first, the rationale for
undertaking this measurement is provided.
Why Quantify the Anaerobic Threshold?
As noted, the initial purpose for estimating the anaerobic threshold was to assess
exercise capacity in cardiac patients (Wasserman and McIlroy, 1964). Clinical as-
sessment or approximation of the anaerobic threshold is also useful in respiratory
disease (Hollmann, 1991). Tests for the detection of anaerobic threshold have also
gained widespread use in athletic populations (Beneke, 1995; Billat, 1996; Jenkins
and Quigley, 1990; Rusko, 1992; Sjodin et al., 1982). Focus has shifted away from
maximal oxygen uptake (V
max) as a predictor of success in endurance perfor-
mance, because studies have shown poor correlations between V
max and per-
formance results when athletes with similar V
max values are compared (Costill
et al., 1973; Hagberg and Coyle, 1983). In addition, endurance performance of
trained athletes continues to improve even after V
max levels have ceased to
improve with further training. For example, despite similar V
max values be-
tween junior-age and adult elite runners, the younger athletes were unable to per-
form at the same level (Murase et al., 1981).
For middle- and long-duration exercise, V
max may not be the best predic-
tor of endurance capability. It has been realized that athletes who can utilize a
larger fraction of their V
max for the duration of an endurance event will perform
better than those who are physiologically limited to completing the event at a lower
intensity. It has been demonstrated that various techniques which purport to mea-
sure anaerobic threshold provide a good estimate of the fraction of V
can be sustained in endurance exercise (Bassett and Howley, 2000; Coyle et al.,
1988; Kindermann et al., 1979). Consequently, submaximal performance indica-
tors, most of which claim to measure some type of “threshold,” have gained wide-
A cautionary note is presented here. It is true that the intensity of exercise
(oxygen uptake, power output, velocity of locomotion) at some measured thresh-
old (lactate threshold, OBLA, ventilatory threshold, etc.) may provide an accurate
prediction of performance in endurance events. However, a high V
max is still a
prerequisite for elite caliber performance in such events.
The concept of anaerobic threshold is also commonly referred to in training
programs. Measurement of the anaerobic threshold provides a benchmark inten-
sity around which training programs can be designed. Exercise performed at an
intensity around the anaerobic threshold would be considered moderate, while
exercise below this intensity would be mild. When the intensity of exercise sub-
stantially exceeds the anaerobic threshold (i.e., approaches V
max), it would be
considered intense. There are differences in the adaptations that occur due to train-
ing at various intensities, and the most appropriate intensity of training depends on
the goal of the program. It is beyond the scope of this review to further evaluate
the consequences of training at these specific intensities.
It is recognized that there are several reasons for identifying the intensity of
exercise associated with the anaerobic threshold, and several methods have been
Anaerobic Threshold • 309
proposed for these purposes. To be useful, the method must be reproducible and
must identify the threshold with some accuracy. The method must also be objec-
tive. Evaluation of the advantages of such tests should include practical consider-
ations for the subjects, including time commitment, invasiveness, and cost. Sev-
eral approaches are described below.
How the Anaerobic Threshold is Detected
An operational definition relates to the manner in which a measurement is ob-
tained. For example, if the intensity of exercise associated with anaerobic thresh-
old is identified by determination of the intensity at which blood lactate remains at
a steady state, then the term “maximal lactate steady state” is a more appropriate
manner of referring to that intensity than to say it is the anaerobic threshold. Maxi-
mal lactate steady state is defined operationally by the method of obtaining this
measure. Definitions and explanations for MLSS and several other terms that should
be operationally defined are presented below.
MAXIMAL LACTATE STEADY STATE
Maximal lactate steady state (MLSS) is defined as “the highest exercise intensity
at which blood lactate concentration does not increase beyond the initial transient
during constant load exercise” (Tegtbur et al., 1993, p. 620). In other words, the
intensity at MLSS represents a point of equilibrium between lactate transport into
the blood and lactate removal from the blood (Heck et al., 1985). Under these
circumstances lactate is not accumulating, measurement of oxygen uptake can ac-
count for the energy requirement of the exercise, and exercise time to exhaustion
is relatively long. As previously noted, MLSS is equivalent to the anaerobic thresh-
old, as long as there is no progressive accumulation of lactate or other glycolytic
intermediates in the muscles.
Although it is unclear as to where the term “maximal lactate steady state”
originated, the term “maximal steady state” was used by Londeree and Ames (1975).
Their work examined the ability of several maximal steady-state criterion mea-
sures to predict level of conditioning. Differences were observed in heart rate and
oxygen uptake at blood lactate concentrations of 2.2 and 4.4 mM between groups
with varying levels of conditioning. No mention was made of a maximal lactate
steady state; however, the exercise intensity at which blood lactate increased from
10 to 15 min of a constant-intensity treadmill test was identified. This intensity
was considered to be that at which glycolysis, leading to the formation of lactic
acid, began to make a net metabolic contribution. The concept and terminology
was explored more extensively by Stegmann and colleagues (Stegmann et al.,
1981; Stegmann and Kindermann, 1982) and Heck et al. (1985). More recently,
the work of Tegtbur et al. (1993) seemingly reintroduced the concept of MLSS as
a valid parameter for athletic testing and training.
The only valid method for measuring MLSS involves blood sampling dur-
ing multiple sessions of constant-intensity exercise over a range of intensities. The
constant-intensity tests should last at least 20 min (Aunola and Rusko, 1992), but
tests lasting 30 min or longer have been used more commonly (Beneke, 1995;
Beneke and von Duvillard, 1996; Jones and Doust, 1998; Swensen et al., 1999). In
310 • Svedahl and MacIntosh
theory, the range of selected intensities should include the intensity corresponding
to MLSS, in addition to an intensity slightly above it. At MLSS, an initial increase
in blood lactate concentration will occur, followed by a steady-state condition for
blood lactate. The curve depicting blood lactate over time that is generated at exer-
cise intensities below the MLSS will also show an initial increase, but this will be
followed by a gradual decrease in blood lactate concentration. Above the MLSS,
blood lactate levels are expected to rise steadily throughout the exercise session
(see Figure 2).
The commonly accepted criterion for achieving MLSS is the highest inten-
sity of exercise for which there is a change in blood lactate concentration of no
more than 1.0 mM during the final 20 min of constant-intensity exercise lasting at
least 30 min (Carter et al., 1999; Heck et al., 1985; Jones and Doust, 1998; Swensen
et al., 1999). However, more stringent criteria have been used, such as changes in
blood lactate concentration of no more than 0.2 to 0.5 mM (Aunola and Rusko,
1992; Haverty et al., 1988).
The increment in exercise intensity that is needed to accurately reflect MLSS
has not been established. Considering the substantial change in endurance that is
expected for a small change in intensity of exercise at the anaerobic threshold (see
Figure 1), the increments between two constant-intensity exercise bouts should be
very small. We have recently observed that an increase in cycling speed of just 0.9
, or approximately 2.5%, gives an increase in plasma lactate concentration
of 0.7 mM during the last 20 min of a 30-min bout of exercise, while at the lower
intensity, plasma lactate was unchanged (MacIntosh et al., 2002). Commonly, re-
searchers involved in the study of MLSS use step increases in intensity of 4 to 5%.
The precision of the estimate of MLSS depends on the size of increment in inten-
sity between tests. Essentially a series of tests will yield an intensity that is clearly
Figure 2. Blood lactate concentration over time for three exercise conditions relative to
maximal lactate steady state (MLSS): below MLSS (diamonds), at MLSS (squares),
above MLSS (triangles).
Anaerobic Threshold • 311
above MLSS (blood lactate increased during the final 20 min of a 30-min test) and
an intensity just less than this, which will be at or below the MLSS.
Narrowing the intensity range over which trials must be conducted is one of
the challenges in devising a strategy to determine MLSS in as few trials as pos-
sible. Investigators have come up with various preliminary tests that permit esti-
mation of a starting point that should be close to MLSS. These preliminary tests
typically predict an apparent anaerobic threshold, and a series of constant-speed
trials would then be conducted to establish the actual intensity at which the blood
lactate remains in steady state.
A number of other tests have been designed in order to predict MLSS, rather
than using direct measurement. Often these methods are based on the average
response to endurance exercise, such as heart rate, velocity, or time trial duration.
Foster et al. (1995) designed a protocol to predict MLSS in speed skaters by calcu-
lating the relative velocities and heart rates associated with constant blood lactate
concentrations. Swensen et al. (1999) expanded upon Foster’s work and applied it
to cycling, using a windload simulator to determine what percentage of 5-km time-
trial velocity corresponded to MLSS. Hoogeveen et al. (1997) had elite cyclists
and triathletes complete a 40-km time trial, which they deemed as representing
MLSS since heart rate and lap times remained constant throughout the test and a
steady-state blood lactate response was observed.
The problem with these approaches is that average physiological responses
are not uniformly applicable to all individuals. Predictive tests tend to overlook
one of the main conceptual advantages of MLSS, which is the fact that it is an
individualized measurement, dependent on individual lactate kinetics rather than
absolute blood lactate concentrations or percent of maximal heart rate. It should be
appreciated that group statistics can result in false confirmation of the validity of a
test. If the results of one test are no different from those of another test, this lack of
difference can be due to either true agreement or large variability between subjects
for a given test. It is not appropriate to use group statistics to validate a technique
for estimating MLSS. Individual results are more relevant. This is true for any test
to estimate the intensity of exercise close to the anaerobic threshold.
Direct measurement of MLSS, however, is not an attractive approach for the
detection of anaerobic threshold. The procedures are too time consuming and al-
ways require multiple laboratory visits for confirmation of the measurement. Al-
though the most common methods of MLSS determination employ somewhat
lengthy protocols, efforts to streamline the process are the focus of recent research
endeavors. In contrast to some of the aforementioned methods, the lactate mini-
mum test seems to be a valid and reliable method of estimating MLSS.
LACTATE MINIMUM SPEED
The lactate minimum speed (LMS) is the speed at which blood lactate reaches a
minimal value during the lactate minimum test (an incremental exercise test with
increments in speed of locomotion which is initiated in the presence of lactic aci-
dosis). The lactate minimum speed is theoretically representative of the MLSS
(Tegtbur et al., 1993).
Tegtbur et al. (1993) hypothesized that MLSS could be predicted using a
protocol consisting of two short-duration, high-intensity efforts, followed by an
312 • Svedahl and MacIntosh
active recovery period and several submaximal workloads of progressively in-
creasing intensity. The above sequence produces a “lactate minimum intensity,”
which is objectively determined by curve fitting of the resultant U-shaped lactate
curve. Verification of the relationship between the lactate minimum point and MLSS
was done with two constant-intensity tests (8-km runs) undertaken at intensities at
and above the LMS. The criterion for MLSS was met if the change in blood lactate
concentration during the final 20 min of constant-load exercise was ≤ 1.0 mM. For
all of Tegtbur’s subjects, this condition was satisfied during the constant-load test
equivalent to the LMS (mean change in blood lactate concentration 0.4 ± 0.4 mM).
However, 5 of the 25 subjects demonstrated a decrease in blood lactate concentra-
tion, indicating that perhaps they were exercising at an intensity below their true
MLSS. Running at 0.2 m·s
above the LMS produced an increase in blood lactate
concentration greater than the 1.0-mM criterion, and 11 of the 25 subjects were
unable to complete the 8-km test at this intensity.
Tegtbur’s work is unique in that it uses an incremental test with previous
lactic acidosis, resulting in a clear change in direction of the resulting lactate curve.
In comparison, other researchers have attempted to predict MLSS using incremen-
tal tests without previous lactic acidosis; therefore the lactate curve shows an ex-
ponential increase rather than a definitive turning point.
The reproducibility of the lactate-minimum test is protocol-dependent. Varia-
tion in stage duration during the incremental portion of the test has significantly
affected test results (Foxdal et al., 1996; Tegtbur et al., 1993). If intervals are not
long enough to allow an indication of the steady-state lactate exchange within the
whole body distribution space, the LMS may be inaccurately predicted. The LMS
is also affected by the initial workload for the incremental test (Carter et al., 1999).
The fact that the lactate minimum point is both intensity- and time-dependent high-
lights the importance of a valid protocol. It is critical that changes in blood lactate
values be related to the true metabolic demand of a given exercise intensity, rather
than being affected by the lactate kinetics of previous exercise stages.
The validity of using the LMS to estimate MLSS has been investigated by
Jones and Doust (1998), who reported that the LMS gave a less accurate estimate
of the velocity at MLSS than did the velocity at lactate threshold, measured by an
incremental test. We have conducted an evaluation of the lactate minimum test in
our laboratory, and have found it to be a reliable and valid predictor of MLSS
(MacIntosh et al., 2002). Clearly, more research in this area is warranted.
Advantages of this method include the fact that it is a single test, and some
of the variability and subjectivity inherent to other methods are avoided by using a
mathematical model. Also, Tegtbur et al.’s (1993) original work demonstrated that
altering glycogen stores did not affect the LMS, although absolute blood lactate
concentrations were different between normal and low-muscle-glycogen condi-
tions. A disadvantage of the LMS method is the level of effort needed for the initial
high-intensity workloads, rendering the test impractical for clinical populations.
As noted, caution must be exercised with regard to the test protocol, as protocol
manipulations have produced variability in results (Carter et al., 1999; Tegtbur et
al., 1993). Furthermore, in our experience (Svedahl and MacIntosh, unpublished),
variable results were obtained when subjects engaged in strenuous exercise up to 2
days before the test. This may represent a problem with the test, or it may reflect a
desirable sensitivity to altered metabolic and performance capabilities.
Anaerobic Threshold • 313
Lactate threshold (LT) is the exercise intensity that is associated with a substantial
increase in blood lactate during an incremental exercise test.
Lactate threshold is probably the term most commonly used in the literature
in association with estimates of the anaerobic threshold, and in most cases the use
of this term is appropriate. The specific criteria used to detect the substantial in-
crease have become important parameters of the definition, and this has led to
specific terms according to the criteria for detection of this threshold (i.e., OBLA,
or individual anaerobic threshold). For example, the substantial increase may be
detected as an increase by a fixed amount above resting blood lactate levels (i.e.,
+1 mM), or by the first intensity at which a given absolute level of blood lactate is
detected (i.e., 2 mM or 4 mM). Figure 3 presents a typical lactate curve, showing
an exponential increase in blood lactate as exercise intensity increases. Several
objective criteria for LT detection are indicated, including departure from linear-
ity, 1-mM increase above resting, absolute 4 mM, and “individual anaerobic thresh-
old” (see below).
All of these techniques will detect an intensity of exercise that is reasonably
close to the anaerobic threshold, but individual variability results in discrepancies
when each measurement is compared with the actual anaerobic threshold (or more
practically, MLSS). Parameters of the incremental test will affect the outcome,
including magnitude of increment, duration of each step, and continuous vs. dis-
Figure 3. A typical lactate curve, showing an apparently exponential increase in blood
lactate as exercise intensity increases. The following objective criteria for lactate thresh-
old detection are shown: departure from linearity (small dotted line); 1-mM increase
above resting (thick dashed line); absolute 4 mM (thick dotted line); and indivdual
anaerobic threshold (IAT, solid line). The IAT is represented by a line drawn tangent to
the blood lactate curve produced during an incremental exercise test, originating at the
time that recovery blood lactate falls to the blood lactate value observed at the highest ex-
314 • Svedahl and MacIntosh
continuous test protocols. Lactate kinetics may be quite different between con-
tinuous and discontinuous incremental tests, with some discontinuous protocols
shifting the lactate curve to the right due to lactate elimination outweighing pro-
duction during the break in exercise (Heck et al., 1985).
Break durations of 30 s have shown negligible effects (Gullstrand et al.,
1994). Workload duration (Foxdal et al., 1996; Wasserman et al., 1973), rate of
increase in work rate (Hughson and Green, 1982), blood sampling site (Robergs et
al., 1990), and measurement error (Aunola and Rusko, 1992) are all potential sources
of variability in measuring the LT. Consequently, it is important to recognize the
sources of variability and realize when results are appropriate (due to physiologi-
cal changes) versus inappropriate (due to error or inconsistency). Under similar
testing procedures and similar physiological conditions, the LT is reasonably re-
producible (r = 0.90) (Dickhuth et al., 1999).
Table 1 presents a sample of various conditions and parameters of reported
tests for LT. It is important to note that there is considerable variability in increment
durations and step sizes, as well as criteria for identification of LT. These differences
could lead to different estimates of the criterion intensity between tests. However,
if a given test provides a reliable estimate of LT, then that test will have utility.
Technology, such as portable lactate analyzers, has made utilization of the
LT more practical and convenient. LT tests are simple to administer, can often be
combined with a maximal oxygen uptake test, and a single test is sufficient for
identifying the intensity of exercise associated with the “substantial change in blood
lactate.” Furthermore, blood sampling is a minimally invasive technique and does
not demand much technical skill. The cost of supplies and equipment is reason-
able, results can be obtained quickly, and on-site lactate sampling can be used to
monitor athletes in their sport-specific environments.
The relationship between LT and MLSS is variable and largely based on
testing protocol. High correlations (r = 0.94) have been noted between running
velocity at LT and running velocity at MLSS (Jones and Doust, 1998). When care-
fully completed, LT tests will yield consistent results, and this is sufficient for
evaluating functional fitness in clinical populations or for assessing benefits from
a training program in either clinical or athlete populations.
Onset of blood lactate accumulation, or OBLA, is defined as the intensity of exer-
cise at which blood lactate reaches 4 mM during an incremental exercise test (Sjodin
et al., 1981).
This approach to the estimation of anaerobic threshold assumes that the
anaerobic threshold is synonymous with an absolute blood lactate concentration
of 4 mM, and was originally described by Mader et al. in 1976 (as cited by Heck et
al., 1985). One reason for selecting 4 mM as the blood lactate concentration asso-
ciated with OBLA was the recognition that at 4 mM muscle lactate, muscle and
blood lactate are related. This is not the case at higher and lower values (Jacobs
and Kaiser, 1982). The transport of lactate out of muscle reaches a peak rate as
muscle lactate reaches 4–5 mmol per kg wet weight (Jorfeldt et al., 1978). How-
ever, the logic of this rationale is limited since the relevant concentration is muscle
and not blood lactate. OBLA is typically measured with an incremental testing
Anaerobic Threshold • 315
Table 1 Various Lactate Threshold Test Parameters
Protocol step duration LT criterion Source
Cycle, continuous 30 W 1, 3, 5 min Breakpoint
28 W 2 min Mathematical
McMorris et al. (2000)
16 W 3 min Nonlinear increase in [la] vs V
max Neary et al. (1985)
Cycle, discontinuous 34 W 3 min Breakpoint
Henritze et al. (1985)
Treadmill, continuous 0.27 m·s
2 min Systematic increase in [la] Haverty et al. (1988)
Treadmill, discontinuous 0.5 km·h
Jones and Doust (1998)
3 min Mathematical
Dickhuth et al. (1999)
5 min Multiple
Nicholson and Sleivert (2001)
Swim, discontinuous 7
3 200 m 5 min Mathematical
Pyne et al. (2001)
Exercise intensity preceding an increase in [la] for successive workloads.
Algorithmic linear regression, log-log and semi-log transformation methods.
Plot of [la]-work rate; highest work rate not associated with an elevation in [la] above resting levels.
Until 95% HR max or 4 mM [la], then increase 1% grade each minute.
Plot of [la]-running velocity; smoothed with an equalizing spline procedure; lowest value of the ratio of [la] to performance.
Velocity preceeding two consecutive increases in [la] ≥ 1 mM; velocity associated with maximum perpendicular distance between nonlinear
regression line and straight line formed by two end-data points of blood lactate profile; velocity corresponding to [la] of 4 m
Velocity at LT calculated as a function of the slope and y-intercept from a plot of [la]-swimming velocity.
316 • Svedahl and MacIntosh
protocol and subsequent interpolation to determine the intensity of exercise that
would be expected to elicit 4 mM blood lactate.
The theoretical basis behind this method was supported by Kinderman et al.
(1979), who reported that elite cross-country skiers could sustain a constant run-
ning speed corresponding to 4 mM blood lactate for at least 45 to 60 min. How-
ever, this support ignores the fact that constant speed with sustained 4 mM blood
lactate is not the same intensity of exercise as that at which blood lactate reaches 4
mM during an incremental exercise test. Furthermore, although the average blood
lactate concentration was 4 mM, there was some variability between subjects.
Clearly, associating a lactate threshold with a fixed blood lactate concentration
ignores individual variability. For example, the sustained blood lactate concentra-
tion at MLSS ranges from 3 to 9 mM among individuals (MacIntosh et al., 2002).
The advantage with using 4 mM lactate as the criterion estimate of OBLA is
that it provides a very objective assessment of lactate threshold. A further advan-
tage is that 4 mM is substantially higher than resting levels, which can be quite
variable. This means that 4 mM will represent a rather narrow region of intensity
during an incremental exercise test (Karlsson and Jacobs, 1982). The problem with
using an absolute blood lactate concentration is the insensitivity to individual physi-
ological differences. As previously noted, many factors affect lactate production
and distribution within the exercising body. For example, since blood lactate con-
centrations are influenced by active muscle mass (Schneider et al., 2000), a fixed
blood lactate concentration represents different relative exercise intensities and
different relative contributions from glycolysis for different activities. Other im-
portant factors to consider while evaluating the usefulness of OBLA are training
status and substrate availability, particularly glycogen stores.
Although prediction of anaerobic threshold using OBLA is very objective in
that it always occurs at 4 mM, performance at the level of OBLA (i.e., workload,
heart rate, oxygen uptake) is not as consistent. In some cases endurance-trained
subjects have been unable to sustain workloads at OBLA (Foxdal et al., 1996).
Conversely, non-endurance-trained subjects have demonstrated the ability to com-
plete 50-min runs at the velocity corresponding to OBLA, but with blood lactate
levels consistently above 4 mM (Foxdal et al., 1996). These results may be attrib-
uted to physiological differences between trained and untrained individuals with
respect to the intensity (relatively higher for trained) at which 4 mM lactate was
reached. For example, the total blood volume of endurance-trained individuals
may be at least 10% greater than that of untrained individuals (Green et al., 1991).
Although it has not been directly investigated, the additional blood volume would
dilute the blood lactate concentration, resulting in a different intensity correspond-
ing to OBLA which may or may not represent a maximal lactate steady state.
Dehydration may have the opposite confounding effect.
Studies have also shown that OBLA is protocol-dependent (Foxdal et al.,
1996; Heck et al., 1985). There is evidence both supporting and refuting the use of
OBLA to predict a maximal steady-state blood lactate response. In some cases no
significant relationship has been reported between OBLA and MLSS (r = 0.57)
(Aunola and Rusko, 1992). In rowing ergometry, high correlations (r = 0.80) have
been shown between the intensity at the individual anaerobic threshold (see be-
low) and OBLA, while both were significantly higher (p < 0.01) than the workload
at MLSS (Beneke, 1995).
Anaerobic Threshold • 317
INDIVIDUAL ANAEROBIC THRESHOLD
The individual anaerobic threshold (IAT) is defined as the exercise intensity iden-
tified by a line drawn tangent to the blood lactate curve produced during an incre-
mental exercise test, originating at the time that recovery blood lactate falls to the
blood lactate value observed at the highest exercise intensity (see Figure 3). Like
OBLA, this is simply a special case of a lactate threshold.
This concept was introduced by Stegman et al. (1981) and was one of the
first attempts at providing a single test to identify the intensity at which MLSS
should occur. Theoretically, this intensity is representative of the metabolic rate
whereby the elimination of blood lactate during exercise is both maximal and equal
to the rate of lactate diffusion into the blood (Stegmann et al., 1981). The IAT is
measured by an incremental exercise test followed by a passive recovery period,
with monitoring of blood lactate levels throughout both phases of the test. Blood
lactate concentration is then plotted versus time, and a line tangent to the rising
blood lactate curve is drawn from the recovery blood lactate value that equals the
final exercise blood lactate concentration. The point of intersection of this line
with the blood lactate curve is referred to as the IAT (see Figure 3).
Essentially, the IAT represents a diffusion/elimination model derived from
blood lactate kinetics during incremental exercise and recovery (Stegmann et al.,
1981). The model presumes to take into account diffusion through biologic mem-
branes, a progressive increase in blood lactate concentration with increasing exer-
cise intensity, the existence of a lactate gradient between working muscle and blood,
and the fact that the rate of elimination approaches maximum at higher workloads.
It is assumed that the rate of diffusion and the lactate gradient are maximal at the
incremental-test end point, and that both decrease during the recovery period
(Stegmann et al., 1981). This model also assumes that the rate of decline in blood
lactate concentration during passive recovery represents the ability to dispose of
lactate. For subjects with a faster decline in blood lactate, the tangent intersects at
a higher blood lactate concentration and represents a higher intensity of exercise.
A recent study has shown that determination of the IAT is insensitive to
small changes in testing protocol (Coen et al., 2001). The protocol manipulations
included previous warm-up, variation in step duration, and test ending point (maxi-
mal or submaximal). However, changing the incremental test starting point pro-
duced significantly different results. Other researchers have reported varying re-
sults due to changes in duration of increments (McLellan, 1985) and test ending
point (McLellan et al., 1991; Urhausen et al., 1993). Under identical testing condi-
tions, the reliability for IAT determination is high (r = 0.98) (Coen et al., 2001;
McLellan and Jacobs, 1993). Endurance trained athletes have been able to sustain
exercise at the IAT for 30 min of cycle ergometry and 45 min of treadmill running
(Urhausen et al., 1993). Subjects in other studies have not been able to maintain a
steady-state lactate response while exercising at the intensity corresponding to
IAT (McLellan and Jacobs, 1993).
This method is advantageous in that it is a single test protocol and permits
individualized measurement, thereby avoiding many of the shortcomings inherent
to OBLA. It is likely not necessary for subjects to put forth maximal effort, al-
though a peak blood lactate concentration of at least 6 mM is recommended
(Urhausen et al., 1993).
318 • Svedahl and MacIntosh
The relationship between IAT and MLSS is somewhat variable. One report
indicates that in rowing, IAT occurs at a higher workload than does MLSS (Beneke,
1995). The relationship is perhaps best summarized by Urhausen et al. (1993),
who report that IAT is a reliable estimation of the range of MLSS, although the
two are not identical in all subjects.
Ventilatory threshold (VT) is defined as the exercise intensity at which the in-
crease in ventilation becomes disproportional to the increase in power output or
speed of locomotion during an incremental exercise test.
Several scientists have noted a nonlinear increase in ventilation when the
exercise intensity associated with anaerobic threshold is exceeded. This observa-
tion has led to the attempt to use ventilation to detect anaerobic threshold, and
various specific techniques have been reported. These include nonlinear increases
in ventilation and carbon dioxide output, and an increase in the respiratory gas
exchange ratio (R). However, it may be difficult to discern a clear breakpoint us-
ing these criteria, and interpretation of the data is not completely objective, with a
number of studies reporting variability between reviewers (Powers et al., 1984;
Yeh et al., 1983).
Additional criteria have been established based on the occurrence of increased
buffering when a net production of lactic acid occurs. In order to minimize the
magnitude of change in blood pH, various buffer systems are involved, including
the bicarbonate system. The reaction of H
with bicarbonate results in the forma-
tion of carbonic acid, which dissociates to H
O and CO
. This excess CO
slight fall in pH stimulate ventilation, and the extra ventilation results in excretion
of the extra CO
. An increase in the ratio of ventilation to oxygen uptake, in con-
junction with no change in the ratio of ventilation to CO
output, represents isocapnic
buffering and is considered to be a more specific method of threshold determina-
tion from gas exchange parameters (Chicharro et al., 2000; Wasserman, 1987).
However, it should be noted that the ability to observe the isocapnic buffering
region is dependent on the increment duration of the exercise protocol (Davis,
1985; Hughson and Green, 1982; Wasserman et al., 1973).
The drawback to this approach (and most of these other single test methods)
is that it does not necessarily detect the exercise intensity that can be called anaerobic
threshold. This may be because several physiological parameters contribute to in-
creased ventilation during exercise. These mechanisms have been reviewed by
Walsh and Banister (1988) and include CO
stimulation of the carotid bodies,
respiratory mechanics, temperature effects, and skeletal muscle neurogenic stimu-
lation. Consequently, the detected increase in ventilation cannot necessarily be
exclusively attributed to buffering of lactic acid. Studies of patients with McArdle’s
disease, a metabolic disorder in which affected individuals do not produce sub-
stantial amounts of lactic acid, have shown ventilatory breakpoints at higher exer-
cise intensities during incremental tests (Hagberg et al., 1982). In healthy indi-
viduals it has been shown that the lactate and ventilatory thresholds do not always
occur together, nor does the LT cause the VT (Neary et al., 1985). This leads one to
question the reliability of predicting anaerobic threshold from noninvasive gas
exchange measurements (Powers et al., 1984).
Anaerobic Threshold • 319
One advantage of using gas exchange measurements to predict anaerobic
threshold is that it is a noninvasive technique. Strong test-retest relationships in
work rate and V
max at VT have been reported (Yamamoto et al., 1991), and
reliability of the method is enhanced if test conditions and personnel are kept con-
stant. In terms of practicality, this method has clinical value, particularly when
maximal exercise is contraindicated and invasive blood sampling is not appropri-
ate or desired.
Yamamoto et al. (1991) reported that the VT measured during a stepwise
incremental test was equivalent to MLSS, and subjects were able to exercise for 30
minutes at constant intensities corresponding to the VT as well as at 4.9% above.
However, during the trial at 4.9% above VT, only 1 of 13 subjects exhibited a rise
in blood lactate concentration greater than 1 mM, while 3 subjects had a slight
decline in blood lactate concentration. Therefore, in the study by Yamamoto et al.,
it is likely that VT underestimated MLSS. This is another case of inappropriate use
of group statistics for validating such a technique.
It has been the purpose of this review to provide insight and clarification as to the
concept and methods of measurement and prediction of anaerobic threshold. The
anaerobic threshold, which is the highest intensity of exercise for which measure-
ment of oxygen uptake can account for the energy requirement of the exercise,
clearly does exist. Years of research have shown that it is a difficult concept to
define and measure. Yet despite the lack of theoretical and methodological con-
sensus, there undoubtedly is value in having a test to estimate the intensity of
exercise associated with the anaerobic threshold. The ideal test should consistently
yield an intensity of exercise that is close to maximal lactate steady state, which is
considered to be the best predictor of anaerobic threshold. It is important for exer-
cise science practitioners to be aware of the different definitions that are com-
monly applied to anaerobic threshold. Interpretation of the literature with different
definitions of these terms relies on taking the meaning of each term from the con-
text in which it is derived. Use of the appropriate operational term when referring
to the estimate of anaerobic threshold is encouraged.
Aunola, S., and Rusko, H. (1992). Does anaerobic threshold correlate with maximal lactate
steady state? J. Sports Sci. 10: 309-323.
Bassett, D.R., Jr., and Howley, E.T. (2000). Limiting factors for maximum oxygen uptake
and determinants of endurance performance. Med. Sci. Sports Exerc. 32: 70-84.
Beneke, R. (1995). Anaerobic threshold, individual anaerobic threshold, and maximal lac-
tate steady state in rowing. Med. Sci. Sports Exerc. 27: 863-867.
Beneke, R., and von Duvillard, S.P. (1996). Determination of maximal lactate steady state
response in selected sports events. Med. Sci. Sports Exerc. 28: 241-246.
Billat, L.V. (1996). Use of blood lactate measurements for prediction of exercise perfor-
mance and for control of training. Sports Med. 22: 157-175.
Billat, V.L., Morton, R.H., Blondel, N., Berthoin, S., Bocquet, V., Koralsztein, J.P., and
Barstow, T.J. (2000). Oxygen kinetics and modelling of time to exhaustion whilst
320 • Svedahl and MacIntosh
running at various velocities at maximal oxygen uptake. Eur. J. Appl. Physiol. 82:
Brooks, G.A. (2000). Intra- and extra-cellular lactate shuttles. Med. Sci. Sports Exerc. 32:
Brooks, G.A., Dubouchaud, H., Brown, M., Sicurello, J.P., and Butz, C.E. (1999). Role of
mitochondrial lactate dehydrogenase and lactate oxidation in the intracellular lactate
shuttle. Proc. Nat. Acad. Sci. 96: 1129-1134.
Carter, H., Jones, A.M., and Doust, J.H. (1999). Effect of incremental test protocol on the
lactate minimum speed. Med. Sci. Sports Exerc. 31: 837-845.
Chicharro, J.L., Hoyos, J., and Lucía, A. (2000). Effects of endurance training on the isocapnic
buffering and hypocapnic hyperventilation phases in professional cyclists. Br. J.
Sports Med. 34: 450-455.
Coen, B., Urhausen, A., and Kindermann, W. (2001). Individual anaerobic threshold: Meth-
odological aspects of its assessment in running. Int. J. Sports Med. 22: 8-16.
Connett, R.J., Gayeski, T.E., and Honig, C.R. (1984). Lactate accumulation in fully aero-
bic, working, dog gracilis muscle. Am. J. Physiol. 246: H120-H128.
Costill, D.L., Thomason, H., and Roberts, E. (1973). Fractional utilization of the aerobic
capacity during distance running. Med. Sci. Sports Exerc. 5: 248-252.
Coyle, E.F., Coggan, A.R., Hopper, M.K., and Walters, T.J. (1988). Determinants of endur-
ance in well-trained cyclists. J. Appl. Physiol. 64: 2622-2630.
Davis, J.A. (1985). Anaerobic threshold: Review of the concept and directions for future
research. Med. Sci. Sports Exerc. 17: 6-18.
Dickhuth, H.-H., Yin, L., Niess, A., Röcker, K., Mayer, F., Heitkamp, H.-C., and Horstmann,
T. (1999). Ventilatory, lactate-derived and catecholamine thresholds during incremental
treadmill running: Relationship and reproducibility. Int. J. Sports Med. 20: 122-127.
Donovan, C.M., and Pagliassotti, M.J. (2000). Quantitative assessment of pathways for
lactate disposal in skeletal muscle fiber types. Med. Sci. Sports Exerc. 32: 772-777.
Febbraio, M.A., Lambert, D.L., Starkie, R.L., Proietto, J., and Hargreaves, M. (1998). Ef-
fect of epinephrine on muscle glycogenolysis during exercise in trained men. J. Appl.
Physiol. 84: 465-470.
Foster, C., Crowe, M.P., Holum, D., Sandvig, S., Schrager, M., Snyder, A.C., and Zajakowski,
S. (1995). The bloodless lactate profile. Med. Sci. Sports Exerc. 27: 927-933.
Foxdal, P., Sjodin, A., and Sjodin, B. (1996). Comparison of blood lactate concentrations
obtained during incremental and constant intensity exercise. Int. J. Sports Med. 17:
Gaesser, G.A., and Poole, D.C. (1996). The slow component of oxygen uptake kinetics in
humans. In: J.O. Holloszy (Ed.), Exerc. Sport Sci. Rev., pp. 35-70.
Gladden, L.B. (1996). Lactate transport and exchange during exercise. Section 12. Exer-
cise: Regulation and Integration of Multiple Systems. In: L.B. Rowell and J.T. Shep-
herd (Eds.), Handbook of Physiology, pp. 614-648. Bethesda, MD: Oxford Univer-
Gladden, L.B. (2000). Muscle as a consumer of lactate. Med. Sci. Sports Exerc. 32: 764-
Graham, T.E. (1991). A review of some issues associated with lactate metabolism during
exercise. In: N. Bachl, T.E. Graham, and H. Löllgen (Eds.), Advances in Ergometry,
pp. 125-148. NewYork: Springer-Verlag.
Graham, T.E., and Saltin, B. (1989). Estimation of the mitochondrial redox state in human
skeletal muscle during exercise. J. Appl. Physiol. 66: 561-566.
Anaerobic Threshold • 321
Green, H.J., Sutton, J.R., Coates, G., Ali, M., and Jones, S. (1991). Response of red cell and
plasma volume to prolonged training in humans. J. Appl. Physiol. 70: 1810-1815.
Gullstrand, L., Sjödin, B., and Svedenhag, J. (1994). Blood sampling during continuous
running and 30-second intervals on a treadmill. Scand. J. Med. Sci. Sports 4: 239-
Hagberg, J.M., and Coyle, E.F. (1983). Physiological determinants of endurance perfor-
mance as studied in competitive racewalkers. Med. Sci. Sports Exerc. 15: 287-289.
Hagberg, J.M., Coyle, E.F., Carroll, J.F., Miller, J.M., Martin, W.H., and Brooke, M.H.
(1982). Exercise hyperventilation in patients with McArdle’s disease. J. Appl. Physiol.
Haverty, M., Kenney, W.L., and Hodgson, J.L. (1988). Lactate and gas exchange responses
to incremental and steady state running. Br. J. Sports Med. 22: 51-54.
Heck, H., Mader, A., Hess, G., Mucke, S., Muller, R., and Hollmann, W. (1985). Justifica-
tion of the 4-mmol/l lactate threshold. Int. J. Sports Med. 6: 117-130.
Henritze, J., Weltman, A., Schurrer, R.L., and Barlow, K. (1985). Effects of training at and
above the lactate threshold on the lactate threshold and maximal oxygen uptake.
Eur. J. Appl. Physiol. 54: 84-88.
Hill, A.V., and Lupton, H. (1923). Muscular exercise and the supply and utilization of oxy-
gen. Q. J. Med. 16: 135-171.
Hogan, M.C. (2001). Fall in intracellular PO
at the onset of contractions in Xenopus single
skeletal muscle fibers. J. Appl. Physiol. 90: 1871-1876.
Hollmann, W. (1991). The anaerobic threshold as a tool in medicine. In: N. Bachl, T.E.
Graham, and H. Löllgen (Eds.), Advances in Ergometry, pp. 1-11. NewYork: Springer-
Holloszy, J.O., and Coyle, E.F. (1984). Adaptations of skeletal muscle to endurance exer-
cise and their metabolic consequences. J. Appl. Physiol. 56: 831-838.
Hoogeveen, A.R., Hoogsteen, J., and Schep, G. (1997). The maximal lactate steady state in
elite endurance athletes. Jpn. J. Physiol. 47: 481-485.
Hughson, R.L., and Green, H.J. (1982). Blood acid-base and lactate relationships studied
by ramp work tests. Med. Sci. Sports Exerc. 14: 297-302.
Jacobs, I., and Kaiser, P. (1982). Lactate in blood, mixed skeletal muscle, and FT or ST
fibres during cycle exercise in man. Acta Physiol. Scand. 114: 461-466.
Jenkins, D.G., and Quigley, B.M. (1990). Blood lactate in trained cyclists during cycle
ergometry at critical power. Eur. J. Appl. Physiol. 61: 278-283.
Jobsis, F.F., and Stainsby, W.N. (1968). Oxidation of NADH during contractions of circu-
lated mammalian skeletal muscle. Respir. Physiol. 4: 292-300.
Jones, A.M., Carter, H., and Doust, J.H. (1999). A disproportionate increase in V
dent with lactate threshold during treadmill exercise. Med. Sci. Sports Exerc. 31:
Jones, A.M., and Doust, J.H. (1998). The validity of the lactate minimum test for determi-
nation of the maximal lactate steady state. Med. Sci. Sports Exerc. 30: 1304-1313.
Jorfeldt, L., Juhlin-Dannfelt, A., and Karlsson, J. (1978). Lactate release in relation to tissue
lactate in human skeletal muscle during exercise. J. Appl. Physiol. 44: 350-352.
Karlsson, J., and Jacobs, I. (1982). Onset of blood lactate accumulation during muscular
exercise as a threshold concept. Int. J. Sports Med. 3: 190-201.
Kindermann, W., Simon, G., and Keul, J. (1979). The significance of the aerobic–anaerobic
transition for the determination of work load intensities during endurance training.
Eur. J. Appl. Physiol. 42: 25-34.
322 • Svedahl and MacIntosh
Lajoie, C., Laurencelle, L., and Trudeau, F. (2000). Physiological responses to cycling for
60 minutes at maximal lactate steady state. Can. J. Appl. Physiol. 25: 250-261.
Londeree, B.R., and Ames, S.A. (1975). Maximal steady state versus state of conditioning.
Eur. J. Appl. Physiol. 34: 269-278.
MacIntosh, B.R., Esau, S., and Svedahl, K. (2002). The lactate minimum test for cycling:
Estimation of the maximal lactate steady state. Can. J. Appl. Physiol. 27: 232-249.
MacIntosh, B.R., Neptune, R.R., and Van den Bogert, A.J. (2000). Intensity of cycling and
cycle ergometry: Power output and energy cost. In: B.M. Nigg, B.R. MacIntosh, and
J. Mester (Eds.), Biomechanics and Biology of Movement, pp. 129-148. Champaign,
IL: Human Kinetics.
Mazzeo, R.S., and Marshall, P. (1989). Influence of plasma catecholamines on the lactate
threshold during graded exercise. J. Appl. Physiol. 67: 1319-1322.
McLellan, T.M. (1985). Ventilatory and plasma lactate response with different exercise pro-
tocols: A comparison of methods. Int. J. Sports Med. 6: 30-35.
McLellan, T.M., Cheung, K.S.Y., and Jacobs, I. (1991). Incremental test protocol, recovery
mode and the individual anaerobic threshold. Int. J. Sports Med. 12: 190-195.
McLellan, T.M., and Jacobs, I. (1993). Reliability, reproducibility and validity of the indi-
vidual anaerobic threshold. Eur. J. Appl. Physiol. 67: 125-131.
McMorris, T., Sproule, J., Draper, S., Child, R., Sexsmith, J.R., Forster, C.D., and Pattison,
J. (2000). The measurement of plasma catecholamine and lactate thresholds: A com-
parison of methods. Eur. J. Appl. Physiol. 82: 262-267.
Murase, Y., Kobayashi, K., Kamei, S., and Matsui, H. (1981). Longitudinal study of aerobic
power in superior junior athletes. Med. Sci. Sports Exerc. 13: 180-184.
Neary, P.J., MacDougall, J.D., Bachus, R., and Wenger, H.A. (1985). The relationship be-
tween lactate and ventilatory thresholds: Coincidental or cause and effect? Eur. J.
Appl. Physiol. 54: 104-108.
Needham, D.M. (1971). Machina Carnis. The Biochemistry of Muscular Contraction
in its Historical Development. Cambridge: Cambridge University Press.
Nicholson, R.M., and Sleivert, G.G. (2001). Indices of lactate threshold and their relation-
ship with 10-km running velocity. Med. Sci. Sports Exerc. 33: 339-342.
Owles, W.H. (1930). Alterations in the lactic acid content of the blood as a result of light
exercise, and associated changes in the CO
-combining power of the blood and in
the alveolar CO
pressure. J. Physiol. (Lond.) 69: 214-237.
Powers, S.K., Dodd, S., and Garner, R. (1984). Precision of ventilatory and gas exchange
alterations as a predictor of the anaerobic threshold. Eur. J. Appl. Physiol. 52: 173-177.
Pyne, D.B., Lee, H., and Swanwick, K.M. (2001). Monitoring the lactate threshold in world-
ranked swimmers. Med. Sci. Sports Exerc. 33: 291-297.
Rall, T.W., Sutherland, E.W., and Berthet, J. (1957). The relationship of epinephrine and
glucagon to liver phosphorylase. IV. Effect of epinephrine and glucagon on the reac-
tion of phosphorylase in liver homogenates. J. Biol. Chem. 224: 463-475.
Richardson, R.S., Noyszewski, E.A., Leigh, J.S., and Wagner, P.D. (1998). Lactate efflux from
exercising human skeletal muscle: Role of intracellular PO
. J. Appl. Physiol. 85: 627-634.
Richter, E.A., Ruderman, N.B., Gavras, H., Belur, E.R., and Galbo, H. (1982). Muscle
glycogenolysis during exercise: Dual control by epinephrine. Am. J. Physiol.
Endocrinol. Metab. 242: E25-E32.
Robergs, R.A., Chwalbinski-Moneta, J., Mitchell, J.B., Pascoe, D.D., Houmard, J., and
Costill, D.L. (1990). Blood lactate threshold differences between arterialized and
venous blood. Int. J. Sports Med. 11: 446-451.
Anaerobic Threshold • 323
Rusko, H.K. (1992). Development of aerobic power in relation to age and training in cross-
country skiers. Med. Sci. Sports Exerc. 24: 1040-1047.
Schneider, D.A., McLellan, T.M., and Gass, G.C. (2000). Plasma catecholamine and blood
lactate responses to incremental arm and leg exercise. Med. Sci. Sports Exerc. 32:
Sjodin, B., Jacobs, I., and Karlsson, J. (1981). Onset of blood lactate accumulation and
marathon running performance. Int. J. Sports Med. 2: 23-26.
Sjodin, B., Jacobs, I., and Svedenhag, J. (1982). Changes in onset of blood lactate accumu-
lation (OBLA) and muscle enzymes after training at OBLA. Eur. J. Appl. Physiol.
Stainsby, W.N. (1986). Biochemical and physiological bases for lactate production. Med.
Sci. Sports Exerc. 18: 341-343.
Stainsby, W.N., Brechue, W.F., and O’Drobinak, D.M. (1991). Regulation of muscle lactate
production. Med. Sci. Sports Exerc. 23: 907-911.
Stainsby, W.N., Sumners, C., and Andrew, G.M. (1984). Plasma catecholamines and their
effect on blood lactate and muscle lactate output. J. Appl. Physiol. 57: 321-325.
Stegmann, H., and Kindermann, W. (1982). Comparison of prolonged exercise tests at the
individual anaerobic threshold and the fixed anaerobic threshold of 4 mmol·L
tate. Int. J. Sports Med. 3: 105-110.
Stegmann, H., Kindermann, W., and Schnabel, A. (1981). Lactate kinetics and individual
anaerobic threshold. Int. J. Sports Med. 2: 160-165.
Swensen, T.C., Harnish, C.R., Beitman, L., and Keller, B.A. (1999). Noninvasive estima-
tion of the maximal lactate steady state in trained cyclists. Med. Sci. Sports Exerc.
Tegtbur, U., Busse, M.W., and Braumann, K.M. (1993). Estimation of an individual equi-
librium between lactate production and catabolism during exercise. Med. Sci. Sports
Exerc. 25: 620-627.
Urhausen, A., Coen, B., Weiler, B., and Kindermann, W. (1993). Individual anaerobic thresh-
old and maximum lactate steady state. Int. J. Sports Med. 14: 134-139.
Walsh, M.L., and Banister, E.W. (1988). Possible mechanisms of the anerobic threshold.
Sports Med. 5: 269-302.
Wasserman, K. (1987). Determinants and detection of anaerobic threshold and consequences
of exercise above it. Circulation 76(Suppl.VI): V1-V29.
Wasserman, K., Hansen, J.E., Sue, S.Y., and Whipp, B.J. (1999). Principles of Exercise
Testing & Interpretation: Including Pathophysiology and Clinical Applications
(3rd ed.). Philadelphia: Lippincott Williams & Wilkins.
Wasserman, K., and McIlroy, M.B. (1964). Detecting the threshold of anaerobic metabo-
lism in cardiac patients during exercise. Am. J. Cardiol. 14: 844-852.
Wasserman, K., Whipp, B.J., Koyal, S.N., and Beaver, W.L. (1973). Anaerobic threshold
and respiratory gas exchange during exercise. J. Appl. Physiol. 35: 236-243.
Yamamoto, Y., Miyashita, M., Hughson, R.L., Tamura, S., Shinohara, M., and Mutoh, Y.
(1991). The ventilatory threshold gives maximal lactate steady state. Eur. J. Appl.
Physiol. 63: 55-59.
Yeh, M.P., Gardner, R.M., Adams, T.D., Yanowitz, F.G., and Crapo, R.O. (1983). “Anaerobic
threshold”: Problems of determination and validation. J. Appl. Physiol. 55: 1178-1186.
Received April 1, 2002; accepted April 18, 2002.