Development
Concerns and approaches: from design to interpretation
Due to the exciting prospects for the clinical relevancy of PEH and the need for more
research on this phenomenon, we were compelled to explore issues related to experimental
design, methodology, and data collection, analysis and interpretation of the PEH studies
that should be carefully considered to promote scientific rigor and high reproducibility
on this topic. As the aim of this review was to stimulate a comprehensive discussion
about methodological concerns in PEH studies, a narrative model was more appropriate,
and was therefore selected [11 ]. For this narrative review, the literature search was conducted in the PubMed database
using the term “post-exercise hypotension,” as well as in reference lists of systematic
reviews about this topic and reference lists of the PEH studies. Moreover, this study
followed all ethical standards [12 ].
Concerns with blood pressure measurements
An important aspect regarding any study of blood pressure measures is the fact that
subjects should be sufficiently familiarized with the lab environment and blood pressure
measurement procedures before the beginning of the experimental protocol.
Given the variability in blood pressure [13 ]
[14 ]
[15 ], clinical guidelines for assessing hypertension recommend the measurement of blood
pressure on two or more visits to determine resting blood pressure, and during each
visit blood pressure should be measured at least twice [16 ]
[17 ]
[18 ]. This is based on studies comparing blood pressure changes between successive visits
when blood pressure decreases substantially from the first to the second visit [19 ]
[20 ]
[21 ], regardless of the interval between the visits (3 days or 14 days) or the initial
level of blood pressure (normotensives or hypertensives) [21 ]. Values obtained on a third visit were similar to those observed on a second visit
[21 ], highlighting the importance of screening visits that include the measurement of
blood pressure to familiarize subjects. So that, a stable resting blood pressure can
be recorded in the experimental protocol, allowing for greater rigor and reproducibility
in the measure of PEH.Despite the importance of familiarization visits to stabilize
resting blood pressure in PEH investigations, some studies have relied on only one
screening visit [9 ]
[22 ]
[23 ]
[24 ]
[25 ]
[26 ]
[27 ]
[28 ]
[29 ]
[30 ]
[31 ]
[32 ]
[33 ]
[34 ]
[35 ]
[36 ]
[37 ]
[38 ]
[39 ]
[40 ]
[41 ]
[42 ]
[43 ]
[44 ]
[45 ]
[46 ]
[47 ] or none at all [48 ]
[49 ]
[50 ]
[51 ]
[52 ], with blood pressure measurements before the experiments, and many studies simply
have not reported whether any blood pressure screening or familiarization was performed
[53 ]
[54 ]
[55 ]
[56 ]
[57 ]
[58 ]
[59 ]
[60 ]
[61 ]
[62 ]
[63 ]
[64 ]
[65 ]
[66 ]
[67 ]
[68 ]
[69 ] ([Table 1 ]). While study designs that incorporate both exercise and control session may mitigate
this concern by showing no difference in pre-exercise blood pressure between the sessions,
studies with a simple pre- versus post-exercise design (i. e., without a control session
and/or one comparison session of exercise) are highly susceptible to this limitation
[24 ]
[29 ]
[47 ]
[49 ].
Table 1 Concerns and approaches in post-exercise hypotension studies.
Author and Year
Days of Screening
Time of Blood Pressure Measurements Post-exercise
Control or Comparative Session yes (y)/no (n)
Approach (I - Pre vs. Post)/(II - E vs. C)/(III - NE)
Angadi (2015)[22 ]
1
3hs/every 15 min
Y
II
Aprile (2016)[98 ]
2
1 h
Y
II
Augeri (2011)[23 ]
1
ABPM
Y
III
Azevêdo (2017)[53 ]
NR
1 h/every 15 min
Y
III
Birk (2012)[54 ]
NR
10 min
Y
I
Bisquolo (2005)[99 ]
NR
45 min
Y
I, II
Bonsu (2016)[24 ]
1
1 h/every 10 min
N
I
Brandão (2002)[82 ]
2
15;30;60;90 min/ ABPM
Y
I, II
Brito (2014)[55 ]
NR
10;30;50;70;90 min
Y
II
Brito (2015)[107 ]
2
45 min
Y
I, II, III
Brito (2017)[89 ]
2
ABPM
Y
II
Carter III (2001)[56 ]
NR
5 min
Y
I, II
Cavalcante (2017)[25 ]
1
1 h/every 15 min
Y
I, II, III
Ciolac (2009)[57 ]
NR
ABPM
Y
II
Cleroux (1992)[84 ]
3
30;60;90 min
Y
II
Coats (1989)[58 ]
NR
1 h/every 10 min
Y
III
Collier (2010)[48 ]
0
40;60 min
Y
I, III
Costa (2016)[73 ]
2
1 h/every 10 min
Y
I
Cucato (2015)[106 ]
2
45 min
Y
I, III
Cunha (2016)[26 ]
1
20;40;60 min
Y
II
Dantas (2016)[27 ]
1
ABPM
Y
II
Dawson (2008)[59 ]
NR
60 min
N
I
Dos Santos (2014)[119 ]
1
1 h/every 15 min
Y
II
Dujic (2006)[49 ]
0
30;60 min
N
I
Endo (2012)a[29 ]
1
1 h/every 15 min
N
I
Endo (2012)b[28 ]
1
1 h/every 15 min
Y
I, II
Esformes (2006)[60 ]
1
45 min
Y
I, II
Figueiredo (2015)[50 ]
2
1 h/every 10 min
Y
I, II
Floras (1989)[85 ]
3
1 h/every 5 min
Y
I
Floras (1992)[93 ]
3
1 h/every 5 min
Y
I
Forjaz (1999)[86 ]
2
90 min/every 5 min
Y
I
Forjaz (2000)[8 ]
2
ABPM
Y
II
Forjaz (2004)[81 ]
2
15;30;60;90 min ABPM
Y
II, III
Franklin (1993)[51 ]
0
1 h/every 15 min
Y
I, II
Gagnon (2012)[30 ]
1
10;30;50;70;90 min
Y
I, II, III
Goessler (2015)[31 ]
1
ABPM
Y
I, II
Hagberg (1987)[83 ]
4
1 h/every 10 min
Y
I
Halliwill (1996)[33 ]
1
60 min
Y
II
Halliwill (2000)[32 ]
1
1 h
Y
II
Hamer (2006)[61 ]
NR
30 min
Y
II
Harvey (2005)[34 ]
1
45;90 min
Y
I
Headley (2008)[76 ]
2
1 h/every 10 min/ABPM
Y
III
Headley (2017)[100 ]
1
1 h/every 10 min/ABPM
Y
I, II, III
Hecksteden (2013)[71 ]
2
10 min/1 h/24 h
N
I
Heffernan (2007)[35 ]
1
20 min
Y
I, II
Isea (1994)[94 ]
NR
30 min;1 h;2 h;3 h;4 h
Y
II
Jones (2007)[36 ]
1
20 min
Y
III
Jones (2008)a[75 ]
1
5;10;15;20
Y
III
Jones (2008)b[74 ]
1
5;10;15;20
Y
III
Keese (2012)[62 ]
NR
2 h/every 10 min
Y
I, II
Lacombe (2011)[37 ]
1
60 min
Y
I, III
Legramante (2002)[108 ]
3
60;90 min
N
I
Lehmkuhl (2005)[38 ]
1
ABPM
Y
II
Liu (2012)[39 ]
1
30 min
Y
I, II
Lockwood (2005)[40 ]
1
30;60;90 min
Y
II
MacDonald (1999)[78 ]
5
1 h/every 15 min
Y
I, II
Mac Donald (2000)a[77 ]
5
1 h/every 15 min
Y
I, II
MacDonald (2000)b[80 ]
5
1 h/every 15 min
Y
I, II
Mach (2005)[52 ]
0
30;60;90 min
Y
II
McCord (2006)[101 ]
1
30;60;90 min
Y
III
Moreira (2014)[72 ]
NR
60 min
N
I
New (2013)[63 ]
NR
30;60;90 min
Y
I, II
Notarius (2005)[64 ]
NR
10 min
Y
I, II
Pricher (2004)[42 ]
1
2 h/every 20 min
N
I
Queiroz (2009)[87 ]
2
1 h/every 5 min ABPM
Y
I, II, III
Queiroz (2013)a[109 ]
2
60 min ABPM
Y
I, II
Queiroz (2013)b[110 ]
2
60 min
Y
I, II, III
Queiroz (2015)[90 ]
2
45 min/ABPM
Y
I, II, III
Queiroz (2017)[111 ]
2
ABPM
Y
I, II
Raine (2001)[43 ]
1
1 h/every 10 min
Y
III
Rezk (2006)[79 ]
NR
15;30;60;90 min
Y
III
Rossow (2010)[44 ]
1
30;60 min
Y
I, II
Santaella (2006)[45 ]
1
30;60 min
Y
III
Santana (2013)[46 ]
1
60 min
Y
I, III
Somers (1991)[65 ]
NR
ABPM
Y
I, II
Souza (2016)[66 ]
NR
2 h/every 10 min
Y
III
Taylor-Tolbert (2000)[7 ]
4
ABPM
Y
II
Takahashi (2000)[68 ]
NR
10 min
Y
I, II
Takahashi (2005)[67 ]
NR
5 min
Y
I, II
Teixeira (2011)[88 ]
2
30;60;90;120 min
Y
III
Terblanche (2012)[112 ]
2
ABPM
Y
I, II, III
Tibana (2015)[120 ]
NR
10;30;60 min
Y
II, III
Wallace (1997)[9 ]
1
ABPM
Y
II
Wilcox (1982)[69 ]
NR
30 min
Y
I
Wilkins (2004)[47 ]
1
2hs/every 20 min
N
I
AMBP – Ambulatory blood pressure; NR – non-reported; Pre vs. Post – Comparing pre
values with post-exercise values; E vs. C – Comparing exercise with control sessions;
NE – Evaluating net effect.
Since post-exercise blood pressure reduction is influenced by its pre-exercise value
[5 ]
[70 ], an “unreal” elevated pre-exercise value, based on a lack of adequate familiarization,
may inflate PEH. As a result, future studies where PEH is the main outcome should
include, at least two familiarization visits during which blood pressure is measured
in accordance with reliability and familiarization studies [13 ]
[14 ]
[15 ]. Researchers may then have more confidence in resting blood pressure obtained during
subsequent experimental sessions.
Timing of post-exercise measurements
The timing of blood pressure measurements after exercise may be one of the main causes
for divergent results, impacting the magnitude and determinants of PEH across studies.
The variability of experimental designs ranges from evaluating PEH during the first
5 min [67 ] to first 3 hrs [22 ] post-exercise, and extends to 24-h ambulatory blood pressure (ABPM) measurements
after exercise. In fact, this variability has made it difficult to determine at what
time after exercise the greatest PEH occurs and has introduced debate as to how long
it lasts.
The magnitude of clinical PEH is important due to its strong correlation with the
chronic hypotensive effect of exercise [39 ]
[71 ]
[72 ], allowing for the detection of responders and non-responders to training. Among
37 studies that analyzed blood pressure in the laboratory setting for at least 20 min
after exercise [22 ]
[24 ]
[25 ]
[26 ]
[28 ]
[29 ]
[35 ]
[36 ]
[42 ]
[43 ]
[45 ]
[47 ]
[50 ]
[51 ]
[52 ]
[53 ]
[54 ]
[55 ]
[56 ]
[62 ]
[64 ]
[65 ]
[66 ]
[67 ]
[68 ]
[71 ]
[73 ]
[74 ]
[75 ]
[76 ]
[77 ]
[78 ]
[79 ]
[80 ]
[81 ]
[82 ]
[83 ], 32 studies reported no decrease in blood pressure until that time (i. e., 20 min)
[25 ]
[26 ]
[29 ]
[35 ]
[36 ]
[42 ]
[43 ]
[44 ]
[47 ]
[50 ]
[51 ]
[52 ]
[53 ]
[54 ]
[55 ]
[56 ]
[58 ]
[62 ]
[64 ]
[65 ]
[66 ]
[67 ]
[68 ]
[71 ]
[73 ]
[74 ]
[75 ]
[76 ]
[77 ]
[79 ]
[80 ]
[81 ], and the other 5 reported PEH was only at 15 min post-exercise [22 ]
[28 ]
[78 ]
[82 ]
[83 ]. Among 31 studies that followed blood pressure throughout the recovery period for
at least 60 min [22 ]
[24 ]
[25 ]
[26 ]
[28 ]
[29 ]
[30 ]
[42 ]
[43 ]
[47 ]
[50 ]
[51 ]
[53 ]
[55 ]
[58 ]
[62 ]
[66 ]
[73 ]
[76 ]
[77 ]
[78 ]
[79 ]
[80 ]
[81 ]
[82 ]
[83 ]
[84 ]
[85 ]
[86 ]
[87 ]
[88 ], 19 studies reported the greatest blood pressure decrease after 30 min of recovery
period [22 ]
[24 ]
[28 ]
[29 ]
[42 ]
[47 ]
[50 ]
[53 ]
[58 ]
[62 ]
[76 ]
[77 ]
[78 ]
[80 ]
[81 ]
[83 ]
[84 ]
[87 ]
[88 ], while 9 studies reported the same magnitude of PEH at all recovery times [43 ]
[51 ]
[55 ]
[66 ]
[73 ]
[79 ]
[82 ]
[85 ]
[86 ], and 3 studies reported the greatest PEH between 15–20 min after exercise [25 ]
[26 ]
[30 ]. Finally, all of the 8 studies that measured blood pressure for more than 1 h in
the laboratory setting after exercise reported PEH lasting at least 120 min [22 ]
[40 ]
[41 ]
[42 ]
[62 ]
[83 ]
[84 ]
[88 ]. Taking into account the aforementioned information, future studies aiming to evaluate
PEH should assess blood pressure between 20 and 120 min after the exercise to increase
the chance for identifying PEH.
The duration of PEH is important to reveal whether the hypotensive effect can have
a clinical impact reducing the subject’s cardiovascular overload and risk for a long
period of time [1 ]. Ambulatory blood pressure monitoring is the only adequate measure to evaluate this
benefit. However, interpretations of ambulatory blood pressure recordings after exercise
are sometimes confounded by the use of prolonged averaging periods (e. g., 12- or
24-h mean) [8 ]
[9 ]
[38 ]
[57 ]
[82 ]
[89 ]
[90 ]. While studies that report a reduction in mean 12- or 24-h blood pressure provide
useful information related to the health benefits of exercise, this approach does
not lend itself to a determination of the duration of PEH, which has erroneously been
reported as lasting 24 h. The problem is that values early in a timeframe can generate
the appearance of a blood pressure reduction, which may be resolved by the end of
the averaging timeframe. As a result, authors should opt for reporting hour-to-hour
analysis, as demonstrated by Pescatello et al. [91 ], where mean blood pressure stayed lower than pre-exercise values for 13 h.
It is also important to note that post-exercise ambulatory blood pressure analysis
allows for investigating the effects of exercise in real life conditions. However,
some care is needed for comparing days with similar daily activities, such as conducting
the experiment on the same day of the week (e. g., exercise and control sessions conducted
on Monday) or at least to avoid conducting experiments on days with very different
activities (e. g., a business day and a weekend day). Another important aspect is
to start ambulatory blood pressure recordings at the same time of day in all experimental
sessions, since average ambulatory blood pressure values are higher when monitoring
was started in the morning than in the afternoon [92 ].
Body position for blood pressure measurements
The body position of subjects before the exercise and during the recovery period after
the exercise is an important experimental consideration for PEH studies and impacts
the interpretation of the results, especially when considering hemodynamic mechanisms.
An extensive review of the literature suggests that studying subjects in the supine
position and in the seated position are both common methods [2 ], but some studies have not reported the adopted body position [93 ]
[94 ]. In addition, no study has, to our knowledge, investigated PEH in the upright position.
This inconsistency across studies limits comparisons. However, for some purposes,
each position may be justified.
The supine position favors hemodynamic measurements without the interference of orthostatic
stress [95 ]. However, it may decrease the generalizability of some findings, as outside of the
construct of research studies, it is less common for individuals to recover from exercise
in the supine position. In contrast, blood pressure measurements performed with subjects
in a seated position may be more relevant, since many persons sit after exercise to
rest, talk or do any other task. Actually, due to the orthostatic stress promoted
by seated position, diastolic blood pressure increases over time when this position
is sustained [95 ]. This is likely to be secondary to decreased venous return, which deactivates the
cardiopulmonary reflex, increasing peripheral sympathetic nerve activity, systemic
vascular resistance and diastolic blood pressure [96 ]. In the laboratory setting, this influence of body position is sufficient to flip
the hemodynamics which underlies PEH, from a decrease of systemic vascular resistance
when recovery occurs in the supine position to a decrease of cardiac output when recovery
occurs in the seated position [43 ]
[67 ]
[68 ].
Therefore, there are valid rationales for studying subjects in either the supine or
the seated position in PEH studies, depending on the experiment’s objectives. However,
authors should report which body position was used in the studies and should use the
same body position pre- and post-exercise in both control and exercise sessions. Likewise,
authors should discuss results within the context of the chosen body position. These
steps may help avoid misinterpretations.
Method of blood pressure measurement
Due to the sometimes-small magnitude of relevant blood pressure reductions associated
with recovery from exercise, rigorous methods of measuring blood pressure are essential
in the study of PEH. While auscultatory measurement is the most common method for
assessing blood pressure [97 ], manual sphygmomanometry, with its susceptibility to observer bias, two-digit preference
and environmental noise, may not be so appropriate, if it was not made by an experienced
evaluator, which minimizes but does not exclude observer bias. As a result, many researchers
of PEH use automated blood pressure monitoring [22 ]
[25 ]
[26 ]
[28 ]
[29 ]
[32 ]
[34 ]
[40 ]
[42 ]
[44 ]
[46 ]
[47 ]
[49 ]
[50 ]
[51 ]
[53 ]
[54 ]
[59 ]
[62 ]
[65 ]
[71 ]
[73 ]
[76 ]
[86 ]
[87 ]
[94 ]
[98 ]
[99 ]
[100 ]
[101 ].
While various groups have established criteria that such devices must fulfill to be
validated (e. g., Association for the Advancement of Medical Instrumentation, British
Hypertension Society, European Society of Hypertension), the criteria are focused
on clinical testing and do not generally require testing specific to exercise or recovery
from exercise. Thus, while there are validations of automated oscillometric devices
at rest [16 ]
[18 ], to the best of our knowledge, no study has tried to validate oscillometric devices
for use during the recovery period after exercise. Nevertheless, oscillometric devices
actually measure mean blood pressure and estimate systolic and diastolic blood pressure
through algorithms based on the range of pulse pressure [97 ]. Thus when pulse pressure changes, the algorithm is affected [102 ]. As blood pressure reductions after exercise are usually greater for systolic than
diastolic blood pressure [5 ]
[6 ]
[70 ], pulse pressure decreases during exercise recovery, and this change may introduce
bias into oscillometric-estimated systolic and diastolic blood pressures. Indeed,
when assessed at the same time, auscultatory blood pressure showed a decrease in systolic
and diastolic blood pressures after exercise, which was not detected by automated
oscillometry that only reveals a reduction of mean blood pressure [87 ]. However, oscillometric method minimizes observer bias [102 ] and is less impacted by environmental noise than manual auscultation. For this reason,
when employing oscillometric devices, mean blood pressure should be considered.
In contrast, at least one automated auscultatory device has been validated for exercise
[103 ] and such device is gaining use in PEH studies [40 ]
[42 ]
[47 ]
[94 ]
[101 ]. Finger monitors that rely on photoplethysmography (e. g., Finometer) have been
validated for the study of PEH [78 ], and this has encouraged other research groups to use this technique [30 ]
[33 ]
[35 ]
[36 ]
[37 ]
[43 ]
[48 ]
[55 ]
[56 ]
[61 ]
[67 ]
[68 ]
[74 ]
[75 ]
[77 ]
[78 ]
[80 ]. However, while these devices can accurately track beat-to-beat changes in blood
pressure, they have limitations for determining absolute values for blood pressure
[104 ].
In brief, automated oscillometric measurements have advantages, mainly in mitigating
investigator bias, but their validity is not as well established as automated auscultatory
measurements for PEH studies. Considering that auscultatory is also widely used for
measurement in PEH studies, its limitation can be minimized by appropriate training
and blinding of the observer.
Different approaches to quantifying PEH
Evaluating the occurrence of PEH and quantifying its magnitude requires a comparison
of post-exercise blood pressure to some reference, and a variety of blood pressure
measurements have been used as this reference value across studies (e. g., measurements
taken immediately before the exercise bout, or from a control session without exercise).
In addition, the magnitude of PEH has been calculated with different methods, with
some incorporating changes over time determined from a control session [105 ], an example can be seen in [Fig. 1 ].
Fig. 1 Demonstrative of post-exercise hypotension (PEH) evaluated by different methods.
Systolic blood pressure responses to exercise (continuous line) and control (dotted
line) session. (Panel a) - PEH_I post-exercise blood pressure - pre-exercise blood
pressure; (Panel b) - PEH_II: post-exercise blood pressure - post-control blood pressure;
(Panel c) - PEH_III: [(post-exercise blood pressure – pre-exercise blood pressure)
– (post-control blood pressure – pre-control blood pressure)].
In many studies [24 ]
[25 ]
[28 ]
[29 ]
[30 ]
[31 ]
[34 ]
[37 ]
[39 ]
[42 ]
[44 ]
[46 ]
[47 ]
[48 ]
[49 ]
[51 ]
[54 ]
[56 ]
[59 ]
[62 ]
[63 ]
[64 ]
[65 ]
[67 ]
[68 ]
[69 ]
[71 ]
[73 ]
[77 ]
[78 ]
[80 ]
[82 ]
[83 ]
[85 ]
[86 ]
[87 ]
[90 ]
[93 ]
[99 ]
[100 ]
[106 ]
[107 ]
[108 ]
[109 ]
[110 ]
[111 ]
[112 ] (shown in [Table 1 ]), PEH has been calculated as simply the difference between post- and pre-exercise
blood pressures, which we will define as “PEH_I” (PEH_I = post-exercise blood pressure
– pre-exercise blood pressure). Other studies which have included both an exercise
and a control session, and have calculated PEH as the difference between post-exercise
and post-control blood pressures, what we define as “PEH_II” (PEH_II=post-exercise
blood pressure – post-control blood pressure) [7 ]
[8 ]
[9 ]
[22 ]
[25 ]
[26 ]
[27 ]
[28 ]
[30 ]
[31 ]
[32 ]
[33 ]
[35 ]
[38 ]
[39 ]
[40 ]
[44 ]
[50 ]
[51 ]
[52 ]
[55 ]
[56 ]
[57 ]
[60 ]
[61 ]
[62 ]
[63 ]
[64 ]
[65 ]
[67 ]
[68 ]
[77 ]
[78 ]
[80 ]
[82 ]
[84 ]
[87 ]
[89 ]
[90 ]
[94 ]
[98 ]
[99 ]
[100 ]
[107 ]
[109 ]
[110 ]
[111 ]
[112 ] ([Table 1 ]). Finally, some studies [23 ]
[25 ]
[30 ]
[36 ]
[37 ]
[43 ]
[45 ]
[46 ]
[48 ]
[53 ]
[58 ]
[66 ]
[74 ]
[75 ]
[76 ]
[79 ]
[81 ]
[87 ]
[88 ]
[90 ]
[100 ]
[101 ]
[106 ]
[107 ]
[110 ]
[112 ] ([Table 1 ]) have taken a more complex approach, calculating PEH as the net effect, i. e., the
difference between responses observed in an exercise session and a control session,
which we define as “PEH_III” [PEH_III=(post-exercise blood pressure – pre-exercise
blood pressure) – (post-control blood pressure – pre-control blood pressure)].
Thus, researchers should consider the advantages and disadvantages for each approach
to data analysis when they are designing their study. PEH_I has the advantage of demanding
only one experimental session, with as few as two blood pressure measurements (pre-
vs. post-exercise); however, we would strongly suggest the inclusion of more measurements
over time rather than the bare minimum of two. In contrast, both PEH_II and PEH_III
demand two data collection sessions, performed on different occasions (exercise vs.
control). Per session, there is a minimum of one (post-exercise and post-control for
PEH_II) or two (pre- and post-exercise/post-control for PEH_III) blood pressure measurements.
Again, more measurements across time are desirable.
As studies using PEH_I require less time to execute and fewer data collection sessions,
it is not surprising that this approach has been employed in many studies. However,
PEH_I does not take into account possible changes in blood pressure that can happen
across time independently of exercise (e. g., circadian effects), which can only be
controlled by the inclusion of a control session in the study design, as done for
PEH_II and PEH_III analysis. The inclusion of a control session as part of the experimental
design and analysis in PEH mitigates against any overlying time effect. For example,
some studies [8 ]
[81 ]
[90 ]
[106 ]
[107 ]
[109 ]
[110 ]
[111 ] observed no decrease in blood pressure after an exercise bout when relying on PEH_I
analysis, but observed an increase in blood pressure from before to after a control
session, showing that time alone is associated with increases in blood pressure under
their specific experimental conditions. In such cases, PEH_II or PEH_III analysis
can demonstrate that the exercise session had a measurable hypotensive effect, blunting
the blood pressure increase that was observed with the control session [8 ]
[81 ]
[90 ]
[106 ]
[107 ]
[109 ]
[110 ]
[111 ]. Based on these arguments, whether or not a particular exercise session can promote
PEH can only be clearly determined if time effects are controlled by a control session.
Then, both PEH_II and PEH_III calculations can be employed to address this issue.
On the other hand, if the objective is to compare expected PEH among different exercises
protocols, instead of detecting its occurrence, PEH_I may be an option to save time.
PEH_II is often employed when studies rely on ambulatory blood pressure monitoring
[7 ]
[8 ]
[9 ]
[23 ]
[27 ]
[38 ]
[57 ]
[82 ]
[89 ]
[90 ]
[109 ]
[112 ] because of the challenges in using an ambulatory blood pressure monitor for more
than 24 hrs (i. e., these studies do not include a pre-exercise 24-h recording, just
a post-exercise and post-control recording). That said, there are also many studies
using this analysis for laboratory-based blood pressure measurements [22 ]
[25 ]
[26 ]
[28 ]
[30 ]
[32 ]
[33 ]
[35 ]
[39 ]
[40 ]
[44 ]
[50 ]
[51 ]
[52 ]
[55 ]
[56 ]
[60 ]
[61 ]
[62 ]
[63 ]
[64 ]
[67 ]
[68 ]
[77 ]
[78 ]
[80 ]
[84 ]
[94 ]
[98 ]
[99 ]
[107 ]. The main limitation of PEH_II analysis is that it does not account for day-to-day
variation in resting (pre-exercise) blood pressure [113 ]. Pre-intervention blood pressure might be different between the exercise and control
session, confounding the results obtained with PEH_II analysis.
PEH_III analysis overcomes the main limitations of PEH_I and PEH_II, controlling for
time effects and day-to-day variation in resting blood pressure. However, it demands
two experimental sessions and at least two blood pressure measurements during each
session (ideally more). In addition, there are some disadvantages beyond the obvious
additional time requirements for conducting studies. The PEH_III value has more degrees
of freedom, which decreases its reproducibility and increases the number of subjects
needed in these studies [105 ]. This makes a prior sample size calculation even more important when designing protocols
which will use PEH_III analysis.
Therefore, as stated before, all these procedures for calculating PEH are being used
in the literature. All of them have advantages and disadvantages, and they may lead
to contradictory outcomes. Thus, when designing, analyzing, and interpreting results
about PEH, researchers should take into consideration these strengths and limitations.
In addition, future research should relate these different approaches with meaningful
clinical outcomes.
Reporting results
Many studies include results for both systolic and diastolic blood pressure [22 ]
[23 ]
[24 ]
[27 ]
[31 ]
[38 ]
[39 ]
[45 ]
[52 ]
[62 ]
[65 ]
[66 ]
[68 ]
[69 ]
[71 ]
[73 ]
[76 ]
[80 ]
[83 ]
[89 ]
[94 ]
[109 ]
[110 ]
[112 ], whereas others report results for mean blood pressure [30 ]
[32 ]
[40 ]
[42 ]
[47 ]
[48 ]
[54 ]
[56 ]
[67 ]
[74 ]
[99 ]
[101 ]. It is possible that this is a reflection of researchers with a more clinical orientation
versus those who focus on systemic hemodynamics and cardiovascular regulation. Based
on the limitations exposed about oscillometric measurement, mean blood pressure should
be reported in PEH studies employing this method of measurement. Using auscultatory
method, systolic and diastolic represent the primary data, while mean blood pressure
is secondary (calculated). Systolic and diastolic blood pressure have been more often
related to risk for mortality and disease development [114 ], reflecting a clinical orientation, while mean blood pressure lends itself to discussions
of the determinants of PEH. Thus, the best conduct would be to report all three blood
pressure parameters in future studies with auscultatory PEH.
Another rationale for reporting all three parameters, is that some studies only report
systolic [33 ]
[36 ]
[37 ]
[46 ]
[51 ]
[52 ]
[53 ]
[63 ]
[66 ]
[73 ] or diastolic blood pressure [64 ] reductions in response to exercise, rather than finding all blood pressure parameters
following the same pattern. This is likely to be more than a statistical underpowering
of studies (and a type II error) and may reflect some of the complicated physiology
which underlies PEH in some contexts. For example, if one assumes, somewhat simplistically,
that changes of systolic blood pressure are more related to cardiac output and changes
of diastolic blood pressure are more related to systemic vascular resistance, divergent
patterns of PEH between systolic and diastolic blood pressure become reasonable predictions.
After performing aerobic exercise, PEH is most often mediated by a decrease of systemic
vascular resistance, except in some special conditions in which cardiac output is
reduced [2 ]
[3 ]. In contrast, resistance exercise generally causes PEH by a decrease in cardiac
output [79 ]
[109 ] or a mixture of patterns across subjects [110 ]
[111 ]. However, studies which find decreased cardiac output can also report a decreased
diastolic blood pressure [49 ]
[82 ]
[107 ]. Likewise for reductions in systemic vascular resistance and systolic blood pressure
[33 ]
[34 ]
[36 ]
[43 ]
[44 ]
[58 ]
[60 ]
[63 ]
[75 ]
[79 ]
[81 ]
[84 ]
[94 ]. Thus, the “simple” attribution of isolated effects on systolic and diastolic blood
pressure does not fit all observations. This is likely due to the more integrated
nature of the responses, with cardiac and vascular changes being linked by physiological
compensation, with superimposed baroreflex regulation of the heart [35 ] and peripheral vasculature [33 ]. Again, it is better to report all three blood pressure parameters in future studies
on PEH, especially if they portray divergent profiles following exercise.
Reporting individual responses: Do group means describe what is happening?
PEH can be observed in individuals with a broad range of characteristics [8 ] and after different exercise protocols, including aerobic, dynamic resistance, and
isometric exercise [5 ]
[6 ]. However, within a given study, it is expected that not all subjects present a uniform
blood pressure decrease after exercise (as shown in [Fig. 2 ]). As a further demonstration, consider [Fig. 2 ], which shows how a study may find no change in mean systolic blood pressure after
a session of aerobic exercise (panel a) despite the presence of PEH in the most of
the subjects (panel b). Indeed, a previous study that evaluated ambulatory blood pressure
after resistance and walking exercises in subjects with peripheral artery disease
did not observe any difference in mean values, but individual responses revealed that
many subjects presented clinically relevant blood pressure reductions [115 ]. Along these lines, Forjaz et al. [8 ] observed no change in the group mean blood pressure for hypertensive subjects and
a decrease in the group mean blood pressure for normotensive subjects, yet in both
groups there were subjects who increased, decreased, or maintained blood pressure
after the exercise session.
Fig. 2 Hypothetical data of systolic blood pressure (mmHg) assessed pre and post an exercise
session, demonstrating how a group’s means can mask important individual responses.
Panel a): data presented by means; Panel b) data presented by individual responses.
As in other areas of research, we recognize that these different individual responses
may be of great importance. There may be a strong rationale to classify subjects as
“responders” and “non-responders” [73 ]
[115 ]. However, similar to debates deriving from randomized control trials, there is still
no consensus on how to define a true “response”: whether it should be based on the
presence of clinically relevant change [116 ] or on clearly measurable change [117 ]. In fact, the clinically relevant change for PEH has yet to be defined. An option
might be when blood pressure response to exercise overcome the error of the blood
pressure measurement [118 ]. However, clearly measurable change is still not clear, since very few studies have
adequately investigated the reproducibility of individual PEH patterns [105 ].
As a result, this is an open area for future investigations that needs to be addressed.
Studies employing analysis of group mean data revealed the clinical relevance of PEH,
but individual response analysis may be necessary to advance comprehension of this
phenomenon, clarifying as well whether there is a role for responders and non-responders.
Analyses of individual responses have been highlighted in the literature, and this
approach shows great promise as a tool to identify who is responsive to a specific
intervention, based on an acute evaluation. However, the lack of standardization of
how PEH is assessed creates a barrier for moving this approach forward.
