Date post: | 25-Dec-2018 |
Category: |
Documents |
Upload: | truongdiep |
View: | 226 times |
Download: | 0 times |
REVIEW
High-intensity interval training: a review of its impact on glucosecontrol and cardiometabolic health
Sophie Cassidy1 & Christian Thoma2 & David Houghton1& Michael I. Trenell1
Received: 26 April 2016 /Accepted: 17 August 2016 /Published online: 28 September 2016# The Author(s) 2016. This article is published with open access at Springerlink.com
Abstract Exercise plays a central role in the manage-ment and treatment of common metabolic diseases, butmodern society presents many barriers to exercise.Over the past decade there has been considerable in-terest surrounding high-intensity interval training(HIIT), with advocates claiming it can induce healthbenefits of similar, if not superior magnitude tomoderate-intensity continuous exercise, despite reducedtime commitment. As the safety of HIIT becomes clearer,focus has shifted away from using HIIT in healthy individualstowards using this form of training in clinical populations. Thecontinued growth of metabolic disease and reduced physicalactivity presents a global health challenge and effective ther-apies are urgently required. The aim of this review is to ex-plore whether the acclaim surrounding HIIT is justified byexamining the effect of HIIT on glucose control, its ability toaffect cardiovascular function and the underlying mechanismsof the changes observed in those with common metabolicdiseases. It also explores translation of the research into clin-ical practice.
Keywords Cardiovascular system . Exercise . Exercisetherapy .Metabolic diseases . Metabolism . Physical fitness .
Review .Weight loss
AbbreviationsEDV End-diastolic volumeFMD Flow mediated dilationHIIT High-intensity interval trainingHRmax Maximum heart rateMICT Moderate-intensity continuous trainingNAFLD Non-alcoholic fatty liver diseasePGC-1α Peroxisome proliferator-activated receptor
gamma, coactivator 1, alphaRPE Rate of perceived exertionV:O2max Maximal oxygen consumption
V:O2peak Peak oxygen consumption
Why exercise?
Before the agricultural, industrial and digital ages, humansexpended large amounts of energy in activities centred onmaintaining shelter and procuring food and water [1]. Fastforward some 350 generations and the barriers to exerciseand physical activity in the 21st century are enormous.Sedentary behaviours, such as the use of mechanised transportand screen-based leisure pursuits have become the norm inmodern society. There is an urgent need therefore to find prac-tical, attractive and effective exercise therapies to combat thewave of inactivity sweeping through the western world.
Not only is exercise part of our nature, it is strongly associ-ated with reduced chronic disease risk. Globally, metabolic dis-orders such as themetabolic syndrome, non-alcoholic fatty liverdisease (NAFLD), type 2 diabetes and the closely associatedcluster of cardiovascular diseases are rapidly increasing [2].European and US treatment algorithms for these obesity drivenepidemics recommend weight loss and maintenance as a mainpriority across all stages [3, 4]. Conceivably, this can beachieved through energy restriction and/or physical exercise.
* Michael I. [email protected]
1 MoveLab, Institute of Cellular Medicine, The Medical School,Newcastle University, 4th Floor William Leech Building,Framlington Place, Newcastle upon Tyne NE2 4HH, UK
2 School of Interprofessional Health Studies, Auckland University ofTechnology, Auckland, New Zealand
Diabetologia (2017) 60:7–23DOI 10.1007/s00125-016-4106-1
Current management guidelines for these common meta-bolic conditions advise individuals to undertake around150 min of moderate-to-vigorous aerobic exercise per week,spread over most days of the week, in addition to resistancetraining on at least 2 days of the week [5, 6]. The emphasisremains on moderate-intensity continuous training (MICT);however there is mounting evidence that high-intensity inter-val training (HIIT) provides an alternative means of achievingthe same or greater health benefits vs MICT, provided thereare no medical contraindications to engaging in HIIT and thatit is well tolerated and preferred by the individual taking part.We refer readers to recent meta-analyses for a comprehensiveanalysis of the metabolic [7] and cardiorespiratory [8] benefitsof HIIT in patient groups. The aim of this review is to assim-ilate existing evidence and provide a clinically relevant narra-tive of the cardiometabolic benefits of HIIT in those withcommon metabolic diseases, before moving onto discussingits safety profile, tolerability and practical considerations fortranslation into clinical care. The information presented in thisreview is not part of a formal systematic review and, therefore,may not have been subjected to the rigor required for such asummary of the data currently available on HIIT.
What is HIIT?
HIIT can be described as ‘brief intervals of vigorous activityinterspersed with periods of low activity or rest’, which in-duces a strong acute physiological response (Fig. 1) [9]. Anumber of HIIT protocols have been adopted in the literature(see Table 1), but the majority of interventions use high-intensity intervals lasting between 1 and 4 min. The goal ofHIIT is to accumulate activity at an intensity that the partici-pant would be unable to sustain for prolonged periods (i.e. 80–95% of peak oxygen consumption (V
:O2peak ) or >90% of
maximum heart rate (HRmax), therefore the recovery time
should be sufficient to allow the subsequent interval to becompleted at the desired intensity. The total duration of aHIIT session tends to be ≥20 min, which actually makes itcomparable with recommendations for MICT, in terms of du-ration. There is also a sub-category of HIIT involving 10–30second intervals and intensities often exceeding 100%V:O2peak, i.e. ‘all-out’ exercise at a workload that is above
maximal aerobic capacity [10]. This is often called sprint in-terval training and has not been substantially tested in clinicalpopulations and will therefore not be covered further.
The vast majority of the published HIIT research, particu-larly in clinical populations, has used exercise modalitiesinvolving cycling, walking, and running, mostly carried outon stationary cycles and treadmills (see Table 1). However,other equipment, such as cross-trainers/ellipiticals [11], arereasonable options for some. Evidently there is clear variationthroughout the literature and it still remains to be determinedwhether an optimal HIIT protocol exists for metabolic diseasemanagement.
HIIT and metabolic health
Skeletal muscle molecular adaptations
A number of molecular adaptations have been identified with-in skeletal muscle following HIIT (Fig. 2). Skeletal muscle isthe primary site for glucose disposal via insulin- and non-insulin-mediated glucose uptake; the latter stimulated by mus-cular contraction. It therefore plays a large role in regulatingmetabolism.
Increased GLUT-4 content GLUT-4 content in the vastuslateralis increased by 369% following six sessions of HIITin type 2 diabetes patients [12]. Insulin resistance underliesmetabolic disease and although decreased GLUT-4 content is
Work Rest5 min warmup Work Rest Work Rest Work Rest Work3 min
cool down
RPE (Borg scale)
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
100
Heart rate(% HRpeak)
80
60
40
20
0
Fig. 1 An example of a HIITprotocol. Schematic of the HIITprotocol adopted by our group inadults with NAFLD [39] and type2 diabetes [38]. Intensity wasbased upon the perceived rate ofexertion (RPE), inducing a strongacute physiological response inheart rate (shown as % peak heartrate [HRpeak]), which increasesacross intervals
8 Diabetologia (2017) 60:7–23
Tab
le1
Effecto
fHIITon
insulin
andglucosemetabolism
inpatientswith
themetabolicsyndrome,NAFL
Dor
type
2diabetes
CG
M: s
ee a
bove
Ref
eren
ceB
asel
ine
popu
lati
onP
roto
cola
Cha
nges
in m
etab
olis
m
Acu
te (
sing
le s
essi
on)
effe
ct
Tjø
nna
et a
l 20
11 [
26]
4 m
en/7
wom
en, t
he m
etab
olic
syn
drom
e, 5
5±13
yea
rs,
BM
I 30
±2,
2m
axO
V34
±3
HII
T: 4
× 4
min
incl
ine
trea
dmill
wal
king
at 9
0–95
% H
Rm
axw
ith 3
m
in r
ecov
ery
peri
ods
at 7
0% H
Rm
ax
FG ↓
~15%
bel
ow b
asel
ine
for
72 h
vs
CO
N
4 m
en/4
wom
en, t
he m
etab
olic
syn
drom
e, 5
2±11
yea
rs,
BM
I 29
±2,
2m
axO
36±
3M
ICT
: 47
min
at 7
0% H
Rm
ax(m
atch
ed f
or e
nerg
y ex
pend
iture
with
H
IIT
)FG
↓~1
5% b
elow
bas
elin
e fo
r 24
h v
s C
ON
5 m
en/4
wom
en, t
he m
etab
olic
syn
drom
e, 5
0±9
year
s,
BM
I 32
±1,
2m
axO
32±
3C
ON
: res
ting
FG n
.s.
Gill
en e
t al
2012
[23
]7
adul
ts (
sex
n.r.
), ty
pe 2
dia
bete
s, 6
3±3
year
s, B
MI
31±
2,
2pea
kO
n.r.
HII
T: 1
0 ×
1 m
in c
yclin
g at
~89
% W
Rpe
akw
ith 6
0 s
pass
ive
rest
pe
riod
s R
elat
ive
to v
alue
s on
a n
on-e
xerc
isin
g co
ntro
l day
: 3 h
pos
tpra
ndia
l A
UG
C ↓
~35%
. Pos
t-mea
l pea
k gl
ucos
e ↓1
6% a
nd ti
me
in
hype
rgly
caem
ia ↓
65%
Kar
stof
t et a
l 20
14 [
25]
7 m
en/3
wom
en, t
ype
2 di
abet
es, 6
0±2
year
s, B
MI
28±
1,
2pea
kO
30±
3 (c
ross
over
des
ign)
HII
T: w
alki
ng a
t 89%
2p
eak
Ofo
r 3
min
inte
rval
s an
d 54
%
2pea
kO
for
3 m
in r
ecov
ery
× 1
h
MM
T: m
ean
gluc
ose
↓12%
vs
CO
N a
nd ↓
6% v
s M
ICT
. Mea
n in
crem
enta
l glu
cose
↓40
% v
s C
ON
and
↓29
% v
s M
ICT
CG
M: s
ame
day
mea
n gl
ucos
e ↓1
2% v
s M
ICT
and
n.s
. vs
CO
N.
Nex
t day
pos
t exe
rcis
e n.
s. v
s C
ON
and
MIC
T
MIC
T: w
alki
ng a
t 73%
2p
eak
O×
1 h
(m
atch
ed f
or e
nerg
y
expe
nditu
re w
ith H
IIT
)
MM
T: m
ean
gluc
ose
n.s.
vs
CO
N
CG
M: s
ame
day
mea
n gl
ucos
e n.
s. v
s C
ON
. Nex
t day
mea
n gl
ucos
e ↑8
% v
s C
ON
CO
N: r
estin
gM
MT
and
CG
M: s
ee a
bove
Litt
le e
t al
2014
[22
]2
men
/8 w
omen
, 6 w
ith I
FG, 4
1±11
yea
rs, B
MI
36±
7,
2pea
kO
22±
2H
IIT
: (m
orni
ng)
10 ×
1 m
in c
yclin
g in
terv
als
at ~
90%
WR
peak
with
1
min
rec
over
y pe
riod
s at
15%
WR
peak
CG
M: d
inne
r A
UG
C ↓
32%
vs
CO
N. D
inne
r po
stpr
andi
al g
luco
se
peak
↓41
% a
nd 3
9% v
s C
ON
and
MIC
T, r
espe
ctiv
ely.
Bre
akfa
st
AU
GC
↓36
% a
nd ↓
33%
vs
CO
N a
nd M
ICT
,resp
ectiv
ely.
B
reak
fast
pos
tpra
ndia
l glu
cose
pea
k ↓3
0% v
s C
ON
and
MIC
T
MIC
T: (
mor
ning
) cy
clin
g m
atch
ed w
ith H
IIT
for
tota
l wor
k at
~3
5% W
Rpe
ak
CG
M: d
inne
r A
UG
C ↓
22%
vs
CO
N. n
.s. f
or o
ther
mea
sure
d va
riab
les
CO
N: n
o ex
erci
se
V V
V
V V
V
V
V
Diabetologia (2017) 60:7–23 9
Tab
le1(continued)
Ref
eren
ceB
asel
ine
popu
lati
onP
roto
cola
Cha
nges
in m
etab
olis
m
Ter
ada
etal
20
16 [
24]
8 m
en/2
wom
en, t
ype
2 di
abet
es, 6
0±6
year
s, B
MI
31±
5,
2pea
kO
23±
7 H
IIT
fast
ed: (
exer
cise
pre
-bre
akfa
st)
trea
dmill
wal
king
1 m
in a
t 100
%
2pea
kO
with
3 m
in a
t 40%
2p
eak
O, f
or 6
0 m
in (
15 h
igh-
inte
nsity
inte
rval
s)
CG
M: 2
4 h
mea
n gl
ucos
e ↓1
6% v
s C
ON
. Tim
e >
10 m
mol
/l ↓5
8%.
Lun
ch p
ostp
rand
ial a
nd to
tal p
ost-m
eal A
UG
C ↓
88%
and
↓37
% v
s C
ON
, res
pect
ivel
y. B
reak
fast
and
din
ner
AU
GC
n.s
.
HII
Tfe
d: (
exer
cise
pos
t-br
eakf
ast)
, as
abov
eC
GM
: n.s
. for
all
mea
sure
s
MIC
Tfa
sted
: (ex
erci
se p
re-b
reak
fast
) tr
eadm
ill w
alki
ng a
t 55%
2pea
kO
for
60 m
inC
GM
: lun
ch a
nd d
inne
r po
stpr
andi
al a
nd to
tal p
ost-m
eal A
UG
C
↓44%
, ↓95
% a
nd ↓
53%
vs
CO
N, r
espe
ctiv
ely.
n.s
. for
24
h m
ean
gluc
ose,
tim
e >
10 m
mol
/l an
d br
eakf
ast p
ostp
rand
ial A
UG
C
MIC
Tfe
d: (
exer
cise
pos
t-br
eakf
ast)
, as
abov
eC
GM
: tot
al p
ost-
mea
l AU
GC
↓31
%. n
.s. f
or a
ll ot
her
mea
sure
s
CO
N: n
o ex
erci
seC
GM
: see
abo
ve
Cum
ulat
ive
(mul
tiple
ses
sion
) ef
fect
Litt
le e
t al
2011
[12
]8
adul
ts (
sex
n.r.
), ty
pe 2
dia
bete
s, 6
3±8
year
s, B
MI
32±
6,
2pea
kO
n.r.
H
IIT
: 10
× 1
min
inte
rval
s at
~90
% H
Rm
axw
ith 1
min
res
t per
iods
; 3
sess
ions
/wee
k fo
r 2
wee
ksC
GM
(24
h): m
ean
gluc
ose
↓13%
, 24
h r
AU
GC
↓14
%, s
um o
f 3
h po
stpr
andi
al A
UG
C ↓
30%
Tjø
nna
et a
l 20
08 a
nd
2011
[20
,26]
4 m
en/7
wom
en, t
he m
etab
olic
syn
drom
e, 5
5±13
yea
rs,
BM
I 30
±2,
2m
axO
34±
3H
IIT
: 4 ×
4 m
in in
clin
e tr
eadm
ill w
alki
ng a
t 90–
95%
HR
max
with
3
min
rec
over
y pe
riod
s at
70%
HR
max
; 3 s
essi
ons/
wee
k fo
r 16
wee
ksIS
(fr
om H
OM
A)
↑24%
vs
MIC
T a
nd C
ON
FG ↓
4.3%
4 m
en/4
wom
en, t
he m
etab
olic
syn
drom
e, 5
2±11
yea
rs,
BM
I 29
±2,
2m
axO
36±
3M
ICT
: 47
min
at 7
0% H
Rm
ax. 3
ses
sion
s/w
eek
for
16 w
eeks
(m
atch
ed f
or e
nerg
y ex
pend
iture
with
HII
T)
IS (
from
HO
MA
) n.
s.
FG n
.s.
5 m
en/4
wom
en, t
he m
etab
olic
syn
drom
e, 5
0±9
year
s,
BM
I 32
±1,
2m
axO
32±
3C
ON
: no
inte
rven
tion
IS (
from
HO
MA
) n.
s.
FG n
.s.
Sten
svol
d et
al
201
0 [3
5]11
adu
lts (
sex
n.r.
), th
e m
etab
olic
syn
drom
e, 5
0±10
yea
rs,
BM
I 31
±4,
2p
eak
O34
±10
, str
atif
ied
by a
ge a
nd s
exH
IIT
: tre
adm
ill w
alki
ng/r
unni
ng a
t 90–
95%
HR
peak
for
4 m
in ×
4
bout
s w
ith a
ctiv
e re
cove
ry p
erio
ds o
f 3
min
at 7
0% H
Rpe
ak;3
se
ssio
ns/w
eek
for
12 w
eeks
IS (
from
HO
MA
) n.
s.
FG n
.s.
HbA
1cn.
s.
11 a
dults
(se
x n.
r.),
the
met
abol
ic s
yndr
ome,
51±
8 ye
ars,
B
MI
32±
4,
2pea
kO
32±
5R
ET
: 3 s
ets
of 8
–12
repe
titio
ns ×
3–5
exe
rcis
es; 3
ses
sion
s/w
eek
for
12 w
eeks
IS (
from
HO
MA
)n.s
.FG
n.s
. H
bA1c
n.s.
VV
V
V
V
V V V V V
10 Diabetologia (2017) 60:7–23
Tab
le1(continued)
Ref
eren
ceB
asel
ine
popu
lati
onP
roto
cola
Cha
nges
in m
etab
olis
m
10 a
dults
(se
x n.
r.),
the
met
abol
ic s
yndr
ome,
53±
10 y
ears
, B
MI
30±
4,
2pea
kO
28±
6H
IIT
+R
ET
: HII
T a
s ab
ove,
2 s
essi
ons/
wee
k, 3
res
ista
nce
exer
cise
s fo
r 8–
12 r
epet
ition
s, 1
ses
sion
/wee
k, a
ll fo
r 12
wee
ksIS
(fr
om H
OM
A)
n.s.
FG n
.s.
HbA
1cn.
s.
11 a
dults
(se
x n.
r.),
the
met
abol
ic s
yndr
ome,
47±
10 y
ears
, B
MI
32±
4,
2pea
kO
34±
10C
ON
: no
exer
cise
IS (
from
HO
MA
) n.
s.
FG n
.s.
HbA
1cn.
s.
Ter
ada
et a
l 20
13 [
46]
4 m
en/4
wom
en, t
ype
2 di
abet
es, 6
2±3
year
s, B
MI
28±
4,
2pea
kO
23±
5H
IIT
: cyc
ling
1 m
in 1
00%
2r
eser
veO
+ 3
min
20%
2p
eak
Ofo
r 30
min
in w
eeks
1–4
, 45
min
in w
eeks
5–8
, 60
min
in w
eeks
9–1
2; 5
se
ssio
ns/w
eek
for
12 w
eeks
FG n
.s.
HbA
1cn.
s.
4 m
en/3
wom
en, t
ype
2 di
abet
es, 6
3±5
year
s, B
MI
33±
5,
2pea
kO
23±
5M
ICT
: 40%
2r
eser
veO
for
30 m
in in
wee
ks 1
–4, 4
5 m
in in
wee
ks
5–8,
60
min
in w
eeks
9–1
2; 5
ses
sion
s/w
eek
for
12 w
eeks
(m
atch
ed
for
ener
gy e
xpen
ditu
re w
ith H
IIT
)
FG n
.s.
HbA
1cn.
s.
Ear
nest
et a
l 20
13 [
33]
21 m
en, ‘
at r
isk
of ty
pe 2
dia
bete
s’, 4
8±9
year
s, B
MI
30+
2,
2pea
kO
30±
3H
IIT
: cyc
ling
6 w
eek
MIC
T r
un-i
n ph
ase,
then
2 m
in in
terv
als
at
90–9
5%
2pea
kO
and
2 m
in 5
0%
2pea
kO
reco
very
, sta
rtin
g at
2
inte
rval
s/se
ssio
n an
d in
crea
sed
by 2
inte
rval
s pe
r w
eek
until
8
inte
rval
s/se
ssio
n in
wee
k 9;
3–4
ses
sion
s/w
eek
for
12 w
eeks
24 h
pos
t exe
rcis
e: 2
h g
luco
se ↓
13%
, HO
MA-
IR ↓
22%
and
fas
ting
insu
linn.
s.
72 h
pos
t exe
rcis
e: H
OM
A-I
R ↓
16%
, 2 h
glu
cose
and
fas
ting
insu
lin n
.s.
16 m
en, ‘
at r
isk
of ty
pe 2
dia
bete
s’, 4
9±9
year
s, B
MI
31±
3,
2pea
kO
28±
5M
ICT
: 40%
2p
eak
O, 3
–4 s
essi
ons/
wee
k fo
r 12
wee
ks (
mat
ched
for
ener
gy e
xpen
ditu
re w
ith H
IIT
)
24 h
pos
t exe
rcis
e: 2
h g
luco
se ↓
12%
, HO
MA-
IR n
.s
72 h
pos
t exe
rcis
e: 2
h g
luco
se, H
OM
A-I
R a
nd f
astin
g in
sulin
n.s
.
Kar
stof
t et a
l 20
13 a
nd
2014
[36
,14]
7 m
en/5
wom
en, t
ype
2 di
abet
es, 5
8±2
year
s, B
MI
29±
1,
2pea
kO
27±
2H
IIT
: wal
king
at >
70%
pea
k en
ergy
exp
endi
ture
× 3
min
inte
rval
s,
<70
% o
f pe
ak e
nerg
y ex
pend
iture
× 3
min
rec
over
y, f
or 1
h; 5
se
ssio
ns/w
eek
for
16 w
eeks
CG
M: 4
8–72
h p
ost e
xerc
ise:
mea
n gl
ucos
e ↓9
% a
nd m
axim
um
gluc
ose
↓20%
vs
CO
N
96–1
20 h
pos
t exe
rcis
e: f
astin
g in
sulin
↓19
% v
s C
ON
. FG
, 2 h
gl
ucos
e, A
UG
C a
nd H
bA1c
n.s.
8 m
en/4
wom
en, t
ype
2 di
abet
es, 6
1±2
year
s, B
MI
30±
2,
2pea
kO
26±
1M
ICT
: wal
king
at >
55%
of
peak
ene
rgy
expe
nditu
re f
or 1
h; 5
se
ssio
ns/w
eek
for
16 w
eeks
(mat
ched
for
ene
rgy
expe
nditu
re w
ith
HII
T)
CG
M: 4
8–72
h p
ost e
xerc
ise:
n.s
. cha
nges
.
96–1
20 h
pos
t exe
rcis
e: f
astin
g in
sulin
, FG
, 2 h
glu
cose
, AU
GC
an
d H
bA1c
n.s.
5 m
en/3
wom
en, t
ype
2 di
abet
es, 5
7±3
year
s, B
MI
30±
2,
2pea
kO
25±
2C
ON
: no
exer
cise
CG
M: 4
8–72
h p
ost e
xerc
ise:
mea
n gl
ucos
e ↑1
7% v
s ba
selin
e
96–1
20 h
pos
t exe
rcis
e: f
astin
g in
sulin
, FG
, 2 h
glu
cose
, AU
GC
and
H
bA1c
n.s.
V V
VV
V
V
VV
V
VV
V V VV
Diabetologia (2017) 60:7–23 11
Tab
le1(continued)
Ref
eren
ceB
asel
ine
popu
lati
onP
roto
cola
Cha
nges
in m
etab
olis
m
Hol
leki
m-
Stra
nd e
t al
2014
[44
]
20 a
dults
(se
x n.
r.),
type
2 d
iabe
tes,
59±
5 ye
ars,
BM
I 30
±3,
2p
eak
O32
±6
HII
T: 4
× 4
min
inte
rval
s at
90–
95%
HR
max
for
40 m
in; 3
se
ssio
ns/w
eek
for
12 w
eeks
(ex
erci
se m
ode
n.r.
)H
bA1c
↓6%
HO
MA
-IR
n.s
.
17 a
dults
(se
x n.
r.),
type
2 d
iabe
tes,
55±
5 ye
ars,
BM
I 30
±4,
2p
eak
O33
±7
MIC
T: ≥
10 m
in/b
out,
210
min
/wee
k ×
12
wee
ks (
sess
ions
wer
e do
ne a
t hom
e w
ithou
t sup
ervi
sion
, exe
rcis
e m
ode
n.r.
) H
bA1c
and
HO
MA
-IR
n.s
.
Mitr
anun
et a
l 20
14 [
41]
5 m
en/9
wom
en, t
ype
2 di
abet
es, 6
1±3
year
s, B
MI
30±
1,
2pea
kO
n.r.
HII
T: t
read
mill
50%
2p
eak
Oor
20 m
in in
wee
ks 1
–2. 4
× 1
min
inte
rval
s at
80%
2p
eak
Ow
ith 4
min
rec
over
y at
50%
2p
eak
Oin
wee
ks 3
–6. 6
× 1
min
inte
rval
s at
85%
2p
eak
Ow
ith 4
min
reco
very
at 6
0%
2pea
kO
in w
eeks
7–1
2; 3
ses
sion
s/w
eek
for
12
wee
ks
48–7
2 h
post
exe
rcis
e: H
OM
A-I
R ↓
19%
vs
CO
N. H
bA1c
↓10%
vs
CO
N. F
G ↓
14%
5 m
en/9
wom
en, t
ype
2 di
abet
es, 6
2±3
year
s, B
MI
30±
1,
2pea
kO
n.r.
MIC
T: t
read
mill
50%
2p
eak
Ofo
r20
min
in w
eeks
1–2
, 60%
VO
2pea
kfo
r 20
min
in w
eeks
3–6
and
65%
2p
eak
Ofo
r 30
min
s in
wee
ks 7
–12;
3 s
essi
ons/
wee
k fo
r 12
wee
ks (
mat
ched
for
ene
rgy
expe
nditu
re w
ith H
IIT
)
48–7
2 h
post
exe
rcis
e: H
OM
A-I
R ↓
18%
vs
CO
N. F
G ↓
13%
. HbA
1c
n.s.
5 m
en/1
0 w
omen
, typ
e 2
diab
etes
, 61±
2 ye
ars,
BM
I 30
±0.
4,
2pea
kO
n.r.
CO
N: n
o ex
erci
se48
–72
h po
st e
xerc
ise:
HO
MA
-IR
, FG
and
HbA
1cn.
s.
Shab
an e
t al
2014
[37
]3
men
/6 w
omen
, typ
e 2
diab
etes
, 40±
10 y
ears
, BM
I 34
±5,
2pea
kO
20±
4 H
IIT
: 30
s cy
clin
g at
100
% e
stim
ated
pea
k w
orkl
oad
× 4
with
4
min
rec
over
y at
25%
est
imat
ed p
eak
wor
kloa
d; 3
ses
sion
s/w
eek
for
2 w
eeks
Glu
cose
imm
edia
tely
pos
t exe
rcis
e ↓0
.95
mm
ol/l
(mea
sure
d af
ter
each
of
the
six
sess
ions
). F
G, f
astin
g in
sulin
, HO
MA
-IR
n.s
.
Hal
lsw
orth
et
al 2
015
[39]
6 m
en/6
wom
en, N
AFL
D, 5
4±10
yea
rs, B
MI
31±
4,
2pea
kO
22±
6H
IIT
: cyc
ling
5 in
terv
als
at R
PE o
f 16
–17
(‘ve
ry h
ard’
), in
terv
alle
ngth
2 m
in in
wee
k 1,
incr
easi
ng 1
0 s/
wee
k, r
ecov
ery
was
3 m
in
incl
udin
g 1
min
ligh
t res
ista
nce
exer
cise
; 3 s
essi
ons/
wee
k fo
r 12
w
eeks
2h g
luco
se ↓
15%
. FG
, fas
ting
insu
lin, H
bA1c
an
d H
OM
A2-
IR n
.s.
10 m
en/1
wom
an, N
AFL
D, 5
2±12
yea
rs, B
MI
31±
5,
2pea
kO
25±
6C
ON
: no
exer
cise
2h g
luco
se, F
G, f
astin
g in
sulin
, HbA
1c an
d H
OM
A2-
IR n
.s.
V V
V V
V
V V V
V
V
V
V
V
V
V
12 Diabetologia (2017) 60:7–23
Tab
le1(continued)
HbA
1c↓4
% a
nd 2
h g
luco
se ↓
6% v
s C
ON
. FG
, 2 h
AU
GC
and
H
OM
A2-
IR n
.s.
Ref
eren
ceB
asel
ine
popu
lati
onP
roto
cola
Cha
nges
in m
etab
olis
m
Rob
inso
n et
al
201
5 [4
5]3
men
/17
wom
en, ‘
elev
ated
ris
k of
type
2 d
iabe
tes
base
d on
HbA
1c’,
52±
10 y
ears
, BM
I 33
±7,
2p
eak
O20
±3
HII
T: (
self
-sel
ecte
d m
ode)
1 m
in in
terv
als
at 8
5–90
% W
max
an
d 1
min
rec
over
y at
20%
Wm
ax ×
4 in
ses
sion
1 (
4 in
terv
als
in s
essi
on 1
an
d 10
inte
rval
s by
ses
sion
10,
incr
emen
t n.r
.); 5
ses
sion
s/w
eek
for
2 w
eeks
48–7
2 h
post
exe
rcis
e: F
ruct
osam
ine
↓22%
. FG
, fas
ting
insu
lin a
nd
HO
MA
-IR
n.s
.
4 m
en/1
4 w
omen
, ‘el
evat
ed r
isk
of ty
pe 2
dia
bete
s ba
sed
on H
bA1c
’, 5
2±10
yea
rs,
BM
I 31
±4,
2p
eak
O21
±5
MIC
T: (
self
-sel
ecte
d m
ode)
at 6
0–65
% W
max
for
20 m
in in
ses
sion
1
(inc
reas
ing
to 5
0 m
in in
ses
sion
10
to m
atch
tota
l wor
k of
HII
T
grou
p). 5
ses
sion
s/w
eek
for
2 w
eeks
(m
atch
ed f
or e
nerg
y ex
pend
iture
with
HII
T)
48–7
2 h
post
exe
rcis
e: F
G ↓
5% v
s H
IIT
. Fru
ctos
amin
e ↓1
3%.
Fast
ing
insu
lin a
nd H
OM
A-I
R n
.s.
Fex
et a
l 201
5 [1
1]4
men
/12
wom
en, I
FG o
r ty
pe 2
dia
bete
s, 6
0±6
year
s,
BM
I 35
±5,
2p
eak
O40
±8
HII
T: e
llipt
ical
/cro
ss-t
rain
er 3
0 s
inte
rval
at 8
0–85
% H
Rm
ax a
nd 9
0 s
activ
e re
cove
ry (
inte
nsity
n.r
.) ×
20
min
; 3 s
essi
ons/
wee
k fo
r 12
w
eeks
FG ↓
8%H
bA1c
n.s
.
Mad
sen
et a
l 20
15 [
34]
3 m
en/7
wom
en, t
ype
2 di
abet
es, 5
2±2
year
s, B
MI
31±
1,
2pea
kO
22±
1H
IIT
: cyc
ling
10 ×
1 m
in in
terv
als
at ~
90%
HR
max
with
1 m
in r
est
(1:1
res
t: in
terv
al r
atio
); 3
ses
sion
s/w
eek
for
8 w
eeks
FG ↓
~11%
. 2 h
glu
cose
↓~1
3%. H
bA1c
↓~4
%. H
OM
A-I
R ↓
~17%
5 m
en/8
wom
en, ‘
heal
thy
cont
rol’
, 56±
2 ye
ars,
BM
I 31
±1,
2p
eak
O26
±2
FG, 2
h g
luco
se a
nd H
bA1c
n.s.
H
OM
A-I
R n
.s.
Cas
sidy
et a
l 20
16 [
38]
10 m
en/2
wom
en, t
ype
2 di
abet
es, 6
1±9
year
s, B
MI
31±
5,
2pea
kO
22±
5H
IIT
: cyc
ling
5 in
terv
als
at R
PE o
f 16
–17
(‘ve
ry h
ard’
), in
terv
al
leng
th 2
min
in w
eek
1 in
crea
sing
10
s/w
eek,
rec
over
y w
as 3
min
in
clud
ing
1 m
in li
ght r
esis
tanc
e ex
erci
se; 3
ses
sion
s/w
eek
for
12
wee
ks
8 m
en/3
wom
en, t
ype
2 di
abet
es, 5
9±9
year
s, B
MI
32±
6,
2pea
kO
20±
6C
ON
: no
exer
cise
FG ↑
17%
. 2 h
glu
cose
↑10
%. 2
h r
AU
GC
↑13
%. H
bA1c a
nd
HO
MA
2-IR
n.s
.
V V
V
V
V
V
V
Studypopulatio
ndemographics(sam
plesize
bysex,healthor
activ
itydescription,age,body
massindex(kg/m
2),andV: O
2peak(m
lkg−
1min−1)o
rV: O
2max
(mlkg−
1min−1)asreported
bystudyauthorsare
provided
asmeans
±SDandroundedto
nearestw
holenumber
Samplesizesarebasedon
thoseincluded
inthefinalanalysis
Resultshave
been
convertedto%
change
from
baselin
eifthechange
was
statisticallysignificant,andalso
notedifthiswas
significantrelativetocomparisongroup(s).W
heredatawas
reported
ingraph
form
itmay
noth
avebeen
feasibleto
accurately
calculate%
change,thusapproxim
ate%
aregiven
a Coreexercise
with
outw
arm
upandcool
downprotocols,consistin
gpredom
inantly
of5–10
min
periodsof
light-to-moderate-intensity
continuous
activ
ity
AUGC,areaunderthe
glucosecurve;CGM,contin
uous
glucosemonito
ring;C
ON,controlgroup;FG
,fastin
gglucose;IFG,impaired
fastingglucose;IS,insulinsensitivity;M
MT,mixed
mealtest;n.r.,not
reported;n.s.,no
statisticallysignificantchange/difference;R
ER,respiratory
exchange
ratio
atrest;R
ET,resistance
exercise
training
group;WRpeak,peakworkload;W
max,m
axim
alworkload;↑,increase;
↓,decrease
Diabetologia (2017) 60:7–23 13
not the cause of insulin resistance, any increase in this proteinimproves glucose transport within skeletal muscle [13].Another study found that 16 weeks of HIIT in type 2 diabetespatients induced higher membrane-bound GLUT-4 andGLUT-4 mRNA levels in comparison with energy matchedMICT, but no rise in overall GLUT-4 protein content wasobserved [14]. The reduced intensity of intervals in this study
(70% V:O2peak) compared with other HIIT protocols is worth
noting, however, as is the fact that biopsies were obtained5–6 days post-training.
Mitochondrial adaptations Reduced mitochondrial content[15], mitochondrial function [16] and markers of mitochon-drial biogenesis in skeletal muscle are commonly observed inindividuals with metabolic disease [17] and have been sug-gested to contribute to insulin resistance. In adults with type 2diabetes, 2 weeks of HIIT (90% HRmax) significantlyincreased mitochondrial capacity evidenced by an increasein citrate synthase activity and raised content of electron trans-port chain complexes [12]. In contrast, 16 weeks of intervalwalking (70% V
:O2peak) only elicited changes in citrate syn-
thase mRNA expression but not in citrate synthase activity.However, as stated above, this could be due to the lowerintensity of intervals used in this study compared with tradi-tionally adopted HIIT protocols and because muscle biopsieswere obtained 5–6 days post-training [14].
Peroxisome proliferator-activated receptor, gamma, coactiva-tor 1, alpha (PGC-1α) regulates muscle mitochondrial
biogenesis [18]. Following HIIT, increases in nuclear PGC-1αlevels have been observed [19], as well as increases in totalPGC-1α vastus lateralis content in biopsy samples. Similarchanges were not observed following energy matched MICT[20]. It has been proposed that the fluctuations in ATP turnoverduring interval training, which differs from the usual steady stateconditions of ATP production, could activate signalling path-ways that lead to this increase in PGC-1α following HIIT [21].
Sarcoplasmic reticulum Sarcoplasmic reticulum Ca2+
handling plays an important role in muscle fatigue andincreased Ca2+ reuptake into the sarcoplasmic reticulum hasbeen demonstrated following HIIT, but not MICT in adultswith the metabolic syndrome [20]. Increases as large as50–60% were observed in Ca2+ reuptake, significantly im-proving the work capacity of the muscle and thereby contrib-uting to improvements in fitness following HIIT. This studyindicates that HIIT provides a more potent stimulus comparedwith MICT to induce skeletal muscle adaptations. However,definite conclusions cannot be drawn from one small study,with a total sample size of 32 participants.
HIIT and glucose control
Key studies on the impact of HIIT on glucose control havebeen summarised in Table 1. These adopt a range of protocolsand cover both the acute and training responses to HIIT.
Cardiometabolic effects of HIIT
3
1
2
Changes to cardiovascular system
1 Torsion = myocardial damage 2 EDV = preload 3 Ca2+ handling = SV + EF4 FMD = O2 supply
3
2
41
Changes in skeletal muscle
1 Ca2+ reuptake into SR = muscle work capacity 2 Mitochondria biogenesis = oxidative capacity 3 GLUT4 = glucose transport
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+ Ca2+
Ca2+
Ca2+
NO
NO
NO
Fig. 2 Cardiometabolic effects ofHIIT. The figure depicts thepreviously reported muscular andcardiovascular impact of HIIT inthose with common metabolicdiseases. In boxes of text: upwardarrow, increase; downward arrow,decrease. EDV, end diastolicvolume; EF, ejection fraction;FMD, flowmediated dilation; SR,sarcoplasmic reticulum; SV,stroke volume
14 Diabetologia (2017) 60:7–23
The acute response Few studies have assessed the acuteresponse to HIIT in patients with metabolic disorders andthose that have are summarised in Table 1. Relative to no ex-ercise, a single session of HIIT reduces same-day postprandialarea under the glucose curve in those with impaired fastingglucose [22] or type 2 diabetes [23, 24]. Similarly, HIIT isassociated with reduced time of glucose being ≥10 mmol/L[23, 24]. However, studies assessing the effect of HIITon mean24 h glucose levels have been less consistent, with some indi-cating no effect [22, 23, 25], and one showing a reduction butonly when the exercise was performed in a fasted state [24].When comparedwith energymatchedMICT, HIIT tended to beslightly superior [22, 25, 26] (see Table 1). As measurementswere not reported much past 24 hours post exercise, the dura-tion of these effects is uncertain. Earlier work relying on chang-es in fasting glucose to assess impact, suggests a measurableeffect may last as long as 72 h post HIIT, but for a shorter periodfollowing MICT [26].
Since postprandial glucose excursions are strong predictorsof cardiovascular disease [27], which may be due to possibleinductions of oxidative stress and micro/macrovascular dam-age [28], the above findings are of clinical relevance.Unfortunately, the available research does not provide clarityon dose–response in terms of intensity, duration or total ener-gy expenditure. Further head-to-head comparisons of differentprotocols are required. However, it is worth noting that a pro-tocol of 1 min intervals with 1 min recovery, repeated tentimes (a modest time investment), improves acute glucosecontrol [22, 23].
The transient nature of changes in glucose metabolism inresponse to exercise is well documented. Insulin-independentglucose disposal is increased during and for approximately60 min post exercise [29]. Insulin-dependent glucose disposalincreases for several hours to a few days following exercise[29, 30]. These effects are localised to contracting muscle[29], thus exercise involving a larger muscle mass is prefera-ble. Higher-intensity exercise has been shown to recruit alarger proportion of muscle fibres compared with moderate-intensity exercise [31], which may explain greater improve-ments in glucose regulation following HIIT. In light of theseacute adaptations, patients should be recommend not tohave more than two exercise-free days, in accordance withguidelines [32].
The training response Of the studies assessing the effects oflonger-term HIIT (≥2 weeks), some report reduced fastingglucose [11, 20, 26, 33, 34], while others report no change[35–39] (Table 1). Where reductions in fasting glucose areobserved, they appear to be similar to those seen followingMICT [7]. Fasting glucose is predominantly a marker ofhepatic insulin sensitivity. After just 1 week of a diet verylow in energy (very low calorie diet; 2510 kJ/day [600 kcal/day]), liver fat content decreased by 30%, hepatic insulin
sensitivity significantly improved and fasting glucose fell by35% in adults with type 2 diabetes [40]. The reduction in fastingglucose following participation in HIIT is generally smaller [11,20, 26, 34, 41], (≤14%, see Table 1), suggesting that exercise(whether HIIT or MICT), lacks potency for improving hepaticinsulin sensitivity, when compared with consumption of a verylow calorie diet. This is most likely because exercise elicits asmaller energy deficit than that achieved with a modest changein eating behaviour. For example, to achieve an energy deficitsimilar to that achieved by reducing energy intake by the equiv-alent of the energy in a blueberry muffin (∼1891 kJ [452 kcal]),a 68 kg female would need to run approximately 38 min at apace of 9.7 km/h [42]. We did, however, show that HIIT wasable to significantly reduce liver fat and, thereby, fasting glu-cose in some type 2 diabetes individuals [38], but the averagereduction in liver fat did not result in a significant reduction infasting glucose levels. Whether an increase in the duration ofthe HIIT intervention (i.e. >12 weeks) would achieve a reduc-tion in fasting glucose levels in this cohort is yet to bedetermined.
HIIT has been shown to improve peripheral insulin sensitiv-ity in those with impaired metabolic control. The molecularadaptations to HIIT described above, including raised GLUT-4content, increased aerobic enzyme capacity and mitochondrialbiogenesis, have all been associated with improved peripheralinsulin sensitivity [13, 43]. Studies assessing the metabolic im-pact of HIIT in those with common metabolic diseases havefound no change in HOMA-IR [37–39, 44, 45], whereas othershave shown an approximate 20% improvement compared witha control group [20, 33, 34, 41] (Table 1). When compared withMICT, HIIT seems to have a small but significant benefit oninsulin resistance [7].
HIIT can also decrease HbA1c [34, 38, 41, 44], yetsome studies have reported no change [11, 36, 39, 46](Table 1). Although there have been a number of studies pub-lished since, a meta-analysis found that a 0.47% absolute re-duction in HbA1c is observed with HIIT in adults with com-mon metabolic diseases, compared with controls [7]. This isslightly lower than the 0.6% absolute HbA1c reductionobserved following aerobic and resistance exercise in type 2diabetes [47]. Both HIIT and other forms of exercise comparewell with improvements achieved throughmetformin [48] andare likely to have clinical benefits, since a 1% absolute rise inHbA1c leads to a 21% increased risk of diabetes related death,a 14% increased risk of myocardial infarction and a 37% in-creased risk of myocardial infarction [47].
Other indicators of glucose control, such as 2 h glucosefollowing an oral glucose challenge and glucose AUC aresimilarly inconsistent across studies (see Table 1). Severalexplanations for the reported inconsistencies across studiesinclude differences in study populations, exercise protocolsand the degree of volunteer supervision during exercise.However, the most plausible explanation is the variation in
Diabetologia (2017) 60:7–23 15
time of post-intervention measures relative to the last bout ofexercise. Studies reporting both the acute and cumulativeeffect of HIIT have consistently shown that changes in indi-cators of glucose control last between 24 and 72 h post exer-cise [25, 26, 33, 36] (Table 1). Only one study has demon-strated a longer-term adaptation, in which fasting insulin wasreduced 96–120 h post exercise [36]. Greater inter-study con-sistency in the timing of post exercise assessments is warrant-ed in the future; continuous glucose monitoring for at least72 h post exercise and HbA1c assessments may also allowus to gauge benefit better.
Collectively, the improvements in glucose control follow-ing HIIT are clinically relevant but do not surpass those seenfollowing the traditionally used MICT with regards to fastingglucose, HbA1c and fasting insulin [7]. HIIT does seem to leadto greater improvements in peripheral insulin sensitivity [7],but overall the use of HIIT for improving glycaemic outcomesshould not be over-emphasised compared with other forms ofexercise training.
HIIT and cardiovascular health
Cardiovascular complications are the leading cause of mortal-ity in those with common metabolic diseases [49, 50]. Theinterval design of HIIT to include rest periods enables patientsto accumulate time at higher exercise intensities, thereby chal-lenging the cardiovascular system. Limited evidence indicatesthat HIIT provides a stronger stimulus thanMICT for elicitingmyocardial improvements. Alongside the beneficial impact ofHIIT on vascular and cardiorespiratory fitness, this suggeststhat the cardiovascular benefit of HIIToutweighs the metabol-ic benefit.
Cardiac adaptations: molecular mechanisms
Because of the difficulty of obtaining human myocardial tissue,most evidence for the molecular adaptations to high-intensityexercise comes from experimental rodent models, the hearts ofwhich bear similarities to human hearts and mimic the humancardiac response to exercise training [51, 52]. The db/dbmouse model provides a good representation of the humanheart in diabetic patients. Following 13 weeks of HIIT, contrac-tility and Ca2+ handling were restored to normal levels as aresult of raised transverse tubule (T-tubule) density, sarcoplas-mic reticulum synchrony of Ca2+ release and sarcoplasmic re-ticulum Ca2+-ATPase (SERCA2a; Ca2+ transporter) activity[53]. These adaptations occurred despite no improvement inglucose or insulin levels, demonstrating the direct impact ofHIIT upon the myocardium. Similar adaptations have been ob-served in heart failure and healthy rodent models [54, 55], withgreater changes occurring following high-intensity exercise
(85–90% maximal oxygen consumption [V:O2max]) compared
with moderate-intensity exercise (65–70% V:O2max) [55].
Exercise also activates the phosphoinositol-3 kinase/Akt/mammalian target of rapamycin (mTOR) signal transductionpathway that leads to higher ribosomal biogenesis and proteinsynthesis, and thus induces physiological hypertrophy toa greater extent following high- (85–90% V
:O2max) vs
moderate- (65–70% V:O2max) intensity exercise [52, 55]. The
exercise-induced pathways activated in disease models maydiffer [54], but both healthy and disease rodent modelsindicate that exercise stimulates important transcription-al, translational and post-translational regulatory mecha-nisms that lead to structural remodelling of cardiac tis-sue and, thereby, improved strength of cardiac contrac-tions [52].
Cardiac structure
Adults with common metabolic diseases display left ventric-ular concentric remodelling, which represents a reduction inend-diastolic volume (EDV) and is also known as pathologi-cal hypertrophy [56, 57]. This reduction in EDV occurs inresponse to stress signals and is reflective of a build-up ofcollagen in the myocardium [58]. HIIT, on the other hand,has been shown to induce physiological hypertrophy [38],increasing left ventricular wall mass and EDV by means of aphysiological response to growth signals [58]. The number ofstudies investigating cardiac structure following HIIT is small;our group showed an 8 ml increase in EDV following12 weeks of HIIT in type 2 diabetes patients [38], but noimprovements in NAFLD patients [39]. Both of these studiescompared HIITwith a non-exercise control, rather thanMICT.That being said, HIIT has been shown to be superior to energymatched MICT in eliciting structural remodelling in thosewith hypertension [59] and heart failure [60].
Cardiac function
Systolic function Stroke volume and ejection fraction, twomeasures of the contractile capabilities of the heart, arereduced in those with metabolic disease [57]. Twelve weeksof HIIT induces systolic improvements in adults with type 2diabetes [38, 44], hypertension [59] and heart failure [60].Following 12 weeks of HIIT in heart failure patients, Wisløffet al [60] demonstrated a 35% and 17% relative increase inejection fraction and stroke volume, respectively, but nochange in these variables following energy matched MICT.These improvements are equal to those seen with commonlyused prescription medications, such as ACE inhibitors or betablockers [61]. Twelve weeks of HIIT in hypertensive patientsimproved early events in systole, which correlate to contrac-tility and are load independent [59]. Furthermore, 12 weeks of
16 Diabetologia (2017) 60:7–23
HIIT in heart failure patients led to a 22% increase in globalcontractility [60]. These improvements were not observedfollowing energy matched MICT [60].
Cardiac torsion describes the twisting motion of the heartduring contraction and reflects the dominance of epicardialfibres over endocardial fibres. In adults with metabolic dis-ease cardiac torsion is raised [62], reflecting damage toendocardial fibres. Interestingly, we observed reductionsin cardiac torsion in adults with type 2 diabetes andNAFLD who partook in 12 weeks of HIIT, when comparedwith controls [38, 39], suggesting a reduction in endocar-dial damage following HIIT.
Diastolic function Diastolic dysfunction is often reportedin those with common metabolic diseases [57, 63].Impaired early filling of the left ventricle is indicative ofstiffer, damaged myocardial fibres that are less compliantduring relaxation; yet evidence suggests that HIIT has thecapacity to target these abnormalities. Two studies havedemonstrated significant improvements in early fillingrates following 12 weeks of HIIT in adults with type 2diabetes [38, 44], which were sustained 1 year later [44].Similar diastolic improvements were also observed inadults with NAFLD [39]. These HIIT-induced elevationsin early filling rate have been demonstrated to be as largeas 49%. In contrast, 12 weeks of MICT fails to have anyimpact upon diastolic variables [44, 59, 60]. These datasuggest that exercise intensity is an important characteris-tic for inducing diastolic improvements. Diastolic dys-function is an independent predictor of mortality [64],therefore any improvements in function are likely to beclinically significant.
Vascular function
Endothelial dysfunction is associated with metabolic dis-ease [65] and considered one of the earliest pathophysio-logical processes in the progression to atherosclerosis.Flow mediated dilation (FMD) is a measure of endothelialdysfunction and is regulated by NO availability. In thosewith common metabolic disease, HIIT has been shown tobe superior [44] or similar [35] to MICT for improvingFMD. Although not limited to common metabolic dis-eases, a meta-analysis of 182 participants demonstratedtwice the improvement in FMD following HIIT, comparedwith MICT [66]. This is most likely due to the greatershear stress experienced during higher-intensity exercise,since shear stress is the main stimuli for increasing NOavailability in the endothelium [59]. Consequently, im-proved FMD results in greater perfusion and oxygen sup-ply to peripheral tissue.
Findings with respect to the effect of HIIT on bloodpressure in individuals with common metabolic diseases
have been inconsistent; some studies demonstrate im-provements [11, 20, 35, 45], whereas some show nochange [36, 38, 39, 44] in blood pressure, despite pos-itive cardiac remodelling [38]. Exercise guidelines forthe treatment of hypertension advise low- to moderate-intensity exercise [67], but these findings suggest furtherwork is required to better define the role of HIIT inhypertension therapy.
Skeletal muscle and cardiac adaptations combine toimprove V
:O2peak following HIIT
It could be argued that the most important outcome fol-lowing HIIT is cardiorespiratory fitness, as measured by
V:O2peak. Large prospective studies have demonstrated fit-
ness to be more important than established risk factors formortality [68], and low V
:O2peak is independently associat-
ed with incident type 2 diabetes [69]. While the exercise-induced increase in V
:O2peak has never been directly linked
to mortality, large scale studies indicate that improvementsin fitness over time leads to significant reductions in mor-tality risk [70, 71].
V:O2peak is the gold standard measure of fitness and a strong
indicator of how well the cardiac, pulmonary, vascularand peripheral systems are working together. A numberof meta-analyses have demonstrated the substantial bene-fits of HIIT for V
:O2peak and its superiority in comparison
to MICT in healthy [72], coronary artery disease [73] andcardiometabolic disease patients [8]. In those with elevat-ed cardiometabolic risk, the increase in V
:O2peak with HIIT
(19.4%) was almost twice that of MICT (10.3%) [8]. Onaverage, V
:O2peak increases by 5.4 ml kg−1 min−1 follow-
ing HIIT, and even a smal le r improvement of3.5 ml kg−1 min−1 has been predicted to improve survivalby 10–25% [74].
Figure 2 provides an overview of the skeletal muscleand cardiac adaptations that are likely to contribute to theimprovements in V
:O2peak observed with HIIT. As dem-
onstrated in the figure, HIIT improves the capacity ofboth aspects of the oxygen supply and demand chain,but it is the cardiovascular adaptations in response toHIIT that are more likely to contribute to these V
:O2peak
improvements [75].
HIIT and weight loss
HIIT induces moderate weight loss (0.5–4 kg reduction)in adults with common metabolic diseases [11, 20, 34,36, 41, 46]. When compared with MICT, however, HIIT
Diabetologia (2017) 60:7–23 17
provides no additional benefit as an exercise therapy forweight loss [7]. The ability of HIIT to induce reductionsin body weight should therefore not be overstated inthose with common metabolic diseases.
Although weight loss is strongly associated with re-duced metabolic complications [76], it does not reflectchanges in body composition; HIIT generally reduceswhole body fat mass by 1–3 kg, even when body weightremains stable [14, 35, 36, 41, 46, 77, 78]. Significantreductions in visceral and hepatic fat have also beenshown with HIIT [14, 38, 39]. These findings are impor-tant since these fat depots increase cardiovascular diseaserisk [79], and metabolic dysfunction [40, 80]. Three pos-sible mechanisms for HIIT-induced fat loss have beensuggested:
1. increased mitochondrial density and capacity fol-lowing HIIT leading to increased fat oxidation[81]
2. large elevations in catecholamines, which have beenshown to drive lipolysis [82], especially in the ab-dominal tissue where there are significantly more β-adrenergic receptors, compared with subcutaneousfat [83]
3. appetite suppression: energy intake the day after HIITwas∼1255 kJ (300 kcal) lower than after MICT, and ∼2510 kJ(600 kcal) lower than after rest [84]
It remains unclear whether HIIT is superior to MICTfor fat loss, with some studies supporting this notion [14,36, 46] and some not [34, 35, 41, 45, 78]. To date, evi-dence to support HIIT over other types of exercise for themanagement of body fat levels is unfounded, but there isenough proof to suggest that HIIT can induce positivechanges in body composition in adults with common met-abolic diseases.
A summary of the effects of HIIT can be found in the textbox ‘Summary of HIIT’.
Is HIIT safe?
Given the strong cardiovascular-focused physiological re-sponse to HIIT, it is appropriate to define the safety ofhigh-intensity activity in those at elevated cardiometabolicrisk. The acute cardiac response to HIIT has beenassessed in a few studies. In patients with coronary heartdisease, no contraindications to HIIT were observed andundesirable changes, such as ST-segment depression, re-covered during interval recovery periods [85, 86]. Also, inpatients with chronic heart failure, cardiac stress (asassessed by rate pressure product) stayed within accept-able values [87]. Furthermore, the studies mentionedabove did not report any serious adverse events withHIIT.
The largest available dataset assessing the safety ofHIIT was derived from a clinical audit of 4846 cardiacrehabilitation patients. It identified two non-fatal cardiacarrests in 46,364 h of supervised HIIT, and one fatalcardiac arrest in 129,456 h of supervised MICT [88].Although the low frequency of events makes the com-parison between the two exercise modalities inconclu-sive, it also highlights that the risk of either approachis low. It is important to note that all patients werereferred to cardiac rehabilitation by their general practi-tioner or hospital cardiologist and underwent a full med-ical screening and cardiopulmonary exercise test prior totaking part, to rule out recurrent ischaemia or chest painduring exercise.
The risk of sudden cardiac death and acute myocardialinfarction is increased following vigorous activity in suscep-tible individuals, including those with structural heart diseaseand congenital complications [89]. The American College ofSports Medicine and the American Heart Association provideguidelines for identifying high risk patients and carrying outpre-exercise screening in such individuals [6, 89]. Accordingto these guidelines those with common metabolic diseases areautomatically considered ‘high risk’. On the whole, however,
Summary of HIIT
1 Leads to modest improvements in metabolic control, of similar magnitude to other forms of exercise training
2 Should not be overstated for its role in weight loss
3 Has strong cardiovascular benefits
4 Leads to large improvements in cardiorespiratory fitness, often superior to other forms of exercise training
18 Diabetologia (2017) 60:7–23
mounting clinical evidence supports HIITas a safe therapy forthe majority of individuals with elevated cardiometabolic risk.
Tolerability of HIIT in patients
The trials published to date illustrate the tolerability ofHIIT among diverse clinical populations and with varyingstudy durations (see Table 1). Although large-scale trialsare lacking, attempts have been made to assess the palat-ability of HIIT in previously sedentary populations. Agroup of obese women, some with type 2 diabetes, werenoted to prefer a HIIT approach to MICT [90], as didvolunteers with coronary heart disease [85]. Within HIITprotocols, enjoyment decreases with increasing intervallength [91]. Specifically, intervals of 30 or 60 s resultedin greater enjoyment than 120 s intervals.
Good adherence to free-living, non-supervised HIIT(<3 months) has been demonstrated in those with type 2 dia-betes [38], NAFLD [39] and those with either impaired glu-cose tolerance or impaired fasting glucose [92]. Good adher-ence was also observed with interval walking (3 min alterna-tive fast and slow walking) in a free-living environment over4 months in patients with type 2 diabetes [36] and over22 months in older adults [93]. Additional longer-term studiesare required but, nonetheless, these results indicate good ad-herence and tolerability to independent HIIT exercise.
Considerations when prescribing HIIT
Beyond the plethora of specific protocols to choose from,the way in which HIIT protocols are often described is, initself, a barrier to clinical implementation. The text box,‘Recommendations for HIIT prescription’ provides a sum-mary of our recommendations for prescribing HIIT in aclinical setting.
In research, HIIT is most commonly carried out asthree sessions per week (Table 1). Such a frequency isconsistent with the probable duration of the metaboliceffects observed. As previously mentioned, the durationof intervals also varies from 1–4 min (see Table 1).Since longer intervals have not been conclusively shownto yield better clinical outcomes but have been shown toreduce enjoyment [91], it makes sense to start at theshorter end of the range. The accumulated time at high-intensity during HIIT has varied from 10–20 min; startingat the lower end of this range allows for greater progres-sion. Likewise, ratios of interval:recovery time also vary,with a 1:1 ratio offering a simple starting point. Last,keeping the intensity of the recovery period to a minimumis likely to increase enjoyment, at least initially.
Intensity is commonly measured using HRmax. At firstglance, heart rate appears like a feasible option given the rel-ative ubiquity of heart rate monitors. However, heart rate risesacross intervals, as shown in Fig. 1. In our research, we haveadopted a very practical approach, using the Rate of PerceivedExertion (RPE) 6–20 Borg scale as a guide of intensity; aspreviously reported, participants were asked to work at a16–17 on the scale (or ‘very hard’) during each interval(Fig. 1) [38, 39]. RPE is an accurate predictor of exerciseintensity in diabetes patients [94], however, like heart rate, itdoes have its limitations. Agreement between RPE and moreobjective measures of intensity is known to suffer both inter-and intra-individual variation. For example, when RPEwas assessed during a set workload protocol, RPE in-creased from the first to the last interval [12]. Thus, usingRPE trades some objectivity, but the benefit is a great dealof practicality.
Exercise selection is ultimately limited by what is available
to the patient. Since peripheral metabolic adaptions are limited
to the muscles undergoing forceful contractions during exer-
cise [29], it is preferable to choose activities involving a large
muscle mass.
Recommendations for HIIT prescription
Frequency: 3 HIIT sessions per week
Intensity: Most easily measured by rating of perceived exertion (although may be variable in practice)
Time: Intervals should last between 1 and 4 minutes, with intervals at the shorter end being preferred by patients. The total time spent doing intervals should be 10–20 minutes per session
Type: Activities involving a large muscle mass
Diabetologia (2017) 60:7–23 19
What’s next for HIIT?
HIIT leads to modest improvements in metabolic control andweight loss. This is in contrast to calorie restriction, whichleads to significant weight loss and improvements in metabol-ic control [40]. Combining HIITwith calorie restriction wouldaccrue the cardiac benefits of HIITand the weight loss benefitsof calorie restriction. Additionally, exercise and calorie restric-tion together have been shown to improve glucose regulationby two-fold compared with the same amount of weight lossinduced by exercise or calorie restriction alone [95]. Thus,there may be additive benefits for metabolic control if HIITwas used adjunct with energy restriction.
The myriad of different HIIT protocols adopted in the lit-erature needs to be addressed. A standardised and consistentapproach for prescribing HIIT protocols is missing, making itdifficult to detect dose–response effects and the thresholdsnecessary to elicit desired changes. Most clinical HIIT studieshave been short term (<4 months, see Table 1) and performedin a laboratory setting. The feasibility, acceptability and effi-cacy of longer-term HIIT in a real world setting requires in-vestigation before it can be accepted as an alternative therapyfor those with elevated cardiometabolic risk.
Conclusion
In circumstances where HIIT is not feasible, considered poten-tially unsafe or not well tolerated by an individual , MICT iseffective at eliciting important health benefits. However,throughout this review we have shown that, provided unstablecardiovascular disease is excluded, HIIT appears to have a goodsafety profile and is well tolerated. Compared with other formsof exercise training, the use of HIIT for improving metaboliccontrol and inducing weight loss should not be overstated.However, there are strong positive cardiovascular adaptationsto HIIT that confer benefit to a population at risk of cardiaccomplications and therein lies the importance of HIIT for meta-bolic disease management. For optimal clinical benefit (im-proved glycaemic control and cardiovascular function), the valueof HIIT appears likely to be adjunct to energy restriction,allowing HIIT to certainly make a hit.
Acknowledgements We thank L. Taylor (MoveLab, NewcastleUniversity, Newcastle, UK) for assistance with creating the images forthis manuscript.
Funding MITwas supported by a Senior Fellowship from the NationalInstitute for Health Research.
Duality of interest The authors declare that there is no duality of inter-est associated with this manuscript.
Contribution statement All authors were involved in drafting the ar-ticle and revising it critically for important intellectual content, and ap-proving the final version for publication.
Open Access This article is distributed under the terms of theCreative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,distribution, and reproduction in any medium, provided you give appro-priate credit to the original author(s) and the source, provide a link to theCreative Commons license, and indicate if changes were made.
References
1. Cordain L, Gotshall R, Eaton S, Eaton S (1998) Physical Activity,Energy Expenditure and Fitness: An Evolutionary Perspective. Int JSports Med 19:328–335
2. WHO (2014) The top 10 causes of death. Fact sheet no. 310.WHO,Geneva
3. Inzucchi SE, Bergenstal RM, Buse JB et al (2015) Management ofhyperglycaemia in type 2 diabetes, 2015: a patient-centered ap-proach. Update to a position statement of the American DiabetesAssociation and the European association for the Study of Diabetes.Diabetologia 58:429–442
4. Jensen MD, Ryan DH, Apovian CM et al (2014) 2013 AHA/ACC/TOS guideline for the management of overweight and obesity inadults: a report of the American College of Cardiology/AmericanHeart Association Task Force on Practice Guidelines and TheObesity Society. Circulation 129:S102–S138
5. Colberg SR, Sigal RJ, Fernhall B et al (2010) Exercise and type 2diabetes: the American College of Sports Medicine and theAmerican Diabetes Association: joint position statement executivesummary. Diabetes Care 33:2692–2696
6. ACSM (2014) ACSM’s guidelines for exercise testing and prescrip-tion, 9th edn. Wolters Kluwer/Lippincott Williams & Wilkins,Philadelphia
7. Jelleyman C, Yates T, O’Donovan G et al (2015) The effects ofhigh-intensity interval training on glucose regulation and insulinresistance: a meta-analysis. Obes Rev 16:942–961
8. Weston KS, Wisløff U, Coombes JS (2014) High-intensity intervaltraining in patients with lifestyle-induced cardiometabolic disease: asystematic review andmeta-analysis. Br J Sports Med 48:1227–1234
9. Gibala MJ, Little JP, Macdonald MJ, Hawley JA (2012)Physiological adaptations to low-volume, high-intensity intervaltraining in health and disease. J Physiol 590:1077–1084
10. Burgomaster KA, Hughes SC, Heigenhauser GJF et al (2005) Sixsessions of sprint interval training increases muscle oxidative po-tential and cycle endurance capacity in humans. J Appl Physiol 98:1985–1990
11. Fex A, Leduc-Gaudet J-P, Filion M-E et al (2015) Effect of ellipticalhigh intensity interval training on metabolic risk factor in pre- andtype 2 diabetes patients: a pilot study. J Phys Act Health 12:942–946
12. Little JP, Gillen JB, Percival ME et al (2011) Low-volume high-intensity interval training reduces hyperglycemia and increasesmuscle mitochondrial capacity in patients with type 2 diabetes.J Appl Physiol 111:1554–1560
13. Ren JM, Semenkovich CF, Gulve EA et al (1994) Exercise inducesrapid increases in GLUT4 expression, glucose transport capacity,and insulin-stimulated glycogen storage in muscle. J Biol Chem269:14396–14401
14. Karstoft K, Winding K, Knudsen SH et al (2014) Mechanismsbehind the superior effects of interval vs continuous training on
20 Diabetologia (2017) 60:7–23
glycaemic control in individuals with type 2 diabetes: a randomisedcontrolled trial. Diabetologia 57:2081–2093
15. Ritov VB, Menshikova EV, Azuma K et al (2010) Deficiency ofelectron transport chain in human skeletal muscle mitochondria intype 2 diabetes mellitus and obesity. Am J Physiol EndocrinolMetab 298:E49–58
16. Schrauwen-Hinderling VB, Kooi ME, Hesselink MKC et al (2007)Impaired in vivo mitochondrial function but similar intramyocellularlipid content in patients with type 2 diabetes mellitus and BMI-matched control subjects. Diabetologia 50:113–120
17. Mootha VK, Lindgren CM, Eriksson K-F et al (2003) PGC-1α-responsive genes involved in oxidative phosphorylation are coordi-nately downregulated in human diabetes. Nat Genet 34:267–273
18. Wu Z, Puigserver P, Andersson U et al (1999) Mechanisms control-ling mitochondrial biogenesis and respiration through the thermo-genic coactivator PGC-1. Cell 98:115–124
19. Little JP, Safdar A,WilkinGP et al (2010) A practical model of low-volume high-intensity interval training induces mitochondrial bio-genesis in human skeletal muscle: potential mechanisms. J Physiol588:1011–1022
20. Tjønna AE, Lee SJ, Rognmo Ø et al (2008) Aerobic interval train-ing versus continuous moderate exercise as a treatment for the met-abolic syndrome: A pilot study. Circulation 118:346–354
21. Daussin FN, Zoll J, Dufour SP et al (2008) Effect of interval versuscontinuous training on cardiorespiratory andmitochondrial functions:relationship to aerobic performance improvements in sedentary sub-jects. Am J Physiol Regul Integr Comp Physiol 295:R264–R272
22. Little JP, JungME,Wright AE et al (2014) Effects of high-intensityinterval exercise versus continuous moderate-intensity exercise onpostprandial glycemic control assessed by continuous glucosemon-itoring in obese adults. Appl Physiol Nutr Metab 39:835–841
23. Gillen JB, Little JP, Punthakee Z et al (2012) Acute high-intensityinterval exercise reduces the postprandial glucose response andprevalence of hyperglycaemia in patients with type 2 diabetes.Diabetes Obes Metab 14:575–577
24. Terada T, Wilson BJ, Myette-Cόté E et al (2016) Targeting specificinterstitial glycemic parameters with high-intensity interval exerciseand fasted-state exercise in type 2 diabetes. Metabolism 65:599–608
25. Karstoft K, Christensen CS, Pedersen BK, Solomon TPJ (2014)The acute effects of interval- vs continuous-walking exercise onglycemic control in subjects with type 2 diabetes: A crossover,controlled study. J Clin Endocrinol Metab 99:3334–3342
26. Tjønna AE, Rognmo Ø, Bye A et al (2011) Time course of endo-thelial adaptation after acute and chronic exercise in patients withmetabolic syndrome. J Strength Cond Res 25:2552–2558
27. Cavalot F, Pagliarino A, Valle M et al (2011) Postprandial bloodglucose predicts cardiovascular events and all-cause mortality intype 2 diabetes in a 14-year follow-up: lessons from the San LuigiGonzaga Diabetes Study. Diabetes Care 34:2237–2243
28. Ceriello A (2005) Postprandial hyperglycemia and diabetes com-plications: is it time to treat? Diabetes 54:1–7
29. Frøsig C, Richter EA (2009) Improved insulin sensitivity after ex-ercise: focus on insulin signaling. Obesity (Silver Spring) 17(Suppl3):S15–S20
30. Mikines KJ, Sonne B, Farrell PA et al (1988) Effect of physicalexercise on sensitivity and responsiveness to insulin in humans.Am J Physiol Endocrinol Metab 254:248–259
31. Gollnick PD, Piehl K, Saltin B (1974) Selective glycogen depletionpattern in human muscle fibres after exercise of varying intensityand at varying pedalling rates. J Physiol 241:45–57
32. America Diabetes Association (2014) Standards of Medical Care inDiabetes—2014. Diabetes Care 37(Suppl 1):S14–S80
33. Earnest CP, Lupo M, Thibodaux J et al (2013) Interval training inmen at risk for insulin resistance. Int J Sports Med 34:355–363
34. Madsen SM, Thorup AC, Overgaard K, Jeppesen PB (2015) Highintensity interval training improves glycaemic control and pancre-atic β cell function of type 2 diabetes patients. PLoS One 10:1–24
35. Stensvold D, Tjønna AE, Skaug E-A et al (2010) Strength trainingversus aerobic interval training to modify risk factors of metabolicsyndrome. J Appl Physiol 108:804–810
36. Karstoft K, Winding K, Knudsen S et al (2013) The effects of free-living interval-walking training on glycemic control, body compo-sition, and physical fitness in type 2 diabetic patients. Diabetes Care36:228–236
37. Shaban N, Kenno K, Milne K (2014) The effects of a 2 weekmodified high intensity interval training program on the homeostat-ic model of insulin resistance (HOMA-IR) in adults with type 2diabetes. J Sports Med Phys Fitness 54:203–209
38. Cassidy S, Thoma C, Hallsworth K et al (2016) High intensityintermittent exercise improves cardiac structure and function andreduces liver fat in patients with type 2 diabetes: a randomisedcontrolled trial. Diabetologia 59:56–66
39. Hallsworth K, Thoma C, Hollingsworth KG et al (2015) Modifiedhigh-intensity interval training reduces liver fat and improves car-diac function in non-alcoholic fatty liver disease: a randomizedcontrolled trial. Clin Sci (Lond) 129:1097–1105
40. Lim EL, Hollingsworth KG, Aribisala BS et al (2011) Reversal oftype 2 diabetes: normalisation of beta cell function in associationwith decreased pancreas and liver triacylglycerol. Diabetologia 54:2506–2514
41. Mitranun W, Deerochanawong C, Tanaka H, Suksom D(2014) Continuous vs interval training on glycemic controland macro- and microvascular reactivity in type 2 diabeticpatients. Scand J Med Sci Sports 24:e69–e76
42. ACSM (2005) ACSM’s resource manual for guidelines for exercisetesting and prescription, 5th edn. Lippincott Williams and Wilkins,Philadelphia
43. Simoneau JA, Colberg SR, Thaete FL, Kelley DE (1995) Skeletalmuscle glycolytic and oxidative enzyme capacities are determinantsof insulin sensitivity and muscle composition in obese women.FASEB J 9:273–278
44. Hollekim-Strand SM, Bjørgaas MR, Albrektsen G et al (2014) High-intensity interval exercise effectively improves cardiac function inpatients with type 2 diabetes mellitus and diastolic dysfunction: arandomized controlled trial. J Am Coll Cardiol 64:1758–1760
45. Robinson E, Durrer C, Simtchouk S et al (2015) Short-term high-intensity interval and moderate-intensity continuous training reduceleukocyte TLR4 in inactive adults at elevated risk of type 2 diabe-tes. J Appl Physiol 110:508–516
46. Terada T, Friesen A, Chahal BS et al (2013) Feasibility and prelim-inary efficacy of high intensity interval training in type 2 diabetes.Diabetes Res Clin Pract 99:120–129
47. Thomas D, Elliott E, Naughton G (2006) Exercise for type 2 diabetesmellitus (Review). Cochrane Database Syst Rev 19:CD002968
48. Johansen K (1999) Efficacy of metformin in the treatment ofNIDDM. Meta-analysis. Diabetes Care 22:33–37
49. Rafiq N, Bai C, Fang Yet al (2009) Long-term follow-up of patientswith nonalcoholic fatty liver. Clin Gastroenterol Hepatol 7:234–238
50. Garcia MJ, McNamara PM, Gordon T, Kannel WB (1974)Morbidity andmortality in diabetics in the Framingham population.Sixteen year follow-up study. Diabetes 23:105–111
51. Hasenfuss G (1998) Animal models of human cardiovascular dis-ease, heart failure and hypertrophy. Cardiovasc Res 39:60–76
52. Wisløff U, Ellingsen Ø, Kemi OJ (2009) High-intensity intervaltraining to maximize cardiac benefits of exercise training? ExercSport Sci Rev 37:139–146
53. Stølen TO, Høydal MA, Kemi OJ et al (2009) Interval trainingnormalizes cardiomyocyte function, diastolic Ca2+ control, andSR Ca2+ release synchronicity in a mouse model of diabetic cardio-myopathy. Circ Res 105:527–536
Diabetologia (2017) 60:7–23 21
54. Wisløff U, Loennechen JP, Currie S et al (2002) Aerobic exercisereduces cardiomyocyte hypertrophy and increases contractility,Ca2+ sensitivity and SERCA-2 in rat after myocardial infarction.Cardiovasc Res 54:162–174
55. Kemi OJ, Haram PM, Loennechen JP et al (2005) Moderate vs.high exercise intensity: differential effects on aerobic fitness, car-diomyocyte contractility, and endothelial function. Cardiovasc Res67:161–172
56. Zile MR, Gottdiener JS, Hetzel SJ et al (2011) Prevalence andsignificance of alterations in cardiac structure and function in pa-tients with heart failure and a preserved ejection fraction.Circulation 124:2491–2501
57. Rijzewijk LJ, van der Meer RW, Lamb HJ et al (2009) Alteredmyocardial substrate metabolism and decreased diastolic functionin nonischemic human diabetic cardiomyopathy: studies with car-diac positron emission tomography and magnetic resonance imag-ing. J Am Coll Cardiol 54:1524–1532
58. Frey N, Katus HA, Olson EN, Hill JA (2004) Hypertrophy of theheart: a new therapeutic target? Circulation 109:1580–1589
59. Molmen-Hansen HE, Stolen T, Tjonna AE et al (2012) Aerobicinterval training reduces blood pressure and improves myocardialfunction in hypertensive patients. Eur J Prev Cardiol 19:151–160
60. Wisløff U, Støylen A, Loennechen JP et al (2007) Superior cardio-vascular effect of aerobic interval training versus moderate contin-uous training in heart failure patients: a randomized study.Circulation 115:3086–3094
61. Coletta AP, Cleland JGF, Freemantle N, Clark AL (2004) Clinicaltrials update from the European Society of Cardiology Heart Failuremeeting: SHAPE, BRING-UP 2 VAS, COLA II, FOSIDIAL,BETACAR, CASINO and meta -ana lys i s o f ca rd iacresynchronisation therapy. Eur J Heart Fail 6:673–676
62. Lumens J, Delhaas T, Arts T et al (2006) Impaired subendocardialcontractile myofiber function in asymptomatic aged humans, asdetected using MRI. Am J Physiol Heart Circ Physiol 291:H1573–H1579
63. Goland S, Shimoni S, Zornitzki Tet al (2006) Cardiac abnormalitiesas a new manifestation of nonalcoholic fatty liver disease: echocar-diographic and tissue Doppler imaging assessment. J ClinGastroenterol 40:949–955
64. Halley CM, Houghtaling PL, Khalil MK et al (2011) Mortality ratein patients with diastolic dysfunction and normal systolic function.Arch Intern Med 171:1082–1087
65. Avogaro A, Albiero M, Menegazzo L et al (2011) Endothelial dys-function in diabetes: the role of reparatory mechanisms. DiabetesCare 34(Suppl 2):S285–S290
66. Ramos JS, Dalleck LC, Tjonna AE et al (2015) The impact of high-intensity interval training versus moderate-intensity continuoustraining on vascular function: a systematic review and meta-analy-sis. Sports Med 45:679–692
67. Pescatello LS, Franklin BA, Fagard R et al (2004) Exercise andhypertension. Med Sci Sports Exerc 36:533–553
68. Barry VW, Baruth M, Beets MW et al (2014) Fitness vs. fatness onall-causemortality: a meta-analysis. Prog Cardiovasc Dis 56:382–390
69. Sui X, Hooker SP, Lee I-M et al (2008) A prospective study ofcardiorespiratory fitness and risk of type 2 diabetes in women.Diabetes Care 31:550–555
70. Blair SN, Kohl HW, Barlow CE et al (1995) Changes in physicalfitness and all-cause mortality. A prospective study of healthy andunhealthy men. JAMA 273:1093–1098
71. Erikssen G, Liestøl K, Bjørnholt J et al (1998) Changes in physicalfitness and changes in mortality. Lancet 352:759–762
72. Milanović Z, Sporiš G, Weston M (2015) Effectiveness of high-intensity interval training (HIT) and continuous endurance trainingfor VO2max improvements: a systematic review and meta-analysisof controlled trials. Sports Med 45:1469–1481
73. Liou K, Ho S, Fildes J, Ooi S-Y (2016) High intensity intervalversus moderate intensity continuous training in patients with cor-onary artery disease: a meta-analysis of physiological and clinicalparameters. Heart Lung Circ 25:166–174
74. Kaminsky LA, Arena R, Beckie TM et al (2013) The importance ofcardiorespiratory fitness in the United States: the need for a nationalregistry: a policy statement from the American Heart Association.Circulation 127:652–662
75. Bækkerud FH, Solberg F, Leinan IM et al (2016) Comparison ofthree popular exercise modalities on VO2max in overweight andobese. Med Sci Sports Exerc 48:491–498
76. Sjöström CD, Lissner L, Wedel H, Sjöström L (1999) Reduction inincidence of diabetes, hypertension and lipid disturbances after in-tentional weight loss induced by bariatric surgery: the SOSIntervention Study. Obes Res 7:477–484
77. Heydari M, Freund J, Boutcher SH (2012) The effect of high-intensity intermittent exercise on body composition of overweightyoung males. J Obes 2012:480467
78. Fisher G, Brown AW, Bohan Brown MM et al (2015) High inten-sity interval- vs moderate intensity-training for improving cardio-metabolic health in overweight or obese males: a randomized con-trolled trial. PLoS One 10:1–15
79. Nakamura T, Tokunaga K, Shimomura I et al (1994) Contributionof visceral fat accumulation to the development of coronary arterydisease in non-obese men. Atherosclerosis 107:239–246
80. Fujioka S, Matsuzawa Y, Tokunaga K, Tarui S (1987) Contributionof intra-abdominal fat accumulation to the impairment of glucoseand lipid metabolism in human obesity. Metabolism 36:54–59
81. Talanian JL, Galloway SDR,Heigenhauser GJF et al (2007) Twoweeksof high-intensity aerobic interval training increases the capacity for fatoxidation during exercise in women. J Appl Physiol 102:1439–1447
82. Zouhal H, Jacob C, Delamarche P, Gratas-Delamarche A (2008)Catecholamines and the effects of exercise, training and gender.Sports Med 38:401–423
83. Rebuffé-Scrive M, Andersson B, Olbe L, Björntorp P (1989)Metabolism of adipose tissue in intraabdominal depots of nonobesemen and women. Metabolism 38:453–458
84. Sim AY, Wallman KE, Fairchild TJ, Guelfi KJ (2014) High-intensity intermittent exercise attenuates ad-libitum energy intake.Int J Obes 38:417–422
85. Guiraud T, Nigam A, Juneau M et al (2011) Acute responses tohigh-intensity intermittent exercise in CHD patients. Med SciSports Exerc 43:211–217
86. Guiraud T, Juneau M, Nigam A et al (2010) Optimization of highintensity interval exercise in coronary heart disease. Eur J ApplPhysiol 108:733–740
87. Meyer K, Samek L, Schwaibold M et al (1996) Physical responsesto different modes of interval exercise in patients with chronic heartfailure—application to exercise training. Eur Heart J 17:1040–1047
88. Rognmo Ø, Moholdt T, Bakken H et al (2012) Cardiovascular riskof high- versus moderate-intensity aerobic exercise in coronaryheart disease patients. Circulation 126:1436–1440
89. Thompson PD, Franklin BA, Balady GJ et al (2007)Exercise and acute cardiovascular events placing the risksinto perspective: a scientific statement from the AmericanHeart Association Council on Nutrition, Physical Activity,and Metabolism and the Council on Clinical Cardiology.Circulation 115:2358–2368
90. Coquart JBJ, Lemaire C, Dubart A-E et al (2008) Intermittent ver-sus continuous exercise: effects of perceptually lower exercise inobese women. Med Sci Sports Exerc 40:1546–1553
91. Martinez N, Kilpatrick MW, Salomon K et al (2015) Affective andenjoyment responses to high-intensity interval training inoverweight-to-obese and insufficiently active adults. J Sport ExercPsychol 37:138–149
22 Diabetologia (2017) 60:7–23
92. Jung ME, Bourne JE, Beauchamp MR et al (2015) High-intensityinterval training as an efficacious alternative to moderate-intensitycontinuous training for adults with prediabetes. J Diabetes Res2015:1–9
93. Masuki S, Mori M, Tabara Y et al (2015) The factors affectingadherence to a long-term interval walking training program inmiddle-aged and older people. J Appl Physiol 118:595–603
94. Colberg SR, Swain DP, Vinik AI (2003) Use of heart rate reserveand rating of perceived exertion to prescribe exercise intensity indiabetic autonomic. Diabetes Care 26:986–990
95. Weiss EP, Albert SG, Reeds DN et al (2015) Calorie restriction andmatched weight loss from exercise: independent and additive ef-fects on glucoregulation and the incretin system in overweightwomen and men. Diabetes Care 38:1253–1262
Diabetologia (2017) 60:7–23 23