Key words brown adipose tissue - infrared thermography - subcutaneous adipose tissue - visceral
adipose tissue
Introducion
The scientific literature is quite robust regarding the importance of adipose tissue
acting as an essential endocrine organ in the regulation of energy metabolism [1 ]. Adipose tissue can be classified according
to its morphological and functional characteristics into white adipose tissue (WAT),
beige adipose tissue (BeAT) and brown adipose tissue (BAT) [2 ]
[3 ]. It
is understood that most body adipose tissue is formed by WAT whose main function is
to store energy in the form of triglycerides [1 ]. WAT may also be called subcutaneous adipose tissue (SAT) when
deposited under the skin and visceral adipose tissue (VAT) when it is deposited
around the viscera [1 ]. BeAT originating from
WAT precursor cells can be defined as a cell with mixed characteristics of WAT and
BAT and localized in small amounts between WAT [3 ]. BAT, whose main function is the production of energy in the form of
heat, is deposited, in adults, mostly in the supraclavicular region (SCV) [2 ].
In the fight against the health problems caused by excess weight and possible
non-transmissible chronic diseases, BAT appears as an important element [4 ]
[5 ]
[6 ]. Through its thermogenic
activity, BAT can help control body weight and reduces risk factors for metabolic
diseases [4 ]
[7 ]. The oxidation of lipids to produce heat stimulated by cold or diet in
BAT cells is caused by the high intensity of mitochondrial and consequently
uncoupling protein-1 (UCP-1) [8 ]. The
uncoupling between oxygen consumption and ATP synthesis induces the energy
dissipation in the form of heat, thus helping to reduce glucose and lipid levels
and, consequently, weight loss [7 ]
[8 ]
[9 ]. It
is important to note that, like WAT, BAT is found in greater amounts in women than
in men [10 ]
[11 ]. Herz et al. [11 ] pointed out
that BAT activation by cold stimulation and PET-CT was greater, but not significant,
in premenopausal women than in men.
Analyzing BAT activation did not seem like such an easy task, because it used to
require examination of data obtained with positron emission tomography (PET-CT)
using 18F-fluordeoxyglucose (FDG), which made it difficult. After all, it is an
expensive and time-consuming test, with a high rate of exposure to radiation [12 ]. On the other hand, with the evolution of
infrared thermography (IRT), evaluating the BAT became simpler, because by measuring
the skin temperature (Tskin) in the SCV region, it is possible to estimate the
activation of the BAT in less time, with low risk and good accuracy [12 ]
[13 ].
BAT activation is inversely correlated with body mass index (BMI), percentage of
total body fat (%BF) and VAT, proving that somehow WAT is negatively
associated with non-shivering thermogenesis [14 ]
[15 ]. A recent survey analyzed
4852 patient scans performed with PE-CT-FDG and found that VAT and SAT were
independent factors related to BAT [15 ].
However, this same study concluded that VAT, age and sex were more important in BAT
activity under thermoneutral conditions [15 ].
Despite these findings, to our knowledge there are no studies that provide joint
information on the possible effects of SAT and VAT, using IRT, on the physiological
functioning of BAT cells when exposed to cold in humans. It is still important to
emphasize that research involving BAT and IRT in only groups of women, who notably
have more adipose tissue than men, is scarce [16 ]
[17 ]
[18 ]. It is hypothesized in this study that VAT
is more closely related to BAT activation, since it eliminates more inflammatory
cytokines when compared to SAT, which may cause some alteration in the non-shivering
thermogenesis process.
Thus, the presented study aimed to verify the relation of SAT and VAT on BAT
activation through IRT and cold stimulation in adult women. This study is justified
by the need for further clarification as to which type of adipose tissue can be more
harmful to the BAT activation process, negatively influencing body weight loss.
Methods
Study population
The present research included a total of 40 healthy female members of the
military divided into two groups using the median of variables: %SAT
(Over vs. Under) and %VAT (Over vs. Under). Female active-duty military
personnel aged between 27 and 56 years were included. Every subject who
underwent surgery in the entire trunk region and in menopause was excluded. This
study was designed according to Ethics Commission approved by Doc103-CE-202 and
CAAE:40495120.9.0000.9433, in agreement with the Declaration of Helsinki [19 ].
Data collection
On the day before the beginning of the data collection, the subjects were
instructed to: 1) remain fasting for 12 hours, 2) not perform physical exercises
24 hours before the collection, 3) not shave the clavicular and sternum regions,
4) not consume alcohol 24 h beforehand, 5) avoid direct sun or UV, and 6) avoid
any medication with a thermogenic substance. On the day of collection, the
subjects could not wear accessories next to the body, not ingest caffeinated
substances, not use cosmetic products and not smoke [20 ]. All data were collected in the morning
starting from 7:30 am to 11:30 am, between the months of February (summer) and
June (autumn) of 2022.
Body composition by anthropometry
Applying the protocol standardized by the International Society for the
Advancement of Kinanthropometry (ISAK), total body mass (BM), height, and body
mass index (BMI) was estimated by a physical education professional [21 ]. For this purpose, a digital scale
model P150M (Lider, São Paulo, Brazil) and a stadiometer model SN ES2030
(Sanny, São Bernardo do Campo, São Paulo, Brazil) were used.
Body composition by dual X-ray absorptiometry (DXA)
This exam was performed by a radiology technician following the protocols defined
by the manufacturer, using a DXA, iLunar model, from GE Healthcare (GE
Healthcare, Madison, WI, USA), with the enCore 2015 software (version
14.10.022). After the scanning process, the variables of total fat mass (FM),
percentage of total body fat (%BF), VAT (g), and fat mass index (FMI)
produced by the equipment were used for analysis [22 ]
[23 ]
[24 ]. Using the FM and VAT
values, the percentage of VAT (% VAT) was estimated. The percentage of
SAT (% SAT) was also calculated, decreasing %BF by %VAT.
The equipment was calibrated before starting the daily assessments.
IRT image acquisition
For the analysis of thermal images, a 3.0 × 2.5-m room with control of
air temperature (22.0±0.5°C) and relative air humidity
(55.0±1.2%) was used. Also, a FLIR E75 camera (FLIR Systems,
Inc., Wilsonville, OR, USA) ([Fig. 1a ]),
with a thermal resolution of 320 × 240 pixels and a 4-inch LCD screen,
spectral range of 7.5–14.0 µm, a 180-pixel rotating lens
platform, and 0.98% emissivity and fixed into a tripod was operated. All
preparation and collection follow the recommended protocol for thermographic
imaging in sports and exercise medicine (TISEM) [20 ]. After 15 minutes of acclimatization, the IRT images were
captured in two moments: 1) right after the end of acclimatization (T0); and 2)
right after the cold stimulation (T5) protocol (
[Fig. 1b ]), which consisted of keeping both
hands in a bowl of water at 13.0±0.6 ºC for 5 min [25 ].
Fig. 1 (a ) Camera FLIR E75 positioned for acquisition
before cold protocol; (b ) immersion of both hands in a container
of cold water for 5 min.
In this research, IRT images of some regions of interest (ROIs) were acquired: 1)
mean Tskin in the right SCV (SCVTR) and mean Tskin in the left SCV (SCVTL) for
BAT activation analysis, and 2) mean Tskin in the sternum as the control region
(CTRLT). The same ROI identification protocol created by da Rosa et al. [25 ] was applied ([Fig. 2 ]).
Fig. 2 Regions of Interest (ROIs) evaluated. SCVTR: right mean
supraclavicular skin temperature, SCVTL: left mean supraclavicular skin
temperature, CTRLT: mean control skin temperature.
The women were positioned sitting on a chair perpendicular to the thermographic
camera, one meter away, keeping their heads in the Frankfurt plane, wearing a
top t-shirt without a shoulder strap. Water temperature were monitored by a
hygrometer (HTC-1, Rio de Janeiro, RJ, Brazil). The images were captured using
the FLIR ResearchIR professional Analyzing software (version 1.5, FLIR
Systems).
Statistical analysis
First, the Shapiro–Wilk test was performed to check the data
distribution. Thereby, a parametric analysis was applied. Descriptive statistics
were applied to characterize the sample using mean and standard deviation. The
Student’s independent t-test was applied to compare descriptive data.
Pearson’s correlation between Tskin, %SAT, and %VAT was
used to assess the relationship between variables. Two-way repeated measures
ANOVA with Tukey’s post-hoc comparisons (Over vs. Under) were conducted
to assess group and moment factors effects of Tskin ROIs (T0 vs. T5). A
p -value<0.05 was considered significant. Data analysis was
conducted using the Jamovi statistical software, version 2.3.16.
Results
The descriptive data of the entire sample point out that the sample had an average
age=36.1±6.6 years, BM=66.1±13.4,
height=164.6±6.2 cm, BF=23.3±8.2 kg,
%BF=34.7 %, SAT=22.9±7.9 kg,
%SAT=33.3±5.5, VAT=0.36 kg,
%VAT=1.4%, BMI=24.4±3.9 kg/m2 and
FMI=8.6±2.7 kg/m2 . [Table 1 ] shows the descriptive data from all samples extracted by type of
fat (%SAT and %VAT) and points out that subjects classified as Over
the amount of adipose tissue have worse health indicators than those classified as
Under.
Table 1 Data of age, anthropometrics, body composition
(n=40 women) extracted Over and Under %SAT and Over and
Under %VAT.
Parameters
Classification by SAT and VAT
%SAT (n=20) Mean±SD
p value*
%VAT (n=20) Mean±SD
p value*
Age
Over
36.1±6.6
0.739
37.9±6.7
0.007*
Under
36.0±6.8
34.2±6.1
Height (cm)
Over
164.2±5.7
0.643
164.0±7.2
0.513
Under
165.2±6.9
165.3±5.2
BM (kg)
Over
69.6±14.7
0.153
71.4±16.0
0.010*
Under
62.6±11.4
60.7±7.2
FM (kg)
Over
27.5±8.5
0.001*
27.3±9.2
0.001*
Under
19.1±5.2
19.2±4.1
%BM
Over
38.9±4.0
<0.001*
37.5±5.1
<0.001*
Under
30.3±4.1
31.7±5.2
SAT (kg)
Over
27.0±8.2
0.001*
26.7±9.0
0.001*
Under
18.8±5.0
19.1±4.1
%SAT
Over
37.3±3.5
<0.001*
35.4±5.0
0.009*
Under
29.1±3.6
31.0±5.1
VAT (kg)
Over
472.9±336.1
0.024*
589.6±275.3
<0.001*
Under
254.6±228.3
137.9±83.6
%VAT
Over
1.6±1.0
0.131
2.1±0.6
<0.001*
Under
1.2±0.8
0.7±0.3
FMI (kg/m2 )
Over
10.1±2.7
0.032*
10.4±2.8
<0.001*
Under
6.9±1.5
7.1±1.4
BMI (kg/m2 )
Over
25.9±4.3
<0.001*
26.5±4.2
<0.001*
Under
22.9±2.8
22.2±1.9
BM: total body mass, FM: total fat mass, %BF: percentage total body
fat, SAT: subcutaneous adipose tissue, %SAT: percentage subcutaneous
adipose tissue, VAT: visceral adipose tissue, %VAT: percentage
visceral adipose tissue, FMI: fat mass index, BMI: body mass index,
* p value<0.05 obtained by Student’s
independent t-test.
[Table 2 ] makes the descriptive data of the
average temperatures within each group quite clear. It is noted that, in both
extracts (%SAT and %VAT), only the subjects classified with the
lowest amount of fat had a significant increase in SCV skin temperatures after
applying the cold stimulation protocol. Sternum ROI did not show a Tskin
increase.
Table 2 Skin temperature before and after cold stimulation
(n=40 women)
%SAT (n=20)
Tskin
Before (T0)
After (T5)
Over
SCVTR
33.4±0.9
33.7±0.8
SCVTL
33.3±0.7
33.6±0.8
CTRLT
31.1±1.0
31.1±0.8
Under
SCVTR
33.8±0.6
34.5±1.1*
SCVTL
33.9±0.7
34.6±0.9*
CTRLT
31.6±0.6
32.1±1.5
%VAT (n=20)
Tskin
Before (T0)
After (T5)
Over
SCVTR
33.4±0.8
33.6±0.7
SCVTL
33.5±0.8
33.7±0.8
CTRLT
31.2±0.9
31.2±0.2
Under
SCVTR
33.7±0.7
34.5±1.1*
SCVTL
33.8±0.7
34.5±1.0*
CTRLT
31.5±0.1
32.0±1.6
%SAT: percentage subcutaneous adipose tissue, %VAT:
percentage visceral adipose tissue, Tskin: skin temperature, T0: before cold
stimulation, T5: after cold stimulation, SCVTR: right mean supraclavicular
skin temperature, SCVTL: left mean supraclavicular skin temperature, CTRLT:
men control skin temperature. *p value obtained by Tukey
In the application of the Pearson’s correlation between SCVT, %SAT,
and %VAT, medium negative degrees of linear variation were found before and
after cold stimulation ([Table 3 ]). However,
the same significance was not found when the values were compared
with the delta (∆) of SCVT ([Table
3 ]).
Table 3 Correlation between skin temperature and %SAT
and %VAT, before and after cold stimulation (n=40
women)
SCVTR(T0)
SCVTL(T0)
SCVTR(T5)
SCVTL (T5)
∆TSCVR
∆TSCVL
%SAT (n=20)
r=–0.359
r=–0.501
r=–0.371
r=–0.407
r=–0.086
r=–0.006
p=0.023*
p<.001*
p=0.018*
p=0.009*
p=0.598
p=0.969
%VAT (n=20)
r=–0.376
r=–0.342
r=–0.442
r=–0.443
r=–0.034
r=–0.006
p=0.017*
p=0.031*
p=0.004*
p=0.004*
p=0.834
p=0.971
%SAT: percentage subcutaneous adipose tissue, %VAT:
percentage visceral adipose tissue, T0: before cold stimulation, T5: after
cold stimulation, SCVTR: right mean supraclavicular skin temperature, SCVTL:
left mean supraclavicular skin temperature, CTRLT: men control skin
temperature. ∆TSCVR=SCVTR(T0)–SCVT(T5);
∆TSCVL=SCVTL(T0)– SCVTL(T5); Pearson’s
correlation *p<0.005 value significance.
The 2-way repeated measurements ANOVA identified a significant effect of the moment
factor (BAT activation) and the group factor (Over vs. Under) on the SCVTR and SCVTL
between subjects divided in the two extracts (%SAT and %VAT) ([Table 4 ]). In intragroup analysis after the
cold stimulation protocol was applied, we noticed that only the Under %SAT
and Under %VAT groups had a significant increase in SCVTR and SCVTL. The
same trend was observed in the intergroup analysis where the SCV Tskin averages of
the group with Under %SAT and Under %VAT were significantly higher
than subjects classified with Over %SAT and Over %VAT. Unexpectedly,
no significant effect of the interaction of the factors was found ([Table 4 ]). Otherwise, there was only a group
factor effect on CTRLT ([Table 4 ]). When we
analyzed the size of the Tskin effects after cold stimulation for moment factor and
group factor with a significant increase, four parameters were classified with a
“medium” and four with a “large” effect size ([Table 4 ]).
Table 4 ANOVA two-way repeated sample (group × moment
factor) for skin temperature and groups by %SAT (n=20)
and by %VAT (n=20), before and after cold
stimulation
%SAT vs. BAT
%VAT vs. BAT
Tskin
Source
df
MS
F
p
η2
df
MS
F
p
η2
SCVTR
Moment Factor (BAT Activation)
1.00
5.25
8.34
0.006*
0.073
Moment Factor (BAT Activation)
1.00
5.23
8.63
0.006*
0.073
Group Factor (%SAT)
1.00
7.50
8.38
0.006*
0.105
Group Factor (%VAT)
1.00
7.53
8.38
0.006*
0.105
Interaction
1.00
0.90
1.43
0.239
0.036
Interaction
1.00
1.71
2.81
0.102
0.024
SCVTL
Moment Factor (BAT Activation)
1.00
4.73
0.04
0.005*
0.072
Moment Factor (BAT Activation)
1.00
4.73
9.44
0.004*
0.072
Group Factor (%SAT)
1.00
12.40
16.8
<0.001*
0.189
Group Factor (%VAT)
1.00
5.46
5.92
0.020*
0.083
Interaction
1.00
0.43
0.828
0.369
0.007
Interaction
1.00
1.27
2.53
0.120
0.019
CTRLT
Moment Factor (BAT Activation)
1.00
1.48
1.88
0.179
0.015
Moment Factor (BAT Activation)
1.00
1.48
1.85
0.182
0.015
Group Factor (%SAT)
1.00
10.88
7.82
0.008*
0.112
Group Factor (%VAT)
1.00
6.44
4.27
0.066
0.101
Interaction
1.00
1.71
2.16
0.149
0.054
Interaction
1.00
1.27
1.59
0.215
0.040
SCVTR: right mean supraclavicular skin temperature, SCVTL: left mean
supraclavicular skin temperature, CTRLT: mean control skin temperature,
%SAT: percentage subcutaneous adipose tissue, %VAT:
percentage visceral adipose tissue,
The presented results can help those in clinical practice understand more clearly,
through IRT, that the high accumulation of SAT and VAT can impair the activation of
BAT and, consequently, hinder the loss of body weight in women.
Discussion
This research aim of the present study was to verify the relation of SAT and VAT on
BAT activation through IRT and cold stimulation in adult women. The main findings
showed that Tskin SCV, right and left, despite small differences follow the same
trend of relation with SAT and VAT levels. This indicates that the higher the level
of body adipose tissue, the lower the increase in Tskin SCV, inferring the inverse
relation between activation of BAT, SAT, and VAT.
The scientific literature points out that there is an inverse correlation between
BAT
activity and BF, and subjects classified with positive BAT activity were female,
younger, and with a normal/overweight BMI when compared to subjects with
negative BAT [26 ].
This finding is in line with our results, which found that the group of women with
a
higher BMI and %BF were those with lower BAT activation. Matsushita et al.
[27 ] confirmed that body fat mass, BMI,
and abdominal fat area were lower in BAT-positive subjects than in BAT-negative
subjects. Our study, despite using another BAT detection technique (IRT) but
maintaining the cold stimulus, found the same trend in women with greater increases
in SCVT. Mean Tskin in the SCV region in women, after cold stimulation in our study,
was slightly higher than that found by da Rosa et al. [25 ] using the same protocol in men. However, in
the study by Hartwig et al. [11 ] using IRT,
after immersing the left hand in water at a temperature of 5 to 9°C for 1
minute, overweight women had lower SCVTL increases than those with normal weight,
corroborating our results where women with lower %SAT and %VAT also
had greater increases.
The scarcity of studies involving analysis of thermal images and BAT activation with
cold stimulation and associations between SAT and mainly VAT make the discussion
somewhat unclear, as we should draw on studies that use PET-CT as a standard method
for comparison [14 ]
[15 ]
[28 ].
However, analyzing the correlations found in our paper between %SAT,
%VAT, and SCVT, right and left, they are consistent with previous research.
For example, Gatidis et al. [14 ] evaluated BAT
activation in 120 patients (44 female, 59 male) and found a significant inverse
correlation between mean SCVT and %SAT (r=–0.65,
p<0.001). In a study with 86 prepubertal children (47 girls and 39 boys)
between 6.5 and 10.2 years of age, with IRT and cold stimulation, Malpique et al.
[17 ] investigated whether the BAT
activation differs between prepubertal children born small-for-gestational-age (SGA)
or appropriate-for-gestational age (AGA) and found significant negative associations
between the change in the area of the thermally active
supraclavicular region after the cold stimulus with SAT (r=–0.243,
p=0.034) and negative and non-significant associations with VAT
(r=–0.194, p=0.077). Another recent finding using PET-CT,
IRT, and cold protocol proved the same significant inverse trend of standardized
uptake values of 18-FDG and SCVT with the percentage of fat
fraction (r=–0.628, p=0.012) [12 ].
In a study analyzing 18 FDG PET-CT scans of 4,852 individuals, it was clear that,
regardless of age groups and gender analyzed, VAT was more associated with BAT
activation than with SAT [15 ]. However, men
showed slightly higher correlation values between BAT, VAT
(r=–0.4945, p<0.000) and SAT (r=–0.3902,
p<0.000) activity, when compared to women’s VAT
(r=–0.3433, p<0.000) and SAT (r=–0.2131,
p<0.000) [15 ].
These existing media correlations between BAT activation through analysis of SCVT,
VAT, and SAT were confirmed in our research, but we did not find great differences
between the correlations when we extract the groups with the highest and lowest
amount of body fat, what points to a balance. To better understand this statement,
we must verify that the effect of the moment and group factor on the SCVTR showed
the same significance and the same large effect size when we extract the general
group by %SAT and %VAT (p=0.006,
η2 =0.073) and (p=0.006,
η2 =0.105). On the SCVTL, the effect of the moment factor
had the same medium effect size between variables and a small p -value
difference. However, we observed that effect size of group factor in %SAT
was classified as very large (p<0.001, η2 =0.189)
and in %VAT as large (p=0.020, η2 =0.089).
Our greatest contribution is to present an original study that, using a
non-invasive, easy-to-apply, and effective technique, confirms that no matter where
the fat is accumulated, it maintains the same degree of correlation with BAT
activation. This minimum difference found in the Pearson’s correlation and
in the value p/size of the effect of the ANOVA are irrelevant to define
which of the types of body fat percentage is more related to BAT activation.
Obviously, it is important to emphasize that within our limitations is the fact that
we did not measure skin thickness in the SCV region. We must take this into account,
as already raised by other authors, since Tskin has a direct influence on the amount
of SAT due to its thermal insulation [29 ]
[30 ]. However, our sample comprised women with
normal BMI and low overall %BF.
Conclusion
The results of the present research infer that, using IRT, were not found difference
in the relation between %SAT, %VAT, and BAT activation in adult
women, pointing out that both types of fat are equally related to BAT thermogenesis
quite, because this is the same thing right in last line. However, subjects with
higher amounts of BF do not seem to achieve a significant increase in SCVT that
characterizes non-thermogenic tremor when exposed to cold.
It is suggested that future studies include a sample of women with a larger age
range, classified by age as well as nutritional status (BMI) and evaluating the
skinfolds of the SCV region to estimate local SAT.
Fundings
Brazilian Army Research Institute Physical Fitness (IPCFEx) —
Pró-Pesquisa 2021/22
FCT-Fundação para a Ciência e Tecnologia —
UID04045/2020
Bibliographical Record
