Obesity rates have increased rapidly worldwide and represent a public health challenge. In Brazil, a large study showed that 55·7 % of the adult and elderly Brazilian population were overweight and that 19·8 % present obesity(1). Obesity has been related to decrease in general health, metabolic co-morbidities, susceptibility to infectious diseases, and predisposition to carcinogenesis and is also a factor of poor prognosis in cancer patients. An umbrella review analysed 204 meta-analyses to investigate the association between clinical and biochemical markers of adiposity and the development of thirty-six primary cancer types. A relationship between body overweight and the risk of at least eleven types of cancer, including haematological neoplasm, was found(Reference Kyrgiou, Kalliala and Markozannes2). Obesity over life course was a modifiable risk for myelodysplastic syndrome (MDS), the most prevalent haematological neoplasm among older people, characterised by cytopenias, ineffective haematopoiesis and increased risk of progression to acute myeloid leukaemia(Reference Poynter, Richardson and Blair3).
There is evidence that obesity disturbs haematopoiesis and the bone marrow niche, with a negative impact on the differentiation and function of haematopoietic cell populations, and unbalances the production of pro-inflammatory cytokines and adipokines(Reference Adler, Kaushansky and Rubin4). Among abundant human adipokines, leptin and adiponectin have an opposing function and play a significant role in metabolic system and weight control.
Leptin is an adipocyte-derived hormone, with an important function not only in regulating appetite, energy consumption, and insulin sensitivity but also in inflammation and immunity system. Higher levels of this adipokine are associated with obesity(Reference Dalamaga, Chou and Shields5). On the other hand, adiponectin is a biologically active adipokine, which plays a role in the metabolism of nutrients, insulin resistance and inflammatory system. Adiponectin inversely correlates with BMI and visceral fat(Reference Marinou, Tousoulis and Antonopoulos6). Several studies have shown the association between higher leptin and lower adiponectin and carcinogenesis and cancer progression(Reference Esfahani, Movahedian and Baranchi7–Reference Ghasemi, Saeidi and Azimi-Nejad10). Previous few studies among MDS individuals found increased leptin and decreased adiponectin concentration when compared with healthy control(Reference Tsiotra, Pappa and Koukourava11–Reference Han and Wang14).
Common polymorphisms of the human leptin (LEP) and adiponectin (ADIPOQ) genes are known to influence the circulating levels of these adipokines. The single-nucleotide polymorphism (SNP) −2548G > A (rs7799039) has been extensively investigated for its influence in leptin expression and association with some chronic diseases (i.e. obesity, diabetes, metabolic syndrome (MetS) and cancer risk)(Reference He, Xi and Ruiter15). The SNP + 276G > T (rs1501299) has been associated with circulating adiponectin concentration and chronic diseases, including obesity, coronary artery disease, MetS and different types of cancer(Reference Lee, Cho and Hong16–Reference Zayani, Omezzine and Boumaiza18). To the best of our knowledge, there are no studies evaluating adipokines gene in patients with MDS.
It is, therefore, very important to assess the relationship between adiposity and MDS not only because of the effect of obesity on haematopoiesis but also because cardiovascular events are the most frequent cause of death unrelated to the disease. In this context, we investigated the association between LEP −2548G > A and ADIPOQ + 276G > T polymorphisms and circulating levels of leptin and adiponectin, adiposity and clinical markers in patients with MDS.
Subjects and methods
Study population
In this cross-sectional study, 102 sequential MDS patients and 102 age- and sex-matched control individuals were evaluated at a single tertiary university hospital.
Patients in the MDS group were included according to the following criteria: aged over 18 years and confirmed diagnosis of MDS. Exclusion criteria were undergoing disease-modifying therapy (i.e. hypomethylating agents and allogeneic stem cell transplantation), HIV infection, and current use of drugs affecting inflammatory system like corticosteroids and other medications. Only healthy adults were included in the control group.
Diagnosis and classification of MDS were established according to the 2016 WHO classification of myeloid neoplasms and acute leukaemia(Reference Arber, Orazi and Hasserjian19) and the Revised International Prognosis Scoring System (IPSS-R)(Reference Greenberg, Tuechler and Schanz20). Clinical and laboratory data (i.e. presence of co-morbidities, blood count, serum ferritin, and HDL and TAG levels) were evaluated by clinical assessment and medical records review. Ferritin level ≥ 1·000 ng/ml was considered Fe overload, and transfusion dependency was defined as transfusion of at least 4 unit/8 weeks.
This study was conducted in accordance with the guidelines set out in the Declaration of Helsinki, and all procedures involving human subjects/patients were approved by the local ethics committee (ethics number: 1.513.488). Written informed consent was obtained from all subjects/patients.
Anthropometric measurements
Weight, height, circumferences and body skinfolds were measured by the trained researchers in a standardised way. BMI was calculated as body weight in kilograms divided by the square of height in metres. Individuals were classified according to BMI into three categories (underweight: BMI < 18·5 kg/m2, eutrophic: BMI of 18·5 to 24·9 kg/m², excess body weight: BMI ≥ 25·0 kg/m² and obesity: BMI ≥ 30·0 kg/m²)(21).
Circumferences were measured using a tape (CESCORF®, length 0–200 cm, accuracy ± 1 mm). Waist circumference (WC) was obtained using a measuring midway between the iliac crest and the lower costal margin, and the hip circumference was measured at the maximum protrusion of the gluteal region(21). WC > 94 cm (male) or > 80 cm (female) were categorised as high abdominal fat accumulation(22).
Skinfold thickness measurements (i.e. biceps, triceps, subscapular and suprailiac) were obtained on the right side of the body by using Premier Cescorf Scientific Adipometer (opening 70 mm and precision of ± 0·2 mm, constant pressure of 10 g/mm²; CESCORF®) in accordance with the procedures previously recommended(Reference Durnin and Womersley23,Reference Harrison, Buskirk, Carter, Lohman, Roche and Martorell24) . Body fat percentage (BF %) was estimated using skinfold measurement equations. Durnin and Womersley equation(Reference Durnin and Womersley23) was used to estimate BF % in the equation where body density was calculated from a skinfold equation. BF % > 25 % (male) and > 30 % (female) was considered increased(Reference Lohman25).
The fat mass index (FMI), a sex-specific measure of fat not confused by lean tissue, was determined using the following equation: FMI = fat mass (kg)/height (m²)(Reference Van Itallie, Yang and Heymsfield26). FMI > 6 (male) or > 9 (female) was categorised as excess of fat(Reference Kelly, Wilson and Heymsfield27).
The visceral adiposity index (VAI) was obtained by the formulas proposed by Amato et al (2010). For males, VAI = (WC/36·58 + (1896 BMI)) × (TG/0·81) × (1·52/HDL) and for females, VAI = (WC/39·68+ (1886 BMI)) × (TG/1·03) × (1·31/HDL). The cutoff points for VAI in the detection of visceral adiposity in adults and the elderly followed as determined by Amato et al.(Reference Amato, Giordano and Galia28,Reference Amato, Giordano and Pitrone29)
Metabolic syndrome diagnoses
MetS criteria were used according to NCEP-ATPIII (2001): abdominal obesity ≥102 cm (male) ≥88 cm (female), HDL-cholesterol <40 mg/dl (male) <50 mg/dl (female), systolic blood pressure ≥130 mmHg or diastolic ≥85 mmHg, TAG ≥150 mg/dl and fasting blood glucose ≥ 110 mg/dl(30).
Adipokines analysis
Serum samples were obtained from MDS patients and controls after 12-h fast. Serum total adiponectin and serum total leptin concentrations were measured using a commercial ELISA kit according to the manufacturer’s protocol (Invitrogen).
DNA isolation and genotyping
DNA extraction
Genomic DNA was extracted from peripheral blood samples of MDS patients and sex- and age-matched healthy volunteers, using Trizol Reagent™ (Invitrogen), according to the manufacturer’s protocol.
Allelic discrimination by real-time-PCR
Genotypes were identified in DNA samples by real-time PCR, using TaqMan SNP Genotyping Assay (ThermoFisher Scientific). SNP were selected based on a known or expected association with obesity, the resulting metabolic complications and cancer. We performed an allelic discrimination assay for rs7799039 and rs1501299 polymorphisms using the TaqMan Genotyping Master Mix kit® (ThermoFisher Scientific), according to the manufacturer’s protocol.
Each 10 μl of PCR reaction contained 50 ng of genomic DNA in the final volume of 4·5 μl, 5·0 μl of 2X TaqMan Genotyping Master Mix and 0·5 μl of 20X TaqMan Genotyping Assay Mix. The following thermal cycling conditions were an initial activation at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s of denaturation step at 95°C and annealing/enzyme extension at 60°C for 1 min. PCR was performed using 7500 Fast System® (Applied Biosystems).
Statistical analysis
Normality was verified by the Shapiro–Wilk test. Student’s t test, Levene’s test, Pearson’s χ 2 test or Fisher’s exact test, Pearson’s correlation and the multinominal logistic regression analysis were performed using SPSS® (SPSS Inc.) software version 20.0. P < 0·05 was considered significant.
The Pearson χ 2 was used to determine the Hardy–Weinberg (H–W) equilibrium and the differences in genotype distribution. All genotypes were divided and analysed in genotype distribution model (wild type v. heterozygous v. mutated), according to the study of Clarke et al.(Reference Clarke, Anderson and Pettersson31) Differences in allele and genotype frequencies and comparisons of variables between the control and MDS groups were evaluated using the Pearson’s χ 2 test or Fisher’s exact test. Multinomial logistic regression analysis was used to measure the association between an exposure and an outcome (odds ratio) with 95 % CI.
Results
Characteristics of myelodysplastic syndrome patients
One hundred and two sequential patients with MDS were included (22·5 % aged <65 years and 77·5 % aged ≥65 years). The mean age of all MDS patients was 72·0 years ± 11·6 (38–96). Most patients were women 64·7 % (n = 66). According to WHO classification of myeloid neoplasms and acute leukaemia(Reference Arber, Orazi and Hasserjian19), 71·6 % of patients were classified as lower risk (9·8 % MDS-SLD, 41·2 % MDS-MLD, 4·9 % MDS-RS-SLD, 11·8 % MDS-RS-MLD and 3·9 % MDS with isolated del5q-) and 28·4 % as higher risk (15·7 % MDS-EB-I, 7·8 % MDS-EB-II and 4·9 % t-MDS). According to the prognostic stratification proposed by the IPSS-R (2012), a higher percentage of patients, 28·4 % (n = 29), had a low-risk disease, with 18·6 % (n = 19) with risk intermediate, 8·8 % (n = 9) very high risk, 7·8 % (n = 8) high risk and 6·9 % (n = 7) very low risk (data not presented in the table).
Transfusion dependency was reported in 36·3 % (n = 37), and Fe overload was observed in 13·7 % (n = 14). The most prevalent cytopenia was anaemia (89·2 %/n = 91). Neutropenia and thrombocytopenia were present in 52·0 % (n = 53) and 58·8 % (n = 60), respectively. As for the percentage of blasts in the bone marrow, 23·5 % of the patients had 5 % or more blasts (data not presented in the table).
According to values proposed by IPSS-R, 63·7 % of patients presented Hb levels below 10 g/dl, 17·6 % of patients presented a number of neutrophils less than 0·8 × 109 l and 43·1 % had platelet counts below 100 × 109/l (data not presented in the table).
Anthropometric assessment
According to the BMI classification, 47·1 % of patients (n = 48) were overweight. There was no significant difference between the BMI classifications between the groups (P = 0·096). There was no significant difference between the mean values of BMI, WC, FMI, VAI and percentage of fat mass (FM%) between both groups (P > 0·05) (Table 1).
Table 1. Mean value of anthropometric parameters of patients with MDS (n = 102) and controls (n = 102)
The data in bold were significant (P value ≤ with the
* Student’s t test and
† Mann–Whitney test.
Data are represented as mean or median and SD.
P values were considered significant less than or equal to 5 %.
Serum adipokines
The mean serum leptin among patients with MDS was 298·3 ± 153·07 ng/ml and among the control group was 184·68 ± 83·61 ng/ml (P = 0·03) (Fig. 1).
Fig. 1. Serum leptin level in MDS and control group. Bars represent means ± SD. Serum leptin level in patients with MDS was significantly different (*P = 0·03) compared with the control group. MDS, myelodysplastic syndrome. Student’s t test. P value was considered significant less than or equal to 5 %.
Female MDS patients had significantly higher serum leptin concentration than male patients (P = 0·036). Higher mean serum leptin was also found among patients with high WC (P < 0·001) and FMI (P = 0·017) (Table 2). There was no difference in serum leptin values between MDS who subtypes and controls (P > 0·05) (data not presented in the table).
Table 2. Comparison of leptin and adiponectin values according to clinical and anthropometric variables of patients with MDS (n = 102)
MDS, myelodysplastic syndrome; IPSS-R, International Prognostic Scoring System Revised.
The data in bold were significant with the
* Student’s t test or ANOVA and
† Mann–Whitney test.
Data are represented as mean or median and SD.
P values were considered significant less than or equal to 0·05.
In the MDS group, there was a significant positive correlation between serum leptin levels and values: BMI (r = 0·264, P = 0·025), WC (r = 0·235, P = 0·047), BF % (r = 0·373, P < 0·001) and FMI (r = 0·371, P < 0·001) (Fig. 2). There was no significant correlation between leptin levels and the variables: age (r = 0·112, P = 0·322), Hb (r = 0·042, P = 0·727), neutrophils (r = −0·199; P = 0·094), platelets (r = −0·165; P = 0·166), percentage of blasts (r = 0·70; P = 0·557) and ferritin (r = −0·017; P = 0·924) (data not presented in the figure).
Fig. 2. (a) Correlation between leptin and BMI. (b) Correlation between leptin and waist circumference. (c) Correlation between leptin and body fat percentage. (d) Correlation between leptin and fat mass index *Pearson’s correlation test (r) P < 0·05.
Serum adiponectin was significantly lower in MDS patients than the control group (P = 0·033). The mean serum adiponectin among the MDS group was 7·10 ± 4·87 µg/ml and among control individuals was 9·00 ± 5·38 µg/ml (Fig. 3).
Fig. 3. Serum adiponectin level in MDS and control group. Bars represent means ± SD. Serum adiponectin level in patients with MDS was significantly different (*P = 0·033) compared with the control group. MDS, myelodysplastic syndrome. Student’s t test. P value was considered significant less than or equal to 5 %.
There was a significant negative correlation between the serum adiponectin levels of MDS patients and BMI values (r = −0·287, P = 0·005), WC (r = −0·226, P = 0·028), FMI (r = −0·266, P = 0·010) and VAI (r = -0·381, P < 0·001) (Fig. 4).
Fig. 4. (a) Correlation between adiponectin and BMI. (b) Correlation between adiponectin and waist circumference. (c) Correlation between adiponectin and fat mass index. (d) Correlation between adiponectin and visceral adiposity index. *Pearson’s correlation test (r) P < 0·05.
There was no significant correlation between adiponectin levels and the variables: age (r = 0·174, P = 0·094), Hb (r = −0·020, P = 0·846), neutrophils (r = -0·055, P = 0·600), platelets (r = 0·078, P = 0·455), percentage of blasts (r = −0·177, P = 0·088), ferritin (r = −0·024, P = 0·878) and fat mass percentage (r = -0·146, P = 0·160) (data not presented in the figure).
According to BMI, overweight MDS patients had significantly lower serum adiponectin than eutrophic patients (P = 0·016). Likewise, a significant difference was observed in adiponectin concentration according to individuals with adequate and higher FMI (P = 0·013). The lower mean of serum adiponectin was also found among patients with high fat mass percentage (P = 0·019) and higher VAI (P = 0·004) (Table 2).
When assessing the presence of co-morbidities and the serum level of adiponectin in patients with MDS, a lower concentration of this adipokine was observed in patients with MetS (P = 0·016). Regarding clinical, haematological and prognostic parameters, MDS patients with excess blasts in bone marrow (≥ 5 %) had a significantly lower median of adiponectin concentration (P = 0·041). There were no significant comparisons with the other variables (Table 2).
Genotype frequencies distributions of LEP and ADIPOQ polymorphisms
LEP rs7799039 and ADIPOQ rs1501299 polymorphisms are in Hardy–Weinberg equilibrium (P > 0·05). The distributions of allele and genotype frequencies of MDS patients and controls are presented in Table 3.
Table 3. Distributions of genotype frequencies of MDS patients and controls (n = 92)
MDS, myelodysplastic syndrome.
* Pearson or Fisher’s χ².
Significant associations were found between the LEP gene rs7799039 polymorphism genotypes and the sex variable in the genotypic distribution model (GG × AG × AA) (P = 0·018) and classification of blast percentage (P = 0·031). Patients with AA genotype of the LEP gene rs7799039 polymorphism had a higher chance of being male (OR = 3·77, 95 % CI (1·08, 13·10)) when compared with patients with AG or GG genotype. In addition, patients with AA genotype had a higher chance ratio (OR: 4·42, 95 % CI (1·23, 15·89)) of having a blast percentage ≥ 5 %. The other variables did not show any significant association (Table 4).
Table 4. Comparison of LEP gene polymorphism rs7799039 with clinical and anthropometric data in patients with MDS (n = 92)
IPSS-R, International Prognostic Scoring System Revised.
* Pearson or Fisher’s χ².
Comparisons between groups: (1) the genotype distribution model was modelled in three categories.
The data in bold were significant at P ≤ 0·05.
Statistically significant value for P ≤ 0·05.
Through the analysis of the polymorphism rs1501299 of the ADIPOQ gene, a significant association was observed in patients with MDS between genotypes and Fe overload (ferritin ≥ 1000 ng/ml) in the genotypic distribution model (GG × GT × TT) (P = 0·033). Patients with TT genotype from SNP rs1501299 had an eight-fold significant chance of having ferritin greater than 1000 ng/ml (OR = 8000 (95 % CI (1·367, 46·812)) when compared with patients with GG or GT genotype. The other variables did not show any significant association (Table 5).
Table 5. Comparison of the polymorphism of the ADIPOQ rs1501299 gene with clinical and anthropometric data in patients with MDS (n = 92)
MDS, myelodysplastic syndrome; IPSS-R, International Prognostic Scoring System Revised.
* Pearson or Fisher’s χ².
Comparisons between groups: (1) the genotype distribution model was modelled in three categories.
The data in bold were significant at P ≤ 0·05.
·
Discussion
This study investigated whether MDS patients have different anthropometric and adipokines status than healthy age- and sex-matched control group. Additionally, we verified whether the leptin gene polymorphism (rs7799039) and the adiponectin gene polymorphism (rs1501299) were associated with clinical and anthropometric status.
First, we found that MDS patients have significantly higher serum leptin concentration and significantly lower mean serum adiponectin than the control group. Considering that anthropometric parameters did not differ significantly, this finding suggests a possible relationship between the dysregulated adipokine production and MDS etiopathogenesis.
Dalamaga et al. (Reference Dalamaga, Nikolaidou and Karmaniolas12), in their case–control study with 101 cases with MDS and 101 controls paired by sex and age, observed that higher serum levels of adiponectin were associated with a lower risk of MDS and individuals with lower levels of leptin had a lower risk of MDS than controls. Likewise, low leptin concentrations were observed in low-risk MDS patients with normal or good prognostic karyotypes after adjustment for age, sex and BMI.
Over the past few years, the relationship between leptin and cancer has drawn attention. Higher levels of leptin were also found in breast, gastric, colon, endometrial cancer and acute lymphoid leukaemia when compared with healthy population(Reference Hosney, Sabet and El-Shinawi32–Reference Aref, Ibrahim and Azmy36). Dalamaga et al. (Reference Dalamaga, Nikolaidou and Karmaniolas12) and Tsiotra et al.(Reference Tsiotra, Pappa and Koukourava11) found higher concentrations of leptin among MDS patients when compared with healthy controls.
Elevated leptin level is harmful to cancer patients. This cytokine can stimulate migration and metastasis stages. Association between increased leptin expression and metastasis was observed in some cancers, such as breast, colon, oesophagus, lung, ovary, pancreas, stomach and thyroid(Reference Ray and Cleary9,Reference Ghasemi, Saeidi and Azimi-Nejad10) . In haematological cancers, leptin can act in the genesis and progression of the disease. This cytokine was shown to have proliferative and apoptosis-inhibitory action in leukaemic cells(Reference Tsiotra, Pappa and Koukourava11,Reference Konopleva, Mikhail and Estrov37) .
In addition, adiponectin levels are reduced in various types of cancers, such as breast cancer(Reference Mantzoros, Petridou and Dessypris38), colorectal cancer(Reference Wei, Giovannucci and Fuchs39), endometrial cancer(Reference Petridou, Mantzoros and Dessypris40), gastric(Reference Ishikawa, Kitayama and Kazama41), liver(Reference Arano, Nakagawa and Tateishi42) and pancreatic(Reference Stolzenberg-Solomon, Weinstein and Pollak43), including haematological cancer(Reference Dalamaga, Karmaniolas and Nikolaidou13,Reference Aref, Ibrahim and Azmy36) .
Although the mechanisms that associate adiponectin and carcinogenesis have not yet been elucidated, some studies report proliferative and anti-apoptotic actions of adiponectin in cancer cell lines(Reference Barb, Williams and Neuwirth44,Reference Lang and Ratke45) . Low levels of adiponectin were associated with increased synthesis of fatty acids and proteins, thus enabling cell proliferation and growth, as well as DNA mutagenesis(Reference Wang, Lam and Xu46). Furthermore, hypoadiponectinemia supports tumour proliferation by increasing anabolic hormones, such as insulin and IGF-1, which can act by inhibiting apoptosis and increasing cell proliferation(Reference Spyrou, Avgerinos and Mantzoros8).
Regarding the concentration of adiponectin and the prognostic parameters for MDS, we found in our study that patients with excess bone marrow blasts (≥5 %) had a significantly lower median of adiponectin when compared with patients with a lower amount at 5 % blasts in the BM (P = 0·041). Aref et al.(Reference Aref, Ibrahim and Azmy36) found a significant negative correlation between serum adiponectin levels and the percentage of blast cells in the bone marrow in patients with acute leukaemia, an important adverse prognostic factor.
A possible explanation for this finding is the fact that the large percentage of blast cells in the bone marrow may lead to a reduction of adipose tissue in the marrow and, thus, to lower levels of adiponectin, since it is produced almost exclusively by adipose tissue. This hypothesis is due to the fact that the number of adipocytes encroaching into the bone marrow space correlates inversely with the number of other cells(Reference Adler, Kaushansky and Rubin4,Reference Azadniv, Myers and McMurray47) .
It is noteworthy that adipose tissue, responsible for the production of adiponectin and leptin, makes up most of the structure of bone marrow, especially among older people. In addition, since it was discovered in 1994, leptin has been associated with obesity. Many studies have found higher levels of leptin in fat individuals when compared with lean population(Reference Halaas, Gajiwala and Maffei48,Reference Friedman49) .
In our study, there was a clear relationship between serum leptin levels and body fat markers. Higher serum leptin concentration was found in patients with higher WC and FMI (Table 2). A weak, but significant, positive correlation was found between serum leptin and BMI, WC and BF % (Fig. 2). Several other studies support our findings. A positive relationship between leptin levels and body adipose tissue volume is well established(Reference Izquierdo, Crujeiras and Casanueva50).
Positive correlation between BMI and leptin was observed in patients with colon, prostate, breast and haematological cancer(Reference Han and Wang14,Reference Stachowicz, Mazurek and Nowakowska-Zajdel51–Reference Pan, Deng and Cui53) . Tsiotra et al. (Reference Tsiotra, Pappa and Koukourava11) found a significant positive correlation between leptin and BMI, but not with other haematological and clinical parameters of the disease in patients with MDS.
Furthermore, female MDS patients had higher leptin mean when compared with male patients. Similar findings were observed in healthy population and MDS individuals Awede et al.(Reference Awede, Adovoekpe and Adehan54), Ayina et al.(Reference Ayina, Noubiap and Etoundingoa55) and Tsiotra et al.(Reference Tsiotra, Pappa and Koukourava11). Oestrogen has been linked to a possible role in stimulating leptin production(Reference Moon, Dalamaga and Kim56,Reference Hu, Gan and Wen57) .
In our study, we also observed significantly lower concentrations of adiponectin in higher body adiposity patients, according to BMI, FM %, FMI and VAI compared with lean individuals (Table 2). Srivastava et al.(Reference Srivastava, Batra and Tyagi58), studying obesity markers in 159 patients with acute leukaemia, including BMI, observed a significantly lower mean level of serum adiponectin in individuals with obesity compared with patients with adequate BMI, a result similar to what was demonstrated in our study.
A weak, but significant, negative correlations between serum adiponectin levels of MDS patients and BMI values were also observed in our study, as well as among other anthropometric markers such as WC, FMI and VAI (Fig. 4).
Adiponectin is secreted by adipose tissue in an amount inversely proportional to BMI, thus being in reduced concentrations in individuals with obesity(Reference Obeid and Hebbard59). In addition, serum levels of this adipokine are more associated with the amount of visceral fat than total body fat(Reference Marinou, Tousoulis and Antonopoulos6). Excess visceral fat, which results in visceral adipose tissue dysfunction, leads to unregulated production of adipokines, such as underproduction of adiponectin. Hypoadiponectinemia and the deposition of excess visceral fat may play important roles in the mechanism associated with the onset of chronic diseases(Reference Roberts, Dive and Renehan60).
Mihu et al. (Reference Mihu, Ciortea and Mihu61), when evaluating the amount of abdominal fat and plasma levels of adiponectin in patients with endometrial cancer, found a significantly lower level of adiponectin compared with the control group, as well as abdominal fat showed a negative correlation with the level of adiponectin, a finding that corroborates our study. Serretta et al.(Reference Serretta, Abrate and Siracusano62) found no significant correlation between BMI, VAI and plasma levels of adiponectin in patients with prostate cancer.
In our study, we also observed a significantly lower amount of adiponectin in MDS patients with MetS. Our group was the first to describe the MetS prevalence in MDS patients. A significantly higher prevalence of MetS was observed in elderly patients when compared with younger patients. In addition, MetS was associated with transfusion dependence, a poor prognostic factor in MDS. Thus, it was concluded that the presence of MetS is an additional factor not related to the disease that potentially increases its morbidity(Reference Aguiar, Mendonça and Ribeiro-Júnior63).
Furthermore, as already noted, visceral fat is inversely correlated with adiponectin levels, thus suggesting a close relationship between visceral obesity, hypoadiponectinemia and metabolic diseases. Cho et al.(Reference Cho, Joo and Cho64) observed a higher incidence of MetS in the group that had a greater area of visceral adipose tissue and a lower serum level of adiponectin, data that corroborate our study.
In our study, there was no difference in genotypes and alleles of LEP rs7799039 G > A polymorphism between MDS and control group. The adipocyte-derived peptide hormone LEP has a well-known role on inflammation, tumour growth and metastasis of cancer. In a recent meta-analysis, the LEP rs7799039 G > A polymorphism indicated a risk of the development of cancer, including haematological cancer(Reference Tang, Kang and Liu65).
We found a significant association between AA genotype and bone marrow blasts percentage ≥ 5 % (Table 4). The concentration of blasts in the bone marrow is an important clinical marker and one of the factors with the greatest negative impact on the prognosis of patients with MDS(Reference Greenberg, Tuechler and Schanz20).
In this study, there was no significant difference between leptin levels and rs7799039 polymorphisms. Similar results were found in South Africans, Japanese and Malays individuals(Reference Okpechi, Rayner and Van Der Merwe66–Reference Fan and Say68). In contrast, studies have shown that the presence of the −2548G/A (rs7799039) polymorphism influences leptin expression, probably by transcription(Reference Hoffstedt, Eriksson and Mottagui-Tabar69–Reference Terrasi, Fiorio and Mercanti71). It is noteworthy that the ethnic influence on the distribution of LEP gene polymorphisms has been recognised(Reference Fan and Say68).
As was observed in the LEP rs7799039 G > A polymorphism, in our study, we also found no difference in genotypes and alleles for ADIPOQ rs1501299 T > G between MDS and the control group. However, it is important to note that the rs1501299 T > G SNP was associated with a lower risk of some solid cancers, such as breast(Reference Kaklamani, Sadim and Hsi72), oesophageal and hepatocellular(Reference Pei, Xu and Niu73). Besides, this SNP was also associated with decreased serum concentrations of adiponectin(Reference Kyriakou, Collins and Spencer-Jones74,Reference Xu, He and Pan75) and increased risk of obesity in some populations(Reference Zayani, Omezzine and Boumaiza18,Reference Ramya, Ayyappa and Ghosh76) . In our study, we did not observe a significant association between adiponectin levels and anthropometric parameters with the gene frequencies of this SNP.
On the other hand, we found a significant association between genotypes and Fe overload (ferritin ≥1000 ng/ml). Patients with the polymorphic TT genotype had a significantly higher chance of having Fw overload when compared with patients with the GT and GG genotype.
Ferritin is considered a marker of inflammation, an acute phase reactant. Furthermore, Fe overload in transfusion-dependent patients results in tissue damage caused by oxidative stress and inflammation. In patients with haematological malignancies, including MDS, it has been observed that Fe overload can lead to liver dysfunction, hepatic sinusoidal obstruction syndrome and type 2 diabetes, which substantially influence the long-term survival of patients(Reference Alessandrino, Della Porta and Bacigalupo77,Reference Yang, Tang and An78) .
Strengths of this study included that it was the first group to assess patients with different anthropometric markers and not just BMI. It is known that BMI is a relatively rough measure of body adiposity, as it does not differentiate between lean mass and fat mass and does not inform about the distribution of body fat(Reference Singer, Blaser and Berger79). Furthermore, adipose tissue dysfunction and its consequences do not occur exclusively in patients with obesity by BMI. Insulin resistance and inflammation have been reported in individuals with normal BMI, as well as individuals with high BMI who were considered metabolically healthy(Reference Chen, Chen and Liu80). Our study showed the importance of assessing body adiposity through the use of other markers of nutritional status, avoiding the use of BMI as an isolated measure of nutritional diagnosis. Furthermore, this was the first study that evaluated adipokine polymorphisms in patients with MDS.
However, there are limitations to consider. As this is a cross-sectional study, the association between exposure to the disease is not the same as that detected at the same point in the population. Perhaps the evaluation of more polymorphisms could find an association with the disease.
Conclusion
It is concluded that body adiposity excess was present in an important percentage of patients with MDS, although not significantly different from control. Higher leptin and hypoadiponectinemia were found in MDS patients compared with controls, with even higher leptin and lower adiponectin values in patients with higher body adiposity. Patients with a less favourable prognosis, according to blast percentage and ferritin levels, showed associations with the LEP −2548G > A and the ADIPOQ + 276G > T SNP, respectively.
These results demonstrate the importance of a comprehensive and systematic evaluation in MDS patients in order to identify and early control negative non-disease-related factors. Nutritional status, circulating adipokines, and leptin and adiponectin polymorphisms may play a potential role in disease pathogenesis. The association between these factors, not only with anthropometric parameters but also with some characteristics of the MDS, deserves further in-depth studies.
Acknowledgements
This study was conducted with partial support from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and Brazilian Company of Hospital Services (EBSERH). CNPq, CAPES and EBSERH had no role in the design, analysis or writing of this article.
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
All authors designed the study and edited the manuscript. All have approved the final version of manuscript before to submission to the British Journal of Nutrition. Study concepts: A. P. N. A., P. S. M., A. G. M. M., H. L. R. J., R. C. P. L. J., R. F. P. and S. M. M. M. Study design: A. P. N. A., P. S. M., R. C. P. L. J., R. T. G. O., H. L. R. J., R. F. P. and S. M. M. M. Data acquisition and quality control of data and algorithms: A. P. N. A., P. S. M., H. L. R. J., R. T. G. O., A. G. M. M., D. V. T. W., R. F. P. and S. M. M. M. Data analysis and interpretation: A. P. N. A., P. S. M., H. L. R. J., R. F. P. and S. M. M. M. Statistical analysis: A. P. N. A., P. S. M., H. L. R. J., R. T. G. O., R. F. P. and S. M. M. M. Manuscript preparation and editing: A. P. N. A., P. S. M., H. L. R. J., R. T. G. O., A. G. M. M., D. V. T. W., R. F. P. and S. M. M. M. Manuscript review: A. P. N. A., P. S. M., H. L. R. J., R. C. P. L. J., D. V. T. W., R. F. P. and S. M. M. M.
The authors declare that there is no conflict of interest.
