Gender and Liver Steatosis Discriminate Different Physiological Patterns in Obese Patients Undergoing Bariatric Surgery: Obesity Center Cohort

Background: Obesity is a major public health problem worldwide. Bariatric surgery can reduce body weight, and it is one of the better ways to improve metabolic disease and lifestyle. The aim of this study was to explore a new cohort of patients with obesity and evaluate the gender differences and the steatosis status within the gender group. Methods: A cohort of 250 adult obese patients with BMI ≥ 30 and age >18 years, eligible for gastric bariatric surgery at Pineta Grande Hospital, Castel Volturno (Italy) was studied. Results: The prevalence in women was higher (72.40%) than men (27.60%). Overall, results indicated many statistically significant gender differences in hematological and clinical parameters. Analysis of the subcohorts based on the severity of steatosis revealed differences of this condition between the genders. Steatosis was more prevalent in the male subcohort, but female patients revealed greater within-group differences. Conclusions: Many differences were found not only in the total cohort but also between the gender subcohorts, both in the presence and absence of steatosis. We can conclude that the pathophysiological, genetic, and hormonal patterns affecting these patients delineate different individual profiles.


Introduction
Over the last few decades, obesity has become a growing public health problem worldwide, and the related conditions differ by geographic region [1]. Italy follows the trend of the progressive increase of the prevalence of obesity, showing a rise by almost 30% in the adult obese population [2]. Obesity, traditionally defined as an excess of body fat causing prejudice to health [3], is the result of complex relationships among genetic, lifestyle, socioeconomic, and cultural factors because obese individuals tend to have a lower grade of instruction and lower income [4]. In addition, obesity may be a risk factor for the development of comorbidity conditions. Bariatric surgery is a primary component in determining weight loss as well as costs associated with obesity. According to the guidelines of the Italian Society for Bariatric Surgery (SICOB), patients with class III or II obesity and with other comorbidities are eligible for bariatric surgery that is fully covered by the Italian National Health System (SSN) [5]. However, only 1.4% of eligible patients, on average, undergo bariatric surgery in Italy, as stated in the Italian report "Osservatorio PariSanità" [6]. The main contraindications to surgery are psychological features, indicating that a patient would not be able to cope with the impact of the procedures. Pragmatic opinion-based decisions after a multidisciplinary team assessment are currently the best solution hospitals can offer patients prior to referring them for bariatric surgery [7], because not all candidates are suitable for surgery. The most commonly performed procedures are the Roux-en-Y gastric bypass (RYGB), adjustable gastric banding (AGB) and sleeve gastrectomy (SG) [8]. All procedures can be performed laparoscopically and have a low rate of complications such as wound infection and incisional hernias [9]. Gender is an important risk factor for the development of obesity. Kappoor et al. [10] had hypothesized that factors leading to gender disparities in obesity could include socioeconomic and sociocultural status, hormone-related comorbidities, and metabolic complications. The literature on this aspect is poor and does not provide details about types of prevention and evaluation. Moreover, the majority of bariatric patients are female, nearly half of which are of peri-or post-menopausal age [11]. Men tend to develop complications from obesity more easily. Obesity induces a state of chronic, lowgrade inflammation responsible for the development of the metabolic syndrome and its pathophysiological consequences, such as insulin resistance, cardiovascular disease and other nonmetabolic complications of obesity [12].
Steatosis is an accumulation of triacylglycerols within hepatocytes based on an unbalanced synthesis or delivery of fatty acids in the liver [13]. The main aspect involved in this gender obesity study was the prevalence of steatosis, which was more prevalent in male than in female patients, even if females account for the majority of the obesity population. The severity of steatosis has been associated with increased visceral fat mass, insulin resistance, lower adiponectin levels, and higher blood pressure [14]. Not all patients with steatosis are overweight or obese [15], and vice versa, not all obese patients have steatosis.
The aim of this article was to study, for the first time, a cohort of obese patients in Southern Italy, exploring the epidemiological and physiological differences between genders and liver conditions.

Materials and Methods
Between 1 January and 31 December 2021, 250 adult obese patients with BMI ≥ 30, and age > 18 years, male and female, eligible for gastric bariatric surgery at a medical center in southern Italy (Pineta Grande Hospital, Castel Volturno, Italy), were included in this study, but the funded investigation was conducted at National Institute of Gastroenterology-IRCCS "Saverio de Bellis". SARS-CoV-2 has negatively affected the number of patients undergoing bariatric surgery, despite the consequent increase in obesity [16]. Before the bariatric surgery, all patients were evaluated by a bariatric team, including internists, dietitians, and psychologists. All patients were prescribed a nutritional regime by the dietitians based on their clinical and physiological characteristics. Three different kinds of diet were assigned: Ketogenic, low-carb or balanced hypocaloric diet, consisting of a maximum daily intake of 800, 1000 and 1200 Kcal, respectively. The diets were created using WinFood 3.29 software (Medimatica Srl Unipersonale, Colonnella, Italy). The detailed bromatological compositions are indicated in Table S1.
Between the prehospitalization and the day before the surgery, there are at least 14 days, which is the minimum to record an effect of the diet on the patient's physiological parameters.

Statistical Analysis
Obese patients' characteristics are reported as mean and standard deviation (M ± SD) for continuous variables and as frequency and percentages (%) for categories. To test the association between independent groups (Female vs. Male, and Steatosis Moderate/Severe vs. Steatosis No/Mild), chi-square or Fisher's exact test was used for categorical variables, where necessary, while the Wilcoxon rank-sum (Mann-Whitney) test was used for continuous variables. In addition, the Wilcoxon matched-pairs signed-rank test for continuous variables was applied to evaluate variations between prehospitalization/surgery, only for variables recorded in repeated tests.
To summarize and visualize the information about prehospitalization blood parameters and to create a profile of patients by gender and steatosis, a principal component analysis (PCA) was performed. This was useful to extract information from the data matrix and reduce it to fit new groups of variables. To test the null hypothesis of no association, the two-tailed probability level was set at <0.05. Analyses were conducted with StataCorp.2021 software, Release 17, StataCorp LLC., College Station, TX, USA, while for the graphics, RStudio ("Elsbeth Geranium" Release) was used.
In Table 2, the same cohort was stratified by gender but also by the presence of liver steatosis (No/Mild vs. Moderate/Severe).
In the same way, marital status showed statistically significant differences both in the total cohort and in the female subcohort (p < 0.001 and p = 0.003, respectively) ( Figure 1C,D). Regarding social status, employed male patients had a higher prevalence of steatosis than employed female patients (84.8% vs. 47.5%, p < 0.001) ( Figure 1E,F).      Within the category with a moderate/severe steatosis clinical status, the BMI was higher in patients in the total and female subcohorts (p = 0.02 and p = 0.008, respectively). The choice of diet and subsequent bariatric surgery reflected a prevalence of a ketogenic diet (p = 0.05 and p = 0.03) and sleeve surgery (p = 0.009 and p = 0.01) in steatosis patients, both in the total cohort and in females, as compared to patients without or with mild steatosis. In relation to the health status, hospitalization days were increased in the total and female subcohorts with moderate and severe steatosis (3.8 ± 2.9 vs. 3.3 ± 2.3 with p = 0.001, and 3.8 ± 3.2 vs. 3.3 ± 2.5 with p = 0.002, respectively). In the total cohort with moderate/severe steatosis, we found high levels of alfpha-2 (%) and alpha-2, triglycerides, GOT, GGT, ferritin, cholinesterase, GPT, PTH Intact, WBC, RBC, Hgb, HCT, neutrophils, HbA1c, and glycosylated hemoglobin, and low levels of HDL, and lymphocytes (p = 0.02, and p = 0.05 respectively). The male subcohort had high levels of alpha-1, GPT, HbA1c, and glycosylated hemoglobin, but lower levels of total and indirect fractional bilirubinemia, p-values < 0.05. Instead, in female patients, differences were more evident between the two steatosis conditions. Patients with moderate/severe steatosis had high levels of alpha-2 (%) and alpha-2, triglycerides, ferritin, cholinesterase, GPT, PTH Intact, and WBC, p < 0.05. Blood group (0) was differently distributed, being most prevalent in female patients with moderate/severe steatosis, while the AB group had a lower prevalence of severe steatosis. Naturally, liver volume was increased in pathological conditions in all groups, and gallbladder stones in the total cohort had a higher prevalence in disease conditions (20.7% vs. 8.9%, p = 0.05). Investigation by gender of patients with no/mild steatosis highlighted that gamma (%), alpha-1, and ESR showed lower levels, as well as indirect fractional bilirubinemia, creatinine, iron, ferritin, potassium, RBC, Hgb, HCT, and MCV, as compared to the steatotic group. Instead, comparison between this disease group by gender showed that females had higher levels of alpha-2, transferrin, TSH, PTH Intact, HDL, platelets, PCT, and ESR, but lower levels of total protein, albumin, A/G ratio, beta-2, glycemia, creatinemia, CPK, GGT, iron, ferritin, FT3, RBC, Hgb, HCT, MCH, MCHC, PDW, monocytes, eosinophils. Hypertension was more prevalent in males than females (55.8% vs. 38.6%, p = 0.03) in moderate/severe steatosis conditions and gallbladder volume, where males manifested worse conditions (such as a contracted or scleroatrophic gallbladder). Table 3 shows changes in blood parameters from prehospitalization to bariatric surgery preparation after prescription of the diet to reduce weight by at least 10% of the initial weight. Both males and female showed a decrease in azotemia, creatinine, GGT, ALP, lipase, RBC, Hgb, HCT, platelets, PCT, eosinophils, and basophils, p-value < 0.05. Female patients also had a decrease in prothrombin time (94.8 ± 18.4 vs. 108.2 ± 20.5, p = 0.0005), MCV (84.9 ± 6.5 vs. 84.9 ± 6.2, p = 0.03), and RDW-CV (14.2 ± 1.2 vs. 14.4 ± 1.1, p = 0.03). Meanwhile, the diet had the effect of increasing sodium, WBC (white blood cells), and neutrophils in both genders. In particular, male patients had an increased MCV (86.9 ± 6.4 vs. 86.7 ± 4.4, p = 0.05), whereas females showed a decrease. Female patients also showed a decrease of I.N.R. (1.0 ± 0.1 vs. 1.0 ± 0.2, p = 0.0003).
To summarize and visualize the information about prehospitalization blood parameters and create a profile of obese patients by gender and grade of steatosis, PCA was performed. Figure 2A shows a scree plot displaying the number of principal components. Firstly, two dimensions explain the 11.8% and 8.1% of variance for males and 8.4% and 7.3% for females, respectively ( Figure 3A). These two dimensions were then visualized in a biplot ( Figures 2C and 3C) in which the variables coordinates are represented as projections inside the circle (i.e., Dimension 1 × Dimension 2 plane). A high value of variable contribution (called "cos2") indicates a quality representation of the variable of the principal component and vice versa, a low value shows an imperfect representation. Variables close to the center of the plot were less important for the first components. This biplot shows a clear separation between parameters, with a high contribution of Albumin (%) and the A/G Ratio both for males and females (Figures 2B and 3D). Figure 2C,D show the variable that constituted this component, for males and for females ( Figure 3C,D).
In the biplot (Figure 4), PCA analysis could not clearly identify a distinct steatosisspecific condition in the three groups analyzed (total, male and female cohorts). In the first two dimensions (Figure 4A), the variance explained was 8.7% and 6.8%, respectively, while it was 10.7% and 8.4% for the male and 8.4% and 7.3% for the female cohort.

Discussion
The aim of this study was to examine a new cohort of patients in southern Italy with pathological obesity and, in particular, to explore differences between genders in regard to their blood parameters.
The prevalence of overweight and obese men and women varies greatly within and between countries. Overall, more women than men were affected by obesity [18]. Differences due to men's employment status and to their social context, as well as to the level of education and civil status, may work against the female sex. Added to these were bad habits, such as cigarette smoking [19]. Finally, these social differences were compounded by genetic, physiological, and hormonal profile differences in the patients analyzed. The differences between gender found in our cohort were reflected in the literature, although the possibility cannot be completely ruled out that obesity could amplify differences between the two genders.
With respect to blood parameters, plasma albumin concentrations did not differ between males and females [20] in healthy patients, but they were lower in our women with obesity, and this could be due to pathological obesity. Albumin levels were involved in appetite regulation because they could serve as carrier ligands for molecular substances, such as ghrelin, involved in stimulating appetite [21,22]. The same applies to PTH and PCT. Gender-related differences were not observed for WBC, neutrophils, and lymphocytes in the current study, unlike in previous findings [23], probably due to the pathological obesity condition. As demonstrated in literature, although its basis is still unclear, the A/G ratio was different by gender [24], but in this context, a lower ratio level (i.e., less albumin and more globulin) could be associated with a worse health condition in female patients. In the same health condition, higher beta globulin 2 was associated with peripheral arterial disease [25], and in our cohort, also with hypertension, that was a risk factor for vascular disorders [26]. ESR, instead, was associated with advancing age in females [27]. Glucose levels were lower in women, probably due to lower fasting plasma glucose in experiments conducted with the administration of glucose during the tolerance test [28]. We found lower triglyceride levels in female patients.
Estrogens could contribute to gender differences by directly influencing lipid metabolism through the suppression of gene expression and activity of lipoprotein lipase (LPL), the rate-limiting enzyme in triglyceride metabolism [29,30]; the modulation of lipolysis via an upregulation of α2-adrenergic receptors might contribute to lower TG concentrations [31] associated with the modulation of lipidic metabolism; females showed larger HDL particles than men, and this could justify the higher HDL levels [32].
Body composition differed between men and women. Men had more lean mass than women and were more likely to accumulate adipose tissue around the trunk and abdomen, whereas women usually accumulate adipose tissue around the hips and thighs [33]. According to these differences, women in this study had lower mean creatinine levels due to the lower muscle mass and meat intake [34] than men, who had a higher muscle energy metabolism than women of the same age [35,36]. Regarding serum levels of GGT, a higher level had been associated with a different distribution of fat in men compared to women [37,38]. Sex hormone fluctuations in females were responsible for lower iron and ferritin levels [39] that were also associated with lower concentrations of MCH, MCHC, PDW, and higher levels of transferrin. Platelets were also involved in these fluctuations, with increased platelets concentrations in females due to hormones or compensation due to menstruation [40,41].
Estradiol produced by the ovaries probably regulates the RBC levels, possibly affecting the nitric oxide (NO) synthesizing pathways [42].
Concerning lower FT3 levels in females, this finding could be due to lower testosterone levels than in males, which affect the peripheral (type 1) deiodinase as shown in a murine model [43,44], but it was also possible that testosterone played a role in the regulation of responses to thyroid hormone influenced by free dihydrotestosterone concentrations [45,46].
Genetic bases, on the other hand, likely justify the Hct levels, where the sequencebased polymorphism for erythropoietin could be partly responsible for this gender-based variation in Hct levels [47], while the immune response could justify the differences between genders regarding monocytes and eosinophils concentrations.
Obesity was well known to be responsible for a higher risk of steatosis [48], and we found by ultrasound examination that obesity was associated to steatosis and increased liver volume. We reported a sexually dimorphic disease and, interestingly, women had a reduced risk of steatosis compared to men of childbearing age, while after menopause women had a comparable prevalence of steatosis to men [49].
Our finding of the major prevalence of steatosis in males had been previously demonstrated [50]. Even if males had a higher prevalence of this disease in our cohort, worse clinical effects were demonstrated in females affected by obesity and steatosis. These results were apparently in contrast with those obtained in other studies performed in obese patients, where males had worse blood parameters [51], although the steatosis condition did not seem to influence the blood profile. Therefore, it seems that the effects of obesity per se were independent of those of steatosis and that the differences shown in this study were based on different genetic, hormone, or microbiota profiles. In fact, the axiom that "fatty people have fatty livers" was not correct because recent studies identified key molecular mediators involved in the progression of steatosis and obesity [52]. A balanced nutrition program diet was a good treatment in order to prepare the patient for surgery and a changed lifestyle.
Diet was important because it reduced liver volume, intraoperative complications, and operating time. Weight loss was also associated with reduction of respiratory failure and sleep apnea [53]. Furthermore, diet was the first step to keep psychological benefits and to try to understand if obese patients were intent to proceed to this new therapeutic path [54].

Conclusions
The present study shows gender differences in morbid obesity and related steatosis. Knowing the possible causes leading to these divergences in morbid obesity is important to try to reduce the impact of this discrimination on the psychological, social and metabolic health of obese patients. Understanding gender differences in obesity and steatosis status could have an important impact on the treatment of obesity and on planning community and individual interventions for the prevention and management of this condition.
The average age of the patients was low, and the real effect on their health status was not yet evident. In the future, it could be interesting to evaluate a follow-up of this un-explored cohort to gain more information on patients undergoing bariatric surgery and evaluate how this procedure could be effective, firstly for losing weight and subsequently for improving their lifestyle and state of health.
It would also be interesting to add the evaluation using instruments (DEXA or MRI) which evaluate the change in lean mass and fat mass in order to have a direct parameter on the efficacy not only of the dietary intervention but also of the surgical contribution.
The creation of personalized diets based on gender could increase efficacy because the dietary adherence was different by gender; women had greater sensitivity to the diet approach than men and ate a healthier diet. This different approach was reflected in the blood parameters, which improved after the diet.
Furthermore, the administration of psychological questionnaires could be useful to better understand the psychological status of patients eligible for bariatric surgery, compare the two genders and liver status, and to improve their emotional status.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nu15102381/s1, Table S1: Daily intake nutrients in different dietary regimen administered to the patient after to bariatric surgery.