Effect of 8 Weeks Aerobic Training and Saffron Supplementation on Inflammation and Metabolism in Middle-Aged Obese Women with Type 2 Diabetes Mellitus

Background: This study aimed to investigate the effects of 8-week aerobic training (AT) and saffron supplementation on inflammation and metabolism in middle-aged obese women with type 2 diabetes mellitus (T2DM). Methods: Thirty-two obese women with T2DM were randomly divided into four groups (n = 8 in all groups): saffron + training (ST), placebo + training (PT), saffron supplementation (SS), and placebo (P). The ST and PT groups performed eight weeks of aerobic training (AT) (three sessions/week at 60–75% HRmax). A daily dose of 400 mg saffron powder was consumed by the ST and SS groups for 8 weeks. Blood samples were taken after 12 h of fasting, 48 h before the first AT session, 48 h and two weeks after the last AT session. Results: AT, saffron supplementation, and their combination affected body mass index (BMI), homeostatic model assessment for insulin resistance (HOMA-IR), and serum levels of insulin, adiponectin, interleukin-6 (IL-6), high-density lipoprotein cholesterol (HDL-C), cholesterol, and triglyceride (TG) (p < 0.05). However, body weight, body fat percentage, and serum levels of glucose, resistin, tumor necrosis factor-alpha (TNF-α), irisin, and low-density lipoprotein cholesterol (LDL-C) showed significant changes in the ST group only (p < 0.05). In addition, a significant difference was seen between all factors in post-training and follow-up in the ST group (p < 0.05). Conclusions: Saffron supplementation at a dose of 400 mg/day, when combined with AT, could improve inflammation, metabolism, glycemic status, and lipid profile in T2DM patients, and these changes are sustainable at up to 2 weeks of detraining.


Introduction
Diabetes mellitus (DM) is a global health problem mainly caused by a sedentary lifestyle, poor dietary habits, and obesity [1][2][3]. Obesity can change adipokine secretion and lead to insulin resistance; a close connection has been reported between obesity and type 2 diabetes mellitus (T2DM) [4]. Adiponectin, the most crucial insulin-sensitizing, anti-inflammatory adipokine, and resistin, a newly discovered adipokine directly related to insulin resistance, are the main adipokines influenced by obesity [5,6]. Adiponectin reduces glucose output from the liver by increasing liver sensitivity to insulin [4,6] and increases glucose uptake in the muscles, thus reducing blood glucose [7]. Similar to adiponectin,

Materials and Methods
This was a quasi-experimental study with repeated measures design: 1: pre-test (48 h before the first AT session), 2: post-test (48 h after the last AT session), and 3: follow-up (2 weeks after the last AT session). Participants consisted of 32 middle-aged obese women with T2DM who lived in Kermanshah Province, Iran. They were selected by convenience sampling method according to the inclusion criteria. The inclusion criteria included no cardiovascular and musculoskeletal disorders, glycosylated hemoglobin lower than 9.9%, no diabetic complications (neuropathy, nephropathy, or retinopathy), no regular AT, no smoking, diabetes history less than 5 years, and a maximum of one type of oral anti-diabetic Sports 2022, 10, 167 3 of 15 tablet a day. Participants with blood glucose levels above 120 mg/dL were considered DM. Participants were divided into four groups: saffron + training (ST) (n = 8), placebo + training (PT) (n = 8), saffron supplementation (SS) (n = 8), and placebo (P) (n = 8). The goals and procedure of the study were explained, and written consent was obtained from the participants. During this study, there were no significant changes in the prescription of drugs to control blood glucose or lipids, as reported by the participants' physicians. The Ethics Committee approved the current study of Mohaghegh Ardabili University (Ethics No: 1605/A. D; date 6 November 2017).

Preparation and Administration of Saffron and Placebo Capsules
Powdered saffron (400 mg) [16] (Food and Drug Organization of Iran Ministry of Health ID: 1021/50 and 111191/50) was placed in capsules and used for two months by the ST and SS groups (once a day at 6 pm). Similar placebo capsules containing 400 mg of wheat flour were prepared for the P and PT groups with the same color and shape as the saffron capsule. All participants were asked to use no drug, except metformin, during the study if possible to control confounding and interfering factors. Major metabolites of saffron were determined using high-performance liquid chromatography.

Controlling Participant Diet
Nutritional data were collected using the 24 h recall diet (two non-holiday days a week and one day at the weekend to determine the mean of nutrient intake). All participants were asked to list all foods and beverages consumed within the last 24 h. All participants completed the questionnaire 30 non-consecutive times (three times a week during the study).

Training Program
An 8 week AT program was performed by the ST and PT groups. Each training session consisted of 10 min warm-up at 55% of target heart rate (THR), 10 min of combined hand-leg movements at 55-60% THR, 30 min of running at 60-75% of THR, and 5 min cool down at 50% THR. Target heart rate was calculated using Karvonen's formula. THR = [(MHR − RHR) × %Intensity] + RHR MHR: heart rate maximum, RHR: resting heart rate. Participant heart rate was monitored using a polar watch during AT sessions (Polar Electro Oy Professoriate 5, Polar, US6584344, FI-90440 Kempele, Finland). Due to the lack of regular AT and the low physical fitness of participants, the AT program consisted of fast walking in the first two weeks (intensity of 60% THR, duration of 15 to 20 min). The intensity and duration gradually increased each week for the remaining 6 weeks ( Table 1). All participants were encouraged to walk and combine hand-leg movements until the end of the session. The P and SS groups did no AT during this period.

Measurement of Variables
A body composition analyzer (In Body version 3.0, made in Seoul, South Korea) was used to measure body weight, Body mass index (BMI), and body fat percentage. Each participant stood on the device without any metal objects and with minimum clothing holding the device electrodes. Measurement was performed at least 3 h after breakfast. Blood sampling was conducted after 10 h of fasting (10 cc venous blood sample from the left radial artery) in 3 stages: pre-test (48 h before the first AT session), post-test (48 h after the last AT session), and follow-up (2 weeks after the last AT session). Serum levels of adiponectin, resistin, IL-6, TNF-α, and irisin were calculated using their optical density at 450 nm by an ELISA reader and using standard curves and applying dilution markers. The ELISA kits were used as follows: adiponectin (Mercodia, Stockholm, Sweden), TNFα (Orgenium, Helsinki, Finland), IL-6 (Boster, Pleasanton, CA, USA), resistin and irisin (Cusabio, Beijing, China), insulin (Diaplus, North York, ON, Canada), and glucose (Pars Azmun, Tehran, Iran). Serum levels of lipid profile-cholesterol, triglyceride (TG), low-Sports 2022, 10, 167 4 of 15 density lipoprotein-cholesterol (LDL-C), and high-density lipoprotein-cholesterol (HDL-C) were also measured enzymatically (Pars Azmun, Tehran, Iran). The insulin resistance index was calculated using homeostatic model assessment for insulin resistance (HOMA-IR) with the following formula [25]: HOMA-IR = fasting blood glucose (mmol/L) × fasting insulin (µu/mL)/22.5

Statistical Analysis
Data were expressed as mean ± standard deviation (SD). Normal data distribution and homogeneity of variances were assessed using the Shapiro-Wilk and Levene tests, respectively. A mixed-model ANOVA with repeated measures across time and group was used to test the main effects and interactions. The Bonferroni post-hoc test was used for checking significant differences among the main effects of each dependent variable. Cohen's d test was used to estimate the effect size. An effect size of less than 0.2 was considered negligible, between 0.2 and 0.5 small, between 0.5 and 0.8 moderate, and more than 0.8 high. Statistical analysis was performed using SPSS version 22 at a significance level of 5%. Table 2, there was no significant difference between the groups regarding age, energy, and nutrient intake.
weight and body fat percentage, ɸ indicates a significant decrease in the ST group compared to the P group. For BMI ɸ, $, #, and * indicate a significant decrease in the ST group compared to the P group, the ST group compared to the SS group, the PT group compared to the P group, and the SS group compared to the P group, respectively. Cohen's d test was used to estimate the effect size. BMI: body mass index.
The results of mixed ANOVA showed a significant interaction between groups and time for glucose (F 6, 56 = 22.88, p < 0.001) (Figure 1), insulin (F 6, 56 = 11.52, p < 0.001) (Figure 2), and HOMA-IR (F 6, 56 = 19.34, p < 0.001) (Figure 3).     The results of mixed ANOVA showed that there was a significant interaction between groups and time in LDL (F 6, 56 = 15.42, p < 0.001) (Figure 4), HDL-C (F 6, 56 = 24.91, p < 0.001) ( Figure 5), cholesterol (F 6, 56 = 5.72, p < 0.001) (Figure 6), and TG (F 6, 56 = 13.65, p < 0.001) (Figure 7). Figure 2. Insulin in the pre-test, post-test, and follow-up (mean ± SD). $ indicates a significant decrease in the ST, PT, and SS groups compared to the P group, and # indicates a significant decrease in the ST group compared to the SS group. The results of mixed ANOVA showed a significant interaction between groups and time for adiponectin (F 6, 56 = 11.39, p < 0.001), resistin (F 6,56 = 7.55, p < 0.001), TNF-α levels (F 6, 56 = 5.87, p < 0.001), IL-6 levels (F 6, 56 = 13.86, p < 0.001), and irisin (F6, 56 = 9.73, p < 0.001) ( Table 5).   ST: Training + saffron, PT: Training + placebo, SS: Saffron supplementation, and P: Plac levels of adiponectin, resistin, TNF-α, IL-6, and irisin in the pre-test, post-test, and foll adiponectin, & indicates a significant increase in the ST group compared to the SS and indicates a significant increase in the PT group compared to the P group, # indicates a increase in the SS group compared to the P group, and $ indicates a significant increas group compared to the PT, SS, and P groups in the follow-up. For resistin, $ indicates a decrease in the ST group compared to the PT, SS, and P groups. For TNF-α, ɸ indicates a decrease in the ST group compared to the P group. For IL-6, $ indicates a significant dec ST group compared to the PT, SS and P groups, * indicates a significant decrease in th compared to the P group, # indicates a significant decrease in the SS group compare group, and ɸ indicates a significant increase in the ST group compared to the P group low-up. For irisin, ɸ indicates a significant increase in the ST group compared to the P g The results of mixed ANOVA showed that there was a significant inter tween groups and time in LDL (   . LDL-C levels in the pre-test, post-test, and follow-up. $ indicates a significant decrease the ST group compared to the P group. $ indicates a significant decrease in the PT group compare to the P group.

Discussion
The present study aimed to investigate the effects of 8-week AT and saffron su plementation on inflammation and metabolism in middle-aged obese women wi T2DM. A growing body of evidence has revealed that saffron supplementation could b considered an essential nutraceutical for T2DM treatment due to its hypoglycemic an anti-inflammatory properties [21,22]. Inhibition of oxidative stress, amelioration of ß-ce function, repression of inflammatory pathways, improvement in insulin sensitivity, an stimulation of GLUT-4 expression could be affected by saffron supplementatio [21,22,26]. Our results showed that when AT was combined with saffron supplement tion, adiponectin, irisin, and HDL-C increased, but resistin, TNF-α, IL-6, body weigh BMI, body fat percentage, LDL-C, cholesterol, blood glucose, insulin, and HOMA-IR d creased. This means that combining AT with saffron supplementation could reduce i flammation, improve metabolism, and ameliorate insulin resistance and body compos tion.
The present study results indicated that insulin levels and resistance decreased aft eight weeks of AT and saffron supplementation, but their combination led to better r sults. These results are consistent with several studies [15,27] but inconsistent with som others [28,29]. For instance, in a study by Choi et al. [29], where the participants we obese women who performed 45 min/session, 300 kcal/session AT, and 20 min/sessio

Discussion
The present study aimed to investigate the effects of 8-week AT and saffron supplementation on inflammation and metabolism in middle-aged obese women with T2DM. A growing body of evidence has revealed that saffron supplementation could be considered an essential nutraceutical for T2DM treatment due to its hypoglycemic and antiinflammatory properties [21,22]. Inhibition of oxidative stress, amelioration of ß-cell function, repression of inflammatory pathways, improvement in insulin sensitivity, and stimulation of GLUT-4 expression could be affected by saffron supplementation [21,22,26]. Our results showed that when AT was combined with saffron supplementation, adiponectin, irisin, and HDL-C increased, but resistin, TNF-α, IL-6, body weight, BMI, body fat percentage, LDL-C, cholesterol, blood glucose, insulin, and HOMA-IR decreased. This means that combining AT with saffron supplementation could reduce inflammation, improve metabolism, and ameliorate insulin resistance and body composition.
The present study results indicated that insulin levels and resistance decreased after eight weeks of AT and saffron supplementation, but their combination led to better results. These results are consistent with several studies [15,27] but inconsistent with some others [28,29]. For instance, in a study by Choi et al. [29], where the participants were obese women who performed 45 min/session, 300 kcal/session AT, and 20 min/session, 100 kcal/day strength training, five times a week for three months, a decrease in insulin levels and insulin resistance was not seen. Different diets, the age and sex of participants, and the type of AT protocol may explain this inconsistency [30].
Blood glucose levels only decreased in the ST group. This means that neither AT nor saffron supplementation could affect blood glucose alone. In line with this result, Cauza et al. [31] and Bello et al. [32] did not report significant changes in blood glucose after 16 and 8 weeks of AT, respectively. It has been suggested that AT cannot significantly affect blood glucose in the middle-aged and the elderly [33]. In addition, our participants had a long history of diabetes and a high level of insulin resistance so they may have needed a longer AT period. These two reasons may explain why we could not see significant changes in the PT or SS groups. However, combining AT and saffron supplementation pro-vided sufficient stimulation for upregulating AMPK signaling, which resulted in increasing GLUT-4 translocation into the cell membrane and therefore increased glucose uptake and insulin sensitivity [34].
In line with other studies [35,36], the present study showed that while eight weeks of AT and saffron supplementation increased serum adiponectin levels, combining these interventions led to better results. A study by Rezaeeshirazi [36] showed that AT is more effective in ameliorating insulin resistance in T2D. Yokoyama et al. [37] reported that improvement in insulin sensitivity induced by diet combined with AT (40 min/session, five sessions/week cycling, and 10,000 steps/session walking) for three weeks could not alter adiponectin levels in T2DM patients, which is inconsistent with our results. The reasons for the inconsistency may be the different types of AT and participants' sex [30] (i.e., female in our study versus both male and female in Yokoyama et al.); high levels of testosterone in men compared to women could inhibit production and secretion of adiponectin [38]. The increase in adiponectin levels in the ST group could be explained by the association of adiponectin with lipid profile (LDL-C, HDL-C, cholesterol, and TG). Adiponectin concentration is directly related to HDL-C levels and inversely related to LDL-C and TG levels [39,40]. This study showed that LDL-C, cholesterol, and TG decreased, and HDL-C increased after combining eight weeks of AT with saffron supplementation.
Furthermore, increased plasma-free fatty acids can stimulate adiponectin secretion. Aerobic training, by increasing lipolysis [39,40] and crocin (one of saffron's components) by inhibiting pancreatic lipase, which indirectly leads to malabsorption of fat, could lead to an increase in plasma-free fatty acids and stimulate adiponectin secretion [41]. In addition, decreasing insulin resistance might be another reason for increasing adiponectin. Based on this result, we concluded that obesity and T2DM are associated with a decrease in adiponectin, which reduces insulin sensitivity. Conversely, the enhancement of serum levels of adiponectin resulting from AT and saffron supplementation could increase insulin sensitivity through insulin-mimicking, insulin-sensitizing, and anti-inflammatory properties [42,43].
Resistin is another crucial link between insulin resistance and obesity. Our results showed no significant decrease in serum resistin levels after eight weeks of AT and saffron supplementation alone but did show a decrease when they were used in combination. Kadoglou et al. [44] reported that after 16 weeks of AT (i.e., 45-60 min running with 50-85% of vo 2max , four sessions/week), serum levels of resistin decreased in T2DM patients. However, no significant changes were reported in other studies after AT [45,46]. Different AT types, durations, and intensities might explain this inconsistency [30]. Many studies have shown that resistin is directly related to changes in body fat, body weight, BMI, HOMA-IR, and TG [47], which is in line with our results since all the above factors showed significant changes in the ST group. Another important factor is diet, as it is the primary regulator of resistin. This is why we supervised the participants' diets, where we observed no significant change in nutrient uptake.
Another insulin resistance influencer is an adipo-myokine, irisin, which has a vital role in energy homeostasis and metabolism [48]. While eight weeks of AT and saffron supplementation could not affect irisin serum levels alone, its levels showed a significant increase in the ST group. This result showed that the isolated effect of AT or saffron supplementation is not enough to increase irisin levels, but the combined effects are needed. However, some studies showed that AT could increase irisin serum levels in diabetic patients [49,50]. Although some studies have suggested a positive correlation between plasma irisin and lactate levels [51], our AT protocol did not have enough load to induce changes in irisin levels.
Our results also showed that while both AT and saffron supplementation could reduce serum IL-6 levels, combining these interventions led to better results. The antioxidant properties of saffron [22] and AT [52] together can reduce reactive oxygen species levels, which play an essential role in the production of pro-inflammatory cytokines [53]. In fact, crocin improves the expression of phosphorylated Akt (p-Akt) and phosphatidylinositol 3-kinase (PI3K). PI3K/Akt signaling plays a significant role in blocking oxidative stress and suppressing the pro-inflammatory response to oxidative stress [54]. The effectiveness of 8 weeks of AT on IL-6 serum levels can also be due to body weight loss, mainly due to the reduced adipose tissue content [55]. IL-6 is secreted by adipose tissue as adipokines, and a decrease in fat mass is associated with a decrease in IL-6 levels [55]. Some studies have reported no significant changes in IL-6 levels after AT [56,57], but others have shown decreased inflammatory and increased anti-inflammatory cytokines in T2DM patients [58].
Another pro-inflammatory cytokine, TNF-α, showed a significant reduction in the ST group only, indicating that it may need a higher amount of AT or saffron to induce isolated changes. In other words, IL-6 seems to be more sensitive to our interventions than TNF-α. The lipid profile, including LDL-C, cholesterol, and TG serum levels, showed a significant reduction in the ST group, and HDL-C showed a significant increase in the ST, SS, and PT groups with no significant difference between them. However, an additive effect was seen for cholesterol and TG highlighting their sensitiveness to AT and saffron. In line with this study, other studies have shown that both saffron and AT [59,60] could improve the lipid profile. Based on previous evidence, saffron supplementation causes an increase in the activity of lecithin-cholesterol acyltransferase (LCAT) [61], which is involved in the regulation of blood lipid metabolism [62]. In addition, regular AT leads to increased daily energy output and fat oxidation and reduces fat mass [63]. As our results show, body weight, BMI, and body fat percentage significantly decreased after eight weeks of AT with saffron supplementation.
The follow-up results (after two weeks) showed that the adaptations disappeared in the PT, SS, and P groups but not in the ST group. It seems that the adaptations in the ST group were high enough to still be detectable two weeks after detraining.

Conclusions
Saffron supplementation at a dose of 400 mg/day, when combined with AT, could improve inflammation, metabolism, glycemic status, and lipid profile in T2DM patients, and these changes are sustainable at up to 2 weeks of detraining.

Limitations and Strength
We gave the saffron/placebo capsules to participants and asked them to take them according to the study plan and followed up with them through phone call. This could be a potential limitation in our study, and a future study should ask the participants to take the supplement in front of the researcher. In addition, we collected participants' nutritional data three days per week, which might have led to a misinterpretation of the participants' diets. What we consider the strength in our study is using a free-style rather than traditional training method. Furthermore, measuring various factors that affect metabolism and inflammation helped us to better interpret our data, which could be another strength of this study.

Conflicts of Interest:
The authors declare no conflict of interest.