Study population
In general, 210 participants were presented in this study. Based on the HBI score, all participants have divided into tertiles that the overall prevalence of them was 75 (35.7%) for tertile 1, 80 (38.1%) for tertile 2, and 55 (26.2%) for tertile 3. The mean (SD) age of participants was 36.09 (8.52) years and the mean BMI of participants was 30.77 (4.22) kg/m2. Forty percent of the participants used supplements. About 73% of the subjects were married, and 96.2% of them were employed. The economic status and education were such that 39% of the participants had a moderate economic status, and 48.1% of respondents were bachelor and higher. The mean (SD) of the total QoL score was 62.08 (29.05).
General characteristics among tertiles of HBI
The general characteristics of study participants categorized based on the HBI tertiles are presented in Table 1. According to this table, P-values for all variables were reported in the crude and adjusted model (adjusting with age, BMI, energy intake, and physical activity). In the crude model, the mean difference of neck circumference (NC) (P= 0.022), total cholesterol (TC) (P= 0.009), high density lipoprotein cholesterol (HDL_C) (P= 0.012), and low-density lipoprotein cholesterol (LDL_C) (P= 0.001) were statistically significant. After controlling for confounding variables, the mean difference of physical activity (PA) (P=0.045), height (P= 0.023), bone mineral content (BMC) (P=0.009), skeletal muscle mass (SMM) (P= 0.020), waist to hip ratio (WHR) (P= 0.032), and triglyceride (TG) (P= 0.020) became significant. In terms of HDL_C and LDL_C, after controlling with potential cofounders, the mean difference remained significant. According to Bonferroni post-hoc testing, the significant mean difference in BMC was between T1 with T2 and T2 with T3 where this mean difference was higher in T1 compared to T2, and also the mean difference of the second tertile was lower than the third. In terms of the fat-free mass index (FFMI), HDL_C, and TG, the mean differences were higher in T1 compared to T2 except in HDL_C where the mean difference of the second tertile was higher. The mean differences of SMM and WHR were higher in the second tertile compared to the first one, and finally, the mean differences of IPAQ and LDL_C were between T2 with T3, where the mean difference of T3 was higher than T2 in IPAQ but in terms of LDL_C, T2 was higher than T3. In categorical variables, the education status was statistically significant after controlling for cofounders. There was no significant difference for other variables in Table 1 (Table 1).
Dietary intakes among tertiles of the HBI in obese and overweight women
Dietary intakes of all participants among tertiles of HBI are presented in Table 2. In the crude model, mean difference of linolenic acid (P= 0.001), vitamin E (P= 0.010), biotin (P= 0.036), whole grains (P= 0.001), fruits (P= 0.022), and red meat (P= 0.041) was statistically significant. After adjusting with energy intake, the mean difference of monounsaturated fatty acid (MUFA) (P= 0.019), carbohydrate (P= 0.017), Potassium (P= 0.016), vitamin B6 (P= 0.027), Pantothenic acid (P= 0.055), and vitamin C (P= 0.041) changed to significant. Linolenic acid, vitamin E, biotin, whole grains, and fruits remained significant after adjusting with cofounders. Other variables in Table 2 had no significant relationship (Table 2).
Total QOL score and its components among HBI tertiles
The association of total QoL score items among HBI tertiles were shown in Table 3. In the crude model, general health (P= 0.073), physical functioning (P= 0.002), and total QoL score (P= 0.031) were statistically significant. After adjusting with age, energy intake, and BMI in model 1, the physical functioning had a significant mean difference among the tertiles of HBI (P= 0.041). In model 2, after adjusting with age, energy intake, BMI, and education status, the mean difference of physical functioning (P= 0.036) and total QoL score (P= 0.079) were statistically significant and marginally significant, respectively (Table 3).
Association between HBI tertiles with total QoL score
Crude and adjusted β and 95 % CI of total QoL score and its components across tertiles of HBI was shown in Table 4. In the crude model, there was a significant association between total QoL score with T2 tertile of HBI (β: 13.73, 95%CI: 3.25, 24.21, P-value= 0.010), in the model 2 adjustment (adjusted for age, energy intake, BMI, and education status), this significant association remained (β: 13.11, 95% CI: 1.52, 24.69, P-value= 0.027) and there was no significant trend based on tertiles in crude (P-trend= 0.463) and model 2 adjustment (P-trend= 0.429).
In terms of general health in the crude model, there was significant negative association in T2 (β: -5.83; 95% CI: -11.48, -0.18; P-value= 0.043) and T3 (β: -6.20; 95% CI: -12.37, -0.03; P-value= 0.049).
In the crude model, there was a significant negative association in term of the physical functioning component in T3 of HBI (β: -11.07; 95% CI: -17.25, -4.90; P-value= 0.001) and a significant decreasing trend was observed with increasing tertiles (P-trend=0.001). In the model 1 adjustment (adjusted for age, energy intake, and BMI), we observed a significant positive association between physical functioning with T3 of HBI (β: 7.51; 95% CI: 1.28, 8.73; P-value= 0.030). A significant increasing trend with rising tertiles was observed (P-trend= 0.042). After controlling potential confounders in model 2, a significant positive association have shown between physical functioning with T3 of HBI (β: 7.74; 95% CI: 0.53, 14.96; P-value= 0.025), which means women with more adherence of HBI, have the high physical functioning score. There was a significant upward trend with the movement from the second to the third tertile (7.74 vs 0.62) (P-trend=0.036).
There was a significant association between vitality and T3 of HBI in the crude (β: 3.07; 95% CI: 1.10, 3.96; P-value= 0.032) and adjusted model 2 (β: 2.03; 95% CI: 0.05, 5.98; P-value= 0.031). In terms of health transition items, in model 2 of adjustment, we observed a significant relationship between this component and HBI tertiles (β: 9.18; 95% CI: 2.20, 20.57; P-value= 0.014). Finally, in model 2 of adjustment, we observed that there was a significant positive association between mental health with T3 of HBI (β: 4.26; 95% CI: 1.51, 5.98; P-value= 0.015) and a significant increasing trend was observed with increasing tertiles (P-trend= 0.045) (Table 4).