Results1. Characteristics of the pa

Results
1. Characteristics of the participants according to serum zinc levels and the correlations between serum zinc levels and metabolic syndrome components The present study was conducted using a total of 1,926
participants. In this population, the prevalence of MetS was 26.4% (n = 248) in men and 26.4% (n = 260) in women. Mean serum zinc levels in men with and without MetS were 142.062.4mg/dL and
141.161.9mg/dL (p = 0.717), respectively, and in women with and without MetS were 127.562.5mg/dL and 129.661.9mg/dL (p = 0.419), respectively. Table 1 shows the characteristics of the study participants according to serum zinc level quartiles, in particular, Q1-3 versus Q4. In men, significant differences in age and fasting glucose and insulin levels were observed according to serum zinc levels, while age, systolic blood pressure, and insulin levels were higher in women in serum zinc level Q1-3 than in Q4.
In both men and women, significant negative correlations were observed between serum zinc levels and fasting glucose (for men: r=20.127, p = 0.003; for women: r =20.078, p = 0.045) and the
HOMA-IR index (r =20.120, p = 0.003 for men, r =20.113, p = 0.006 for women), and, in women, between serum zinc levels and systolic blood pressure (r =20.082, p = 0.015) and insulin levels (r =20.097, p = 0.023) (Table 2).

2. Mean metabolic syndrome component values
according to serum zinc level quartile The mean values of MetS components adjusted for age,
smoking, alcohol consumption, physical activity, BMI, and eGFR levels according to serum zinc level quartile are shown in Table 3. In men, as serum zinc levels increased, fasting glucose levels
decreased (p for trend = 0.013). HDL-cholesterol levels were not significantly different according to quartiles of serum zinc levels in both men and women (p = 0.398 and 0.308, respectively), but as
serum zinc levels increased, HDL-cholesterol levels showed a decreasing trend (p for trend = 0.088 and 0.083, respectively).

3. Associations between serum zinc levels and metabolic
syndrome and its components Unadjusted odds ratios (ORs), age-adjusted ORs (model 1), and
multivariate-adjusted ORs (model 2) of serum zinc levels according to the presence of MetS and its components are shown in Table 4. Men with elevated fasting glucose levels were more
likely to have low serum zinc levels than were those with normal fasting glucose levels (unadjusted OR 0.50, 95% confidence interval [CI] 0.33–0.77, p = 0.001), and this negative association
remained significant after adjusting for covariates (adjusted OR 0.58, 95% CI 0.36–0.93, p = 0.023). The multivariate-adjusted OR of serum zinc levels for elevated triglyceride levels in men was
1.47 (95% CI 1.01–2.13, p = 0.044). However, no significant association between MetS components and serum zinc levels was found in women. No association was detected between the presence of MetS and serum zinc levels in either men or women.

4. Serum zinc levels and the percentage of the highest
zinc level group (Q4) according to the number of metabolic syndrome components Figure 1 shows mean serum zinc levels, and the percentage of the highest zinc level group (Q4) according to the number of MetS components. After adjusting for age, smoking, alcohol drinking, physical activity, BMI, and eGFR levels, in women, a difference in serum zinc levels was observed based on the number of MetS
components (p = 0.002). Furthermore, in women with MetS (the number of MetS components: 3, 4 and 5), serum zinc levels showed a decreasing trend as the number of MetS components increased. In terms of the percentage of Q4, the difference showed according to number of MetS components, and the percentage of Q4 in women with MetS showed a decreasing trend as the number of MetS components increased (p = 0.050). No differences in mean serum zinc levels and the percentages of Q4 according to number of MetS components were observed in men (p = 0.727 and p = 0.741, respectively).

5. Percentages of participants according to the MetS
component combinations (MetS phenotypes) and serum zinc levels The participant distribution according to the MetS component combinations and serum zinc levels (Q1-3, and Q4) is shown in
Figure 2. In men with the MetS phenotype manifesting as increases in waist circumference, blood pressure and fasting glucose, the percentage of participants in Q4 was lower than in
Q1-3 (p = 0.021); on the other hand, in men with the MetS phenotype manifesting as increased waist circumference, elevated triglyceride, and reduced HDL-cholesterol, the percentage in Q4
was higher than in Q1-3 (p = 0.012). There were no significant differences in the percentage of men with other MetS phenotypes. Among women of almost every MetS phenotype, the percentage
of participants in Q4 was significantly lower than those in Q1-3, with the exception of the MetS phenotypes manifesting as increased waist circumference, elevated fasting glucose and
reduced HDL-cholesterol, and as increased waist circumference, elevated triglycerides and reduced HDL cholesterol.

Discussion
We investigated the associations of serum zinc levels with MetS or its metabolic risk factors in Korean adults. The results of this study showed associations between serum zinc levels and certain
MetS components. Serum zinc levels in men were negatively associated with elevated fasting glucose and positively associated with elevated triglycerides. In both men and women, as serum zinc
levels increased, HDL-cholesterol levels showed a decreasing trend. Although there was no significant association found between serum zinc levels and the prevalence of MetS in either men or women, there were differences in serum zinc levels