DIABETICMedicine DOI: 10.1111/dme.12643
Research: Epidemiology Low transferrin saturation is associated with impaired fasting glucose and insulin resistance in the South Korean adults: the 2010 Korean National Health and Nutrition Examination Survey R. J. Park1,2 and J. D. Moon3 1 Department of Occupational and Environmental Medicine, Gwangyang Sarang General Hospital, Gwangyang-si, Jeollanam-do, 2Department of Medicine, Graduate School of Chonnam National University, Gwangju and 3Department of Occupational and Environmental Medicine, Chonnam National University Medical School, Gwangju, Republic of Korea
Accepted 25 November 2014
Abstract Aims The associations of transferrin saturation with diabetes have not been well evaluated and conflicting results have been reported. The purpose of this study is to examine the association of iron indices (serum ferritin and transferrin saturation) with risk of impaired fasting glucose and insulin resistance. Methods We conducted a cross-sectional study in 2413 individuals (1150 men and 1263 women) aged 20–50 years who participated in the 2010 Korean National Health and Nutrition Examination Survey. Participants were free of diabetes, malignancy, liver cirrhosis, chronic renal failure, anaemia, pregnancy and menopause. Fasting plasma glucose, insulin and the homeostasis model assessment of insulin resistance (HOMA–IR) were measured as the outcomes.
Impaired fasting glucose was more prevalent in the highest compared with the lowest serum ferritin quartile among men (odds ratio [OR], 1.97; 95% confidence interval [CI], 1.20–3.24) after adjustment for multiple covariates. Following the same adjustment, impaired fasting glucose was less prevalent in the highest compared with the lowest transferrin saturation quartile among men (OR, 0.45; 95% CI, 0.25–0.80) and women (OR, 0.33; 95% CI, 0.14–0.77). Moreover, a higher ferritin level was significantly associated with higher HOMA–IR after adjusting for confounders in men. Lower transferrin saturation was also significantly associated with higher insulin levels and HOMA–IR in both sexes.
Results
Conclusions Lower transferrin saturations were associated with an increased risk of impaired fasting glucose and insulin resistance among general South Korean population.
Diabet. Med. 32, 673–678 (2015)
Introduction Iron is vital for humans because it is essential for many metabolic processes, including oxygen transport, DNA synthesis and electron transport [1]. However, iron can also damage tissues via the generation of oxygen free radicals that attack cellular membranes, proteins and DNA [2]. Thus, balancing the deleterious and beneficial effects of iron is an essential aspect of cell survival. Ferritin is responsible for the storage of iron and transferrin is responsible for the transport of iron from sites of absorption to sites of use [3]. It is well known that the serum
ferritin concentration closely parallels body iron reserves [3]. However, serum ferritin concentrations reflect not only body iron stores, but also inflammation [4], and transferrin saturation is used together with serum ferritin concentration to estimate the true iron status [5]. Previous studies have reported that increased serum ferritin concentrations are associated with hyperinsulinaemia, reduced insulin function and diabetes [6–11]. However, the associations between transferrin saturation and diabetes have not been well evaluated and conflicting results have been reported [5,9,12–17]. This study was performed to examine the association of iron indices (serum ferritin and transferrin saturation) with risk of impaired fasting glucose and insulin resistance and to
Correspondence to: Jai Dong Moon. E-mail: [email protected]
© 2014 The Authors. Diabetic Medicine © 2014 Diabetes UK
673
DIABETICMedicine
Transferrin saturation and insulin resistance in Korean adults R. J. Park and J. D. Moon
What’s new?
Data collection
• The associations of iron indices (serum ferritin and transferrin saturation) with impaired fasting glucose and insulin resistance were examined.
Anthropometric measurements of participants were conducted at local community health centres and clinics as a part of the health examination. The anthropometric variables (height and body weight) were recorded using a standard protocol. BMI was calculated as weight in kilograms divided by height in metres squared. Waist circumference was measured midway between the costal margin and iliac crest at the end of a normal expiration. In accordance with the Korean Society for the Study of Obesity criteria for waist circumference [19], abdominal obesity was defined as a waist circumference 90 cm in men and 85 cm in women. Information on the smoking and drinking status of the participants was obtained from a questionnaire survey. People who reported having smoked fewer than 100 cigarettes in their lives were classified as having never smoked, and the remainder were asked whether they were current or former smokers. Individuals who had 12 alcohol drinks or more per year were considered as drinkers. Level of physical activity was calculated using metabolic equivalent of task values (MET) based on self-reported frequency and duration of activities during the week. MET-minutes per week was computed by multiplying the MET value of a particular activity (walking = 3.3 METs, moderate physical activity = 4.0 METs and vigorous physical activity = 8 0 METs) with minutes spent in that particular activity [20]. Total weekly physical activity was calculated by summing MET-minutes per week of walking, moderate and vigorous activity. Participants who had 3000 MET-minutes per week were classified as vigorous physical activity. Blood pressure was measured with participants in the sitting position following a 5–minute rest period. Blood pressure was measured on three occasions with a mercury sphygmomanometer on the right arm and averaged for a final blood pressure reading. Mean arterial pressure (mm/Hg) was calculated as 2/3 diastolic blood pressure + 1/3 systolic blood pressure. Blood samples were obtained in the morning following an overnight fast. Levels of fasting plasma glucose, total cholesterol, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase and serum creatinine were measured (Hitachi Automatic Analyzer 7600, Hitachi, Tokyo, Japan). Haemoglobin and white blood cell count were measured using the sodium lauryl sulfate haemoglobin method (XE-2100D, Sysmex, Kobe, Japan) and laser flow cytometry (XE-2100D, Sysmex), respectively. Serum ferritin and insulin were measured by immunoradiometric assay (1470 Wizard gamma counter, PerkinElmer, Turku, Finland). Serum iron and iron-binding capacity were measured using the bathophenanthroline direct method (Hitachi Automatic Analyzer 7600, Hitachi). Transferrin saturation is
• Study participants included 2413 Korean individuals aged 20–50 years. • Lower transferrin saturations were associated with an increased risk of impaired fasting glucose and insulin resistance.
determine whether the association differed between men and women using a national representative sample of Korean adults who were free of comorbidities.
Participants and methods Study population
The study used data from the 2010 Korean National Health and Nutrition Examination Survey (KNHANES). A stratified multistage clustered probability design was used to select representative samples of non-institutionalized Korean civilians for the KNHANES [18]. The survey data were publicly available. This survey was completed by 8958 participants (81.9% of the total target population of 10,938). Our analysis was restricted to participants aged 20–50 years for whom fasting glucose (8 h fasting) levels were available (n = 2971). We did not study participants with diabetes (n = 99, defined as a plasma glucose level 7 mmol/l after fasting for a minimum of 8 h or physiciandiagnosed diabetes or self-reported current use of oral hypoglycaemic medication or insulin). Other exclusion criteria included malignancy (n = 16), liver cirrhosis (n = 2), chronic renal failure (n = 0), anaemia (n = 254, defined as haemoglobin < 13 g/dl for men and < 12 g/dl for women), pregnancy (n = 33) or menopause (n = 60, defined as self-reported cessation of menstruation for > 1 year or hysterectomy). Individuals with exceptionally increased serum ferritin levels (> 500 ng/ml; n = 3), increased serum liver enzyme values (aspartate aminotransferase 80 U/l and alanine aminotransferase 70 U/l; n = 7), increased serum creatinine level (> 132 lmol/l; n = 0), increased white blood cell counts (> 12.0 9 109/l; n = 14) or missing values for exposure variables (n = 70) were not studied. After these exclusions, 2413 individuals (1150 men and 1263 women) were included in our final analysis. The study was approved by the Korea Centers for Disease Control and Prevention Institutional Review Board. Written informed consent was obtained from all study participants.
674
© 2014 The Authors. Diabetic Medicine © 2014 Diabetes UK
Research article
the ratio of serum iron and total iron-binding capacity, multiplied by 100. All clinical analyses were performed by the Neodin Medical Institute, a laboratory certified by the Korean Ministry of Health and Welfare. Impaired fasting glucose was defined as a fasting plasma glucose of 5.5–6.9 mmol/l [21]. Insulin resistance was determined using the homeostasis model assessment of insulin resistance (HOMA–IR = fasting insulin [in lU/ml] 9 fasting glucose [mmol/l]/22.5) and fasting insulin levels [22].
Statistical analysis
We categorized the level of ferritin and transferrin saturation into quartiles (ferritin: < 69, 69–105, 106–157, 157 for men and < 17, 17–28, 29–47, 47 for women; transferrin saturation: < 30, 30–41, 42–53, 53 for men and < 24, 24–33, 34–43, 43 for women). The characteristics of the study sample were analysed using t-tests for continuous variables and chi-square tests for categorical variables. All analyses were performed separately for men and women. We obtained the odds ratio (OR) and 95% confidence interval (CI) for impaired fasting glucose using a multiple logistic regression analysis after adjusting for confounding variables across serum ferritin and transferrin saturation quartiles for both sexes. The covariates for the adjusted OR calculation were age, drinking status, smoking status, physical activity, abdominal obesity, BMI, mean arterial pressure, total cholesterol, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, white blood cell count and serum creatinine. We also conducted association analyses between iron indices and insulin resistance (fasting insulin, HOMA– IR) using analysis of covariance (ANCOVA) models in the same adjustment model. Two-tailed P-values of < 0.05 were considered statistically significant. Variables with skewed distributions were logtransformed before analysis. All statistical analyses were conducted using the SPSS statistical software package v. 18.0 (SPSS Inc., Chicago, IL, USA), which incorporates sample weights and adjusts the analyses for the complex sample design of the survey. Survey sample weights were calculated taking into consideration the sampling rate, response rate and age/sex proportion of the reference population (2010 National Census Registry).
Results Among these 2413 participants, 315 had impaired fasting glucose. Participant characteristics according to glucose regulation status are presented in Table 1. Those with impaired fasting glucose were older and more likely to be abdominally obese. In both sexes, they also had higher mean BMI, fasting glucose level, insulin concentration, HOMA– IR, mean arterial pressure, total cholesterol, aspartate aminotransferase and alanine aminotransferase. There was
© 2014 The Authors. Diabetic Medicine © 2014 Diabetes UK
DIABETICMedicine
a higher mean ferritin concentration in men and lower transferrin saturation in women. The prevalence of impaired fasting glucose increased significantly according to the increasing serum ferritin quartiles in men and decreased significantly according to the increasing transferrin saturation in women (Table 2). As shown in Table 3, impaired fasting glucose was more prevalent in the highest compared with the lowest serum ferritin quartile among men (OR, 1.97; 95% CI, 1.20–3.24) after adjustment for age, drinking status, smoking status, physical activity, abdominal obesity, BMI, mean arterial pressure, total cholesterol, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, white blood cell count and serum creatinine. Following the same adjustment, impaired fasting glucose was less prevalent in the highest compared with the lowest transferrin saturation quartile among men (OR, 0.45; 95% CI, 0.25–0.80) and women (OR, 0.33; 95% CI, 0.14–0.77). A higher ferritin level was significantly associated with higher HOMA–IR after adjusting for confounders in men (Table 4). Lower transferrin saturation was also significantly associated with higher insulin levels and HOMA–IR in both sexes.
Discussion To our knowledge, this is the first study examining the associations between transferrin saturation and impaired glucose metabolism among Korean population. In this large nationally representative cross-sectional study, we found that lower transferrin saturations were associated with impaired fasting glucose and insulin resistance. Our data corroborated previous observations. Cheung et al. [5] suggested that higher ferritin and lower transferrin saturation were associated with a higher risk of prediabetes in a general US population. A Chinese study [9] reported a similar magnetic resonance imaging-estimated hepatic iron concentration, but higher serum ferritin levels and lower transferrin saturation in the diabetes group than in the prediabetes group, which implied a decreased circulating iron in diabetes even though hepatic iron stores remained higher. We observed a trend toward increasing insulin resistance (HOMA–IR) with increasing serum ferritin levels in men (median ferritin: 105 ng/ml), but not in premenopausal women (median ferritin: 28 ng/ml). Similarly, Lee et al. [10] reported that an elevated serum ferritin level was significantly associated with insulin resistance in men (median ferritin: 96 ng/ml) and postmenopausal women (median ferritin: 54 ng/ml), but not in premenopausal women (median ferritin: 26 ng/ml). Kim et al. [11] found that HOMA–IR was associated with increased serum ferritin concentration in men (mean ferritin: 167 ng/ml), but not in women (mean ferritin: 65 ng/ml). These observations may be because serum ferritin concentrations of premenopausal women were significantly lower than men. The difference is
675
Transferrin saturation and insulin resistance in Korean adults R. J. Park and J. D. Moon
DIABETICMedicine
Table 1 General and clinical characteristics according to glucose regulation status Men [1,150/10,044,496]*
Variable
Normal fasting glucose [940/8,360,696]
Current alcohol intake (%) Current smoking (%) Vigorous physical activity (%) Abdominal obesity (%) Age (years) BMI (kg/m2) Fasting plasma glucose (mmol/l) Insulin (pmol/l) HOM–-IR Mean arterial pressure (mm/Hg) Total cholesterol (mmol/l) Aspartate aminotransferase (U/l) Alanine aminotransferase (U/l) Alkaline phosphatase (U/l) White blood cell count (9 109/l) Serum creatinine (lmol/l) Ferritin (ng/ml) Transferrin saturation (%)
81.0 53.4 32.4 19.5 33.33 23.85 4.92 69.76 2.21 88.08 4.74 22.33 25.28 225.82 6.51 84.43 114.07 42.95
1.6 1.7 1.9 1.6 0.40 0.12 0.01 1.24 0.04 0.46 0.03 0.47 0.55 2.05 0.05 0.41 2.86 0.77
Women [1,263/8,062,370]
Impaired fasting glucose [210/1,683,800] 85.9 50.5 27.6 35.8 40.31 25.64 5.91 88.02 3.36 94.44 5.28 26.72 33.18 227.59 6.63 83.03 146.19 41.22
3.0 4.7 3.7 3.5 0.58 0.33 0.03 5.36 0.23 1.03 0.09 1.06 1.66 4.15 0.11 0.89 7.88 1.25
P
< < < < < < < <
Research: Epidemiology Low transferrin saturation is associated with impaired fasting glucose and insulin resistance in the South Korean adults: the 2010 Korean National Health and Nutrition Examination Survey R. J. Park1,2 and J. D. Moon3 1 Department of Occupational and Environmental Medicine, Gwangyang Sarang General Hospital, Gwangyang-si, Jeollanam-do, 2Department of Medicine, Graduate School of Chonnam National University, Gwangju and 3Department of Occupational and Environmental Medicine, Chonnam National University Medical School, Gwangju, Republic of Korea
Accepted 25 November 2014
Abstract Aims The associations of transferrin saturation with diabetes have not been well evaluated and conflicting results have been reported. The purpose of this study is to examine the association of iron indices (serum ferritin and transferrin saturation) with risk of impaired fasting glucose and insulin resistance. Methods We conducted a cross-sectional study in 2413 individuals (1150 men and 1263 women) aged 20–50 years who participated in the 2010 Korean National Health and Nutrition Examination Survey. Participants were free of diabetes, malignancy, liver cirrhosis, chronic renal failure, anaemia, pregnancy and menopause. Fasting plasma glucose, insulin and the homeostasis model assessment of insulin resistance (HOMA–IR) were measured as the outcomes.
Impaired fasting glucose was more prevalent in the highest compared with the lowest serum ferritin quartile among men (odds ratio [OR], 1.97; 95% confidence interval [CI], 1.20–3.24) after adjustment for multiple covariates. Following the same adjustment, impaired fasting glucose was less prevalent in the highest compared with the lowest transferrin saturation quartile among men (OR, 0.45; 95% CI, 0.25–0.80) and women (OR, 0.33; 95% CI, 0.14–0.77). Moreover, a higher ferritin level was significantly associated with higher HOMA–IR after adjusting for confounders in men. Lower transferrin saturation was also significantly associated with higher insulin levels and HOMA–IR in both sexes.
Results
Conclusions Lower transferrin saturations were associated with an increased risk of impaired fasting glucose and insulin resistance among general South Korean population.
Diabet. Med. 32, 673–678 (2015)
Introduction Iron is vital for humans because it is essential for many metabolic processes, including oxygen transport, DNA synthesis and electron transport [1]. However, iron can also damage tissues via the generation of oxygen free radicals that attack cellular membranes, proteins and DNA [2]. Thus, balancing the deleterious and beneficial effects of iron is an essential aspect of cell survival. Ferritin is responsible for the storage of iron and transferrin is responsible for the transport of iron from sites of absorption to sites of use [3]. It is well known that the serum
ferritin concentration closely parallels body iron reserves [3]. However, serum ferritin concentrations reflect not only body iron stores, but also inflammation [4], and transferrin saturation is used together with serum ferritin concentration to estimate the true iron status [5]. Previous studies have reported that increased serum ferritin concentrations are associated with hyperinsulinaemia, reduced insulin function and diabetes [6–11]. However, the associations between transferrin saturation and diabetes have not been well evaluated and conflicting results have been reported [5,9,12–17]. This study was performed to examine the association of iron indices (serum ferritin and transferrin saturation) with risk of impaired fasting glucose and insulin resistance and to
Correspondence to: Jai Dong Moon. E-mail: [email protected]
© 2014 The Authors. Diabetic Medicine © 2014 Diabetes UK
673
DIABETICMedicine
Transferrin saturation and insulin resistance in Korean adults R. J. Park and J. D. Moon
What’s new?
Data collection
• The associations of iron indices (serum ferritin and transferrin saturation) with impaired fasting glucose and insulin resistance were examined.
Anthropometric measurements of participants were conducted at local community health centres and clinics as a part of the health examination. The anthropometric variables (height and body weight) were recorded using a standard protocol. BMI was calculated as weight in kilograms divided by height in metres squared. Waist circumference was measured midway between the costal margin and iliac crest at the end of a normal expiration. In accordance with the Korean Society for the Study of Obesity criteria for waist circumference [19], abdominal obesity was defined as a waist circumference 90 cm in men and 85 cm in women. Information on the smoking and drinking status of the participants was obtained from a questionnaire survey. People who reported having smoked fewer than 100 cigarettes in their lives were classified as having never smoked, and the remainder were asked whether they were current or former smokers. Individuals who had 12 alcohol drinks or more per year were considered as drinkers. Level of physical activity was calculated using metabolic equivalent of task values (MET) based on self-reported frequency and duration of activities during the week. MET-minutes per week was computed by multiplying the MET value of a particular activity (walking = 3.3 METs, moderate physical activity = 4.0 METs and vigorous physical activity = 8 0 METs) with minutes spent in that particular activity [20]. Total weekly physical activity was calculated by summing MET-minutes per week of walking, moderate and vigorous activity. Participants who had 3000 MET-minutes per week were classified as vigorous physical activity. Blood pressure was measured with participants in the sitting position following a 5–minute rest period. Blood pressure was measured on three occasions with a mercury sphygmomanometer on the right arm and averaged for a final blood pressure reading. Mean arterial pressure (mm/Hg) was calculated as 2/3 diastolic blood pressure + 1/3 systolic blood pressure. Blood samples were obtained in the morning following an overnight fast. Levels of fasting plasma glucose, total cholesterol, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase and serum creatinine were measured (Hitachi Automatic Analyzer 7600, Hitachi, Tokyo, Japan). Haemoglobin and white blood cell count were measured using the sodium lauryl sulfate haemoglobin method (XE-2100D, Sysmex, Kobe, Japan) and laser flow cytometry (XE-2100D, Sysmex), respectively. Serum ferritin and insulin were measured by immunoradiometric assay (1470 Wizard gamma counter, PerkinElmer, Turku, Finland). Serum iron and iron-binding capacity were measured using the bathophenanthroline direct method (Hitachi Automatic Analyzer 7600, Hitachi). Transferrin saturation is
• Study participants included 2413 Korean individuals aged 20–50 years. • Lower transferrin saturations were associated with an increased risk of impaired fasting glucose and insulin resistance.
determine whether the association differed between men and women using a national representative sample of Korean adults who were free of comorbidities.
Participants and methods Study population
The study used data from the 2010 Korean National Health and Nutrition Examination Survey (KNHANES). A stratified multistage clustered probability design was used to select representative samples of non-institutionalized Korean civilians for the KNHANES [18]. The survey data were publicly available. This survey was completed by 8958 participants (81.9% of the total target population of 10,938). Our analysis was restricted to participants aged 20–50 years for whom fasting glucose (8 h fasting) levels were available (n = 2971). We did not study participants with diabetes (n = 99, defined as a plasma glucose level 7 mmol/l after fasting for a minimum of 8 h or physiciandiagnosed diabetes or self-reported current use of oral hypoglycaemic medication or insulin). Other exclusion criteria included malignancy (n = 16), liver cirrhosis (n = 2), chronic renal failure (n = 0), anaemia (n = 254, defined as haemoglobin < 13 g/dl for men and < 12 g/dl for women), pregnancy (n = 33) or menopause (n = 60, defined as self-reported cessation of menstruation for > 1 year or hysterectomy). Individuals with exceptionally increased serum ferritin levels (> 500 ng/ml; n = 3), increased serum liver enzyme values (aspartate aminotransferase 80 U/l and alanine aminotransferase 70 U/l; n = 7), increased serum creatinine level (> 132 lmol/l; n = 0), increased white blood cell counts (> 12.0 9 109/l; n = 14) or missing values for exposure variables (n = 70) were not studied. After these exclusions, 2413 individuals (1150 men and 1263 women) were included in our final analysis. The study was approved by the Korea Centers for Disease Control and Prevention Institutional Review Board. Written informed consent was obtained from all study participants.
674
© 2014 The Authors. Diabetic Medicine © 2014 Diabetes UK
Research article
the ratio of serum iron and total iron-binding capacity, multiplied by 100. All clinical analyses were performed by the Neodin Medical Institute, a laboratory certified by the Korean Ministry of Health and Welfare. Impaired fasting glucose was defined as a fasting plasma glucose of 5.5–6.9 mmol/l [21]. Insulin resistance was determined using the homeostasis model assessment of insulin resistance (HOMA–IR = fasting insulin [in lU/ml] 9 fasting glucose [mmol/l]/22.5) and fasting insulin levels [22].
Statistical analysis
We categorized the level of ferritin and transferrin saturation into quartiles (ferritin: < 69, 69–105, 106–157, 157 for men and < 17, 17–28, 29–47, 47 for women; transferrin saturation: < 30, 30–41, 42–53, 53 for men and < 24, 24–33, 34–43, 43 for women). The characteristics of the study sample were analysed using t-tests for continuous variables and chi-square tests for categorical variables. All analyses were performed separately for men and women. We obtained the odds ratio (OR) and 95% confidence interval (CI) for impaired fasting glucose using a multiple logistic regression analysis after adjusting for confounding variables across serum ferritin and transferrin saturation quartiles for both sexes. The covariates for the adjusted OR calculation were age, drinking status, smoking status, physical activity, abdominal obesity, BMI, mean arterial pressure, total cholesterol, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, white blood cell count and serum creatinine. We also conducted association analyses between iron indices and insulin resistance (fasting insulin, HOMA– IR) using analysis of covariance (ANCOVA) models in the same adjustment model. Two-tailed P-values of < 0.05 were considered statistically significant. Variables with skewed distributions were logtransformed before analysis. All statistical analyses were conducted using the SPSS statistical software package v. 18.0 (SPSS Inc., Chicago, IL, USA), which incorporates sample weights and adjusts the analyses for the complex sample design of the survey. Survey sample weights were calculated taking into consideration the sampling rate, response rate and age/sex proportion of the reference population (2010 National Census Registry).
Results Among these 2413 participants, 315 had impaired fasting glucose. Participant characteristics according to glucose regulation status are presented in Table 1. Those with impaired fasting glucose were older and more likely to be abdominally obese. In both sexes, they also had higher mean BMI, fasting glucose level, insulin concentration, HOMA– IR, mean arterial pressure, total cholesterol, aspartate aminotransferase and alanine aminotransferase. There was
© 2014 The Authors. Diabetic Medicine © 2014 Diabetes UK
DIABETICMedicine
a higher mean ferritin concentration in men and lower transferrin saturation in women. The prevalence of impaired fasting glucose increased significantly according to the increasing serum ferritin quartiles in men and decreased significantly according to the increasing transferrin saturation in women (Table 2). As shown in Table 3, impaired fasting glucose was more prevalent in the highest compared with the lowest serum ferritin quartile among men (OR, 1.97; 95% CI, 1.20–3.24) after adjustment for age, drinking status, smoking status, physical activity, abdominal obesity, BMI, mean arterial pressure, total cholesterol, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, white blood cell count and serum creatinine. Following the same adjustment, impaired fasting glucose was less prevalent in the highest compared with the lowest transferrin saturation quartile among men (OR, 0.45; 95% CI, 0.25–0.80) and women (OR, 0.33; 95% CI, 0.14–0.77). A higher ferritin level was significantly associated with higher HOMA–IR after adjusting for confounders in men (Table 4). Lower transferrin saturation was also significantly associated with higher insulin levels and HOMA–IR in both sexes.
Discussion To our knowledge, this is the first study examining the associations between transferrin saturation and impaired glucose metabolism among Korean population. In this large nationally representative cross-sectional study, we found that lower transferrin saturations were associated with impaired fasting glucose and insulin resistance. Our data corroborated previous observations. Cheung et al. [5] suggested that higher ferritin and lower transferrin saturation were associated with a higher risk of prediabetes in a general US population. A Chinese study [9] reported a similar magnetic resonance imaging-estimated hepatic iron concentration, but higher serum ferritin levels and lower transferrin saturation in the diabetes group than in the prediabetes group, which implied a decreased circulating iron in diabetes even though hepatic iron stores remained higher. We observed a trend toward increasing insulin resistance (HOMA–IR) with increasing serum ferritin levels in men (median ferritin: 105 ng/ml), but not in premenopausal women (median ferritin: 28 ng/ml). Similarly, Lee et al. [10] reported that an elevated serum ferritin level was significantly associated with insulin resistance in men (median ferritin: 96 ng/ml) and postmenopausal women (median ferritin: 54 ng/ml), but not in premenopausal women (median ferritin: 26 ng/ml). Kim et al. [11] found that HOMA–IR was associated with increased serum ferritin concentration in men (mean ferritin: 167 ng/ml), but not in women (mean ferritin: 65 ng/ml). These observations may be because serum ferritin concentrations of premenopausal women were significantly lower than men. The difference is
675
Transferrin saturation and insulin resistance in Korean adults R. J. Park and J. D. Moon
DIABETICMedicine
Table 1 General and clinical characteristics according to glucose regulation status Men [1,150/10,044,496]*
Variable
Normal fasting glucose [940/8,360,696]
Current alcohol intake (%) Current smoking (%) Vigorous physical activity (%) Abdominal obesity (%) Age (years) BMI (kg/m2) Fasting plasma glucose (mmol/l) Insulin (pmol/l) HOM–-IR Mean arterial pressure (mm/Hg) Total cholesterol (mmol/l) Aspartate aminotransferase (U/l) Alanine aminotransferase (U/l) Alkaline phosphatase (U/l) White blood cell count (9 109/l) Serum creatinine (lmol/l) Ferritin (ng/ml) Transferrin saturation (%)
81.0 53.4 32.4 19.5 33.33 23.85 4.92 69.76 2.21 88.08 4.74 22.33 25.28 225.82 6.51 84.43 114.07 42.95
1.6 1.7 1.9 1.6 0.40 0.12 0.01 1.24 0.04 0.46 0.03 0.47 0.55 2.05 0.05 0.41 2.86 0.77
Women [1,263/8,062,370]
Impaired fasting glucose [210/1,683,800] 85.9 50.5 27.6 35.8 40.31 25.64 5.91 88.02 3.36 94.44 5.28 26.72 33.18 227.59 6.63 83.03 146.19 41.22
3.0 4.7 3.7 3.5 0.58 0.33 0.03 5.36 0.23 1.03 0.09 1.06 1.66 4.15 0.11 0.89 7.88 1.25
P
< < < < < < < <
