Abstract
The present study investigated the relationship between hyperglycemia and both serum carotenoids and intake of vegetables and fruits. Subjects with a history of diabetes mellitus (DM group, n=133) or with hyperglycemia diagnosed using a 5.6% cutoff value for hemoglobin A1c (High HbA1C group, n=151) were recruited from among inhabitants of a rural area in Hokkaido, Japan. Intake frequencies of vegetables and fruits were assessed using a questionnaire administered by public health nurses. Serum levels of carotenoids and retinol were measured using high-performance liquid chromatography. The relationships between high HbA1C or DM and both serum carotenoids and intake frequencies of vegetables and fruits were analyzed using logistic regression modeling for a case-control study; each case (High HbA1C or DM) was matched to two controls (healthy subjects without any history of disease) matched for sex and age (within 3 years). The odds ratio (OR) for high HbA1C was 0.49 (95% confidence interval: 0.29-0.85) on high intake frequency of carrot and pumpkin and the OR for DM was 1.21 (95% CI: 0.79-1.84). No significant relationships were observed between high HbA1c and intake frequencies of other vegetables and fruits. The ORs on high serum levels of α- and β-carotenes, lycopene, β-cryptoxanthin and zeaxanthin and lutein were 0.38 (0.22-0.65), 0.35 (0.21-0.59), 0.57 (0.35-0.93), 0.35 (0.20-0.59), and 0.88 (0.54-1.46) for high HbA1c, respectively. In conclusion, intake of vegetables and fruits rich in carotenoids might be a protective factor against hyperglycemia.
Key words: intake of fruit and vegetables, carotenoids, beta-carotene, hyperglycemia, diabetes mellitus
Diabetes mellitus (DM) is characterized by chronic hyperglycemia due to insufficiency of insulin. Chronic hyperglycemia leads to complications such as retinopathy, nephropathy, and neuropathy.1 In addition, DM is a risk factor for coronary heart disease and cerebrovascular disease.2 In Japan, the prevalence of diabetes among those aged 40 years and older is estimated to be 6-7%3 and 6.3 million DM patients were identified in 1995.3 The number of DM patients in Japan is estimated to reach 8.5 million in 2025.4
Hyperglycemia increases the generation of free radicals by glucose autoxidation and increases levels of the glycoprotein Hemoglobin A1c (HbA1c).1,5-7 Hyperglycemia enhances oxidation stress such as increasing serum lipid peroxidation.1,7,8 The relationship between pathogenesis of DM and oxidation stress is unclear, but a close relation exists between the pathogenesis of DM complications and tissue injury from free radicals.9
Carotenoids such as β-carotene, contained in fruit and vegetables, protect cells from oxidative stress by quenching free radicals.9,10 Numerous epidemiologic studies have demonstrated that subjects with higher dietary consumption of fruit and vegetables rich in carotenoids, or with high serum level of β-carotene have lower risks of certain cancers and cardiovascular disease.9,11,12 A recent study13 found that serum levels of β-carotene and lycopene were lower among Americans with DM than those without. However, few reports have investigated serum carotenoids levels of subjects with chronic hyperglycemia, a pre-DM condition. HbA1c is a biomarker of long-term glucose homeostasis that reflects blood glucose concentrations over the previous 1 or 2 months.14 High HbA1c concentrations indicate a condition of chronic hyperglycemia.14
In the present study, we investigated the relationship between high HbA1c or DM and intake frequencies of fruit and vegetables and serum carotenoids levels in rural Japanese subjects.
SUBJECTS AND METHODS
Health examinations of inhabitants aged 40 years or older (range: 40 to 86 years) have been performed in an area of Hokkaido, Japan, every August since 1982.12 The population for this study comprised 1,691 subjects (656 males and 1,035 females) who attended the health examinations from 1998 through 2000. We used the first data as to repeaters. Of these, 151 subjects (81 males and 70 females) displayed HbA1c values of 5.6% or more and no history of DM (High HbA1c group). Other 133 subjects (73 males and 60 females) displayed a history of DM (DM group).
We randomly selected two controls for each subject in the High HbA1c group (Healthy 1 group) and DM group (Healthy 2 group) from health examination participants with no history of illness such as apoplexy, myocardial infarction, angina pectoris, hypertension, liver disease, renal disease, gout, DM, or cancer. Control subjects were pair-matched by sex and age (within 3 years).
Trained public health nurses administered a questionnaire regarding health and daily lifestyle habits, assessing smoking (current smoker, ex-smoker, or non-smoker), alcohol consumption (regular drinker, ex-drinker, or non-drinker), dietary intake of major foods (rarely, 1-2 times/month, 1-2 times/week, 3-4 times/week, or almost daily), physical exercise (rarely, 1-2 hours/week, 3-4 hours/week, or 5+ hours/week), and histories of major illnesses at the time of the health examination.
Fasting serum samples were taken during the health examinations and sera were separated from blood cells by centrifugation within 1 hour. Biochemical analysis of the sampled sera was performed using an auto-analyzer (JCA-RX20, Nihon Denshi Co. Ltd.). Whole blood HbA1c was measured by high-performance liquid chromatography.15
Serum levels of α- and β-carotene, lycopene, β-cryptoxanthin, zeaxanthin and lutein (zeaxanthin & lutein), canthaxanthin, retinol, and α-tocopherol were measured separately by high-performance liquid chromatography.16
Serum level of thiobarbituric acid-reactive substance (TBARS) was determined using the thiobarbituric acid reaction method.17
Body height, weight and blood pressure were measured during the health examinations. Body mass index (BMI) was calculated as body weight (kg) divided by height (m) squared.
We obtained informed consent with the participants’ signatures for providing information and serum to our epidemiologic study.
All statistical analyses, such as t tests, chi-square tests, and logistic regression analyses, were conducted using the statistical package (Stat View ver.5.0). To assess the relationship between intake frequencies of fruit and vegetables and the risks of high HbA1c and DM, intake frequency of fruit and vegetables were divided into two groups (low: 1-2 times/week or less, high: 3-4 times/week or more), and logistic regression analyses were performed after adjusting for smoking and alcohol consumption. To assess the relationship between serum levels of carotenoids and the risks of high HbA1c and DM, serum levels of carotenoids were divided into three groups (low, middle, and high) according to the distribution of serum carotenoids levels in Healthy 1 or 2 groups, and logistic regression analyses were performed after controlling for smoking habit and alcohol consumption, systolic blood pressure, BMI, and serum levels of triglyceride, total cholesterol, and γ-GTP activity. These potential confounding factors were strongly associated with serum carotenoids levels. Serum levels of carotenoids such as β-carotene are lower in smoker or alcohol drinker.18 Serum β-carotene levels were negatively associated with systolic blood pressure and BMI,18,19 and positively associated with serum levels of total cholesterol and triglyceride.19,20 Serum carotenoids levels were also affected by liver function, because absorbed carotenoids are stored in several organs, especially in the liver and adipose tissue.21 Serum level of γ-GTP activity, a biomarker of liver function, was added in adjusting factors.
RESULTS
Table 1 shows the characteristics of subjects in this study. The percentage of current alcohol drinkers was significantly higher in the High HbA1c group than in the Healthy 1 group. No significant differences in smoking, alcohol consumption, or exercise were observed between the DM and Healthy 2 groups.
Table 1. Characteristics of the subjects.
| Items | High HbA1c | Healthy 1 | p | DM | Healthy 2 | p | ||||||||||
| No .a | 151 | (100.0) | 302 | (100.0) | 133 | (100.0) | 266 | (100.0) | ||||||||
| male | 81 | (53.6) | 162 | (53.6) | n.s.$ | 73 | (54.9) | 146 | (54.9) | n.s.$ | ||||||
| female | 70 | (46.4) | 140 | (46.4) | 60 | (45.1) | 120 | (45.1) | ||||||||
| Age (year)b | 63.6 ± 9.9 | 63.1 ± 9.4 | n.s.# | 63.8 ± 9.7 | 63.7 ± 9.61 | n.s.# | ||||||||||
| Smoking habit a | ||||||||||||||||
| never smoker | 74 | (49.0) | 170 | (56.3) | n.s.$ | 63 | (47.4) | 142 | (53.4) | n.s.$ | ||||||
| ex-smoker | 35 | (23.2) | 51 | (16.9) | 34 | (25.6) | 54 | (20.3) | ||||||||
| current smoker | 42 | (27.8) | 81 | (26.8) | 36 | (27.1) | 70 | (26.3) | ||||||||
| Alcohol consumption a | ||||||||||||||||
| non-drinker | 79 | (52.3) | 206 | (68.2) | ***$ | 77 | (57.9) | 183 | (68.8) | n.s.$ | ||||||
| ex-drinker | 7 | (4.6) | 7 | (2.3) | 17 | (12.8) | 8 | (3.0) | ||||||||
| regular drinker | 64 | (42.4) | 87 | (28.8) | 39 | (29.3) | 74 | (27.8) | ||||||||
| Physical exercise a | ||||||||||||||||
| none | 100 | (66.2) | 207 | (68.5) | n.s.$ | 80 | (60.2) | 179 | (67.3) | n.s.$ | ||||||
| 1-2 hours/week | 27 | (17.9) | 43 | (14.2) | 20 | (15.0) | 35 | (13.2) | ||||||||
| 3-4 hours/week | 9 | (6.0) | 25 | (8.2) | 14 | (10.5) | 25 | (9.4) | ||||||||
| 5+ hours/week | 13 | (8.6) | 27 | (8.9) | 16 | (12.0) | 25 | (9.4) | ||||||||
High HbA1c: the subjects with no history of diabetes mellitus with HbA1c values of 5.6% or more.
DM: the subjects with history of diabetes mellitus
Healthy1: the healthy control group for High HbA1c group, Healthy2: the healthy control group for DM group.
a : The number and percentage in parenthesis.
b : mean ± standard deviation.
#: t test,$ : chi-square test, ***: p<0.001
Table 2 shows the levels of HbA1c, serum components, blood pressure, and BMI among the study groups. The levels of HbA1c, serum fasting glucose, triglyceride, TBARS, γ-GTP, systolic blood pressure, diastolic blood pressure and BMI were significantly higher in the High HbA1c group than in the Healthy 1 group. HDL-cholesterol was significantly lower in the High HbA1c group than in the Healthy 1 group. HbA1c, serum fasting glucose, triglyceride, γ-GTP, systolic blood pressure and BMI were significantly higher in the DM group than in the Healthy 2 group. HDL-cholesterol was significantly lower in the DM group than in the Healthy 2 group. HbA1c was significantly higher and serum total cholesterol, systolic blood pressure and diastolic blood pressure were significantly lower in the DM group than in the High HbA1c group.
Table 2. Comparison of serum components, HbA1c, blood pressure and BMI among the study groups.
| Items | High HbA1c | Healthy 1 | p | DM | Healthy 2 | p | |
| Fasting glucose a | (mg/dl) | 111.6 ± 34.4 | 94.2 ± 8.3 | *** | 120.1 ± 38.5 | 94.3 ± 8.4 | *** |
| HbA1c a | (%) | 6.0 ± 0.8 | 5.0 ± 0.3 | *** | 6.4 ± 1.2 | 5.0 ± 0.3 | *** |
| Total cholesterol a | (mg/dl) | 220.2 ± 38.1 | 215.5 ± 37.1 | 209.0 ± 31.6 | 210.4 ± 33.9 | ||
| HDL-cholesterol a | (mg/dl) | 55.0 ± 15.7 | 59.0 ± 16.3 | * | 55.6 ± 15.5 | 58.9 ± 17.5 | * |
| Triglyceride b | (mg/dl) | 123.3 | 97.5 | *** | 117.9 | 91.7 | *** |
| (66.6-264.2) | (54.0-186.3) | (64.6-234.4) | (52.0-173.9) | ||||
| TBARS b | (nmol/ml) | 3.17 | 2.91 | ** | 3.08 | 2.90 | |
| (2.40-3.82) | (2.30-3.70) | (2.45-3.80) | (2.20-3.60) | ||||
| γ-GTP b | (IU/l) | 28.6 | 18.7 | *** | 26.4 | 18.2 | *** |
| (12.0-85.4) | ( 9.0-51.0) | (11.0-72.2) | (10.0-46.9) | ||||
| SBP a | (mmHg) | 145.9 ± 20.7 | 133.8 ± 19.2 | *** | 141.4 ± 18.7 | 134.9 ± 20.4 | ** |
| DBP a | (mmHg) | 85.4 ± 10.7 | 81.4 ± 10.1 | *** | 81.0 ± 11.6 | 80.8 ± 9.7 | |
| BMI a | (kg/m2) | 24.8 ± 3.6 | 23.5 ± 2.9 | *** | 24.5 ± 3.3 | 23.2 ± 2.8 | *** |
High HbA1c: the subjects with no history of diabetes mellitus with HbA1c values of 5.6% or more.
DM: the subjects with history of diabetes mellitus.
Healthy1: the healthy control group for High HbA1c group, Healthy2: the healthy control group for DM group.
TBARS: tiobarbituric acid-reactive substances
SBP: systolic blood pressure, DBP: diastolic blood pressure, BMI: body mass index (Weight / [Hight] 2)
a : mean ± standard deviation.
b : geometric mean and range of 10-90% in parenthesis.
p: t test (*:p<0.05, **: p<0.01, ***: p<0.001)
Table 3 shows intake frequency of fruit and vegetables among the study groups. The percentage of subjects with high intake frequency of carrot and pumpkin was significantly lower, and that of tomato tended to be lower in the High HbA1c group than in the Healthy 1 group. The percentage of subjects with intake frequency of oranges over 3-4 times/week tended to be lower in the DM group than in the Healthy 2 group.
Table 3. Comparison of intake frequency of vegetables or fruits between the study groups.
| Intake frequency of vegetables or fruits | High HbA1c | Healthy 1 | p | DM | Healthy 2 | p | |||||
| Green leaf vegetables | Low | 78 | (51.7%) | 146 | (48.7%) | 65 | (48.9%) | 134 | (51.1%) | ||
| High | 73 | (48.3%) | 154 | (51.3%) | 68 | (51.1%) | 128 | (48.9%) | |||
| Carrot or Pumpkin | Low | 101 | (67.8%) | 176 | (58.3%) | * | 76 | (57.1%) | 164 | (61.7%) | |
| High | 48 | (32.2%) | 126 | (41.7%) | 57 | (42.9%) | 102 | (38.3%) | |||
| Tomato | Low | 82 | (55.4%) | 141 | (46.8%) | # | 72 | (54.1%) | 132 | (49.8%) | |
| High | 66 | (44.6%) | 160 | (53.2%) | 61 | (45.9%) | 133 | (50.2%) | |||
| Cabbage or Lettuce | Low | 69 | (45.7%) | 134 | (44.5%) | 58 | (43.6%) | 117 | (44.0%) | ||
| High | 82 | (54.3%) | 167 | (55.5%) | 75 | (56.4%) | 149 | (56.0%) | |||
| Chinese cabbage | Low | 116 | (77.3%) | 232 | (77.6%) | 105 | (80.2%) | 197 | (74.9%) | ||
| High | 34 | (22.7%) | 67 | (22.4%) | 26 | (19.8%) | 66 | (25.1%) | |||
| Seaweeds | Low | 67 | (45.0%) | 155 | (51.8%) | 68 | (52.3%) | 131 | (49.8%) | ||
| High | 82 | (55.0%) | 144 | (48.2%) | 62 | (47.7%) | 132 | (50.2%) | |||
| Oranges | Low | 117 | (80.1%) | 227 | (77.2%) | 113 | (86.9%) | 205 | (79.2%) | # | |
| High | 29 | (20.0%) | 67 | (22.8%) | 17 | (13.1%) | 54 | (20.8%) | |||
| Other Fruits | Low | 88 | (59.5%) | 161 | (54.4%) | 75 | (58.6%) | 160 | (61.1%) | ||
| High | 60 | (40.5%) | 135 | (45.6%) | 53 | (41.4%) | 102 | (38.9%) | |||
Low: 1-2 times/week or less, High: 3-4 times/week or more
High HbA1c: the subjects with no history of diabetes mellitus with HbA1c values of 5.6% or more.
DM: the subjects with history of diabetes mellitus.
Healthy 1: the healthy control group for High HbA1c group, Healthy 2: the healthy control group for DM group.
p : chi-square test (#: p<0.1, *: p<0.05)
Table 4 shows the odds ratios of intake frequency of fruit and vegetables for high HbA1c and DM. A significantly lower odds ratio for high HbA1c was observed in the high intake group of carrot and pumpkin compared with the low intake group (0.49, 95%CI: 0.29-0.85). Significantly lower odds ratio for DM was not observed between the high intake group of vegetables and fruits and the low intake group.
Table 4. Odds ratios (ORs) and 95% confidence intervals (CIs) of vegetables for high HbA1c and diabetes mellitus.
| Intake frequency of vegetables or fruits | OR for high HbA1c | OR for DM | |||
| Crude OR | Adjusted OR | Crude OR | Adjusted OR | ||
| Green leaf vegetables | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 0.92(0.50-1.72) | 0.96(0.50-1.83) | 1.10(0.72-1.66) | 1.08(0.71-1.66) | |
| Carrot or Pumpkin | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 0.49(0.29-0.85)* | 0.52(0.29-0.95)* | 1.21(0.79-1.84) | 1.19(0.76-1.85) | |
| Tomato | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 0.69(0.39-1.23) | 0.71(0.39-1.28) | 0.84(0.55-1.28) | 0.84(0.55-1.29) | |
| Cabbage or Lettuce | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 1.19(0.58-2.45) | 1.14(0.53-2.44) | 1.02(0.67-1.55) | 1.03(0.61-1.59) | |
| Chinese cabbage | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 0.92(0.55-1.55) | 0.94(0.55-1.60) | 0.74(0.44-1.23) | 0.71(0.44-1.21) | |
| Seaweeds | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 1.02(0.55-1.90) | 1.07(0.56-2.05) | 0.91(0.59-1.38) | 0.83(0.53-1.28) | |
| Oranges | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 0.75(0.44-1.27) | 0.87(0.50-1.52) | 0.57(0.32-1.03) | 0.59(0.32-1.10) | |
| Other Fruits | Low | 1.00 | 1.00 | 1.00 | 1.00 |
| High | 0.87(0.53-1.43) | 1.08(0.61-1.91) | 1.11(0.72-1.71) | 1.23(0.79-1.94) | |
Low: 1-2 times/week or less, High: 3-4 times/week or more
95% confidence intervals in parentheses.
Adjusted OR : Odds ratio adjusted for smoking habit and alcohol consumption.
*: p<0.05
Table 5 shows the levels of serum carotenoids, α-tocopherol and retinol among the study group. Serum levels of carotenoids, excluding canthaxanthin, were significantly lower in the High HbA1c group than in the Healthy 1 group. Serum β-cryptoxanthin levels were significantly lower in the DM group than in the Healthy 2 group, although serum levels of other carotenoids, α-tocopherol and retinol displayed no significant differences between the DM and Healthy 2 groups. Serum levels of carotenoids, excluding canthaxanthin and β-cryptoxanthin, were significantly lower in the High HbA1c group than in the DM group.
Table 5. Comparison of serum levels of carotenoids, α-tocopherol and retinol among the study groups.
| Items (µmol/l) | High HbA1c | Healthy 1 | p | DM | Healthy 2 | p |
| β-carotene | 0.87 | 1.24 | *** | 1.14 | 1.21 | |
| (0.28-2.39) | (0.45-3.29) | (0.41-3.46) | (0.47-3.09) | |||
| α-carotene | 0.10 | 0.13 | *** | 0.12 | 0.12 | |
| (0.04-0.22) | (0.05-0.31) | (0.05-0.32) | (0.05-0.28) | |||
| Lycopene | 0.38 | 0.46 | * | 0.50 | 0.45 | |
| (0.14-1.02) | (0.17-1.23) | (0.17-1.25) | (0.16-1.23) | |||
| β-cryptoxanthin | 0.20 | 0.26 | *** | 0.21 | 0.25 | * |
| (0.08-0.42) | (0.10-0.63) | (0.08-0.48) | (0.10-0.59) | |||
| Zeaxanthin & Lutein | 0.97 | 1.17 | *** | 1.10 | 1.11 | |
| (0.56-1.63) | (0.62-1.58) | (0.57-2.01) | (0.57-1.87) | |||
| Canthaxanthin | 0.02 | 0.02 | 0.02 | 0.02 | ||
| (0.010-0.037) | (0.010-0.044) | (0.010-0.040) | (0.010-0.043) | |||
| α-tocopherol | 29.22 | 27.60 | 28.92 | 26.92 | ||
| (19.83-46.08) | (20.24-39.21) | (20.31-42.32) | (19.28-38.14) | |||
| Retinol | 3.13 | 2.98 | 3.02 | 2.91 | ||
| (2.18-4.46) | (2.13-4.05) | (2.08-4.10) | (2.00-4.05) | |||
High HbA1c: the subjects with no history of diabetes mellitus with HbA1c values of 5.6% or more.
DM: the subjects with history of diabetes mellitus.
Healthy 1: the healthy control group for High HbA1c group, Healthy 2: the healthy control group for DM group.
Data represented as geometric mean and range of 10-90% in parentheses.
p: t test (log-transformed )
*: p<0.05, ***: p<0.001
The results from the analyses conducted to compare the intake frequencies of fruit and vegetables and the serum carotenoids levels between DM and Healthy 1 group were also similar to those using Healthy 2 group as controls.
Table 6 shows the odds ratios of serum carotenoids, α-tocopherol and retinol for high HbA1c. Significantly lower odds ratios for high HbA1c were observed in the high group for serum α- and β-carotene, lycopene, β-cryptoxanthin, and zeaxanthin & lutein than in the low group. However, the odds ratios of serum retinol and α-tocopherol were not significantly different between the high and low groups. Even after adjusting for confounding factors, these associations did not change.
Table 6. Odds ratios (ORs) and 95% confidence intervals (CIs) of serum levels of carotenoids, α-tocopherol and retinol for high HbA1c.
| Items | Range(µmol/l) | No. of High HbA1c / Healthy1 |
Crude odds ratio | Adjusted odds ratioa | |
| β-carotene | |||||
| Low | 0.08-0.96 | 81 / 102 | 1.00 | 1.00 | |
| Middle | 0.97-1.88 | 42 / 100 | 0.53(0.33-0.84) * | 0.52(0.30-0.90) * | |
| High | 1.89-8.66 | 28 / 100 | 0.35(0.21-0.59) ** | 0.36(0.22-0.80) ** | |
| α-carotene | |||||
| Low | 0.01-0.08 | 75 / 111 | 1.00 | 1.00 | |
| Middle | 0.09-0.19 | 52 / 97 | 0.79(0.51-1.24) | 0.94(0.56-1.57) | |
| High | 0.20-2.63 | 24 / 94 | 0.38(0.22-0.65) * | 0.51(0.27-0.96) * | |
| Lycopene | |||||
| Low | 0.01-0.33 | 66 / 102 | 1.00 | 1.00 | |
| Middle | 0.34-0.69 | 48 / 100 | 0.74(0.47-1.18) | 0.76(0.45-1.29) | |
| High | 0.70-3.19 | 37 / 100 | 0.57(0.35-0.93) * | 0.58(0.35-0.99) * | |
| β-cryptoxanthin | |||||
| Low | 0.03-0.18 | 72 / 101 | 1.00 | 1.00 | |
| Middle | 0.19-0.36 | 54 / 100 | 0.76(0.48-1.19) | 0.73(0.43-1.24) | |
| High | 0.37-1.91 | 25 / 101 | 0.35(0.20-0.59) ** | 0.41(0.22-0.79) ** | |
| Zeaxanthin & Lutein | |||||
| Low | 0.27-0.95 | 71 / 101 | 1.00 | 1.00 | |
| Middle | 0.96-1.40 | 53 / 103 | 0.73(0.47-1.15) | 0.66(0.39-1.09) | |
| High | 1.41-6.67 | 27 / 98 | 0.39(0.23-0.66) * | 0.37(0.20-0.60) ** | |
| Canthaxanthin | |||||
| Low | 0.010-0.019 | 52 / 84 | 1.00 | 1.00 | |
| Middle | 0.020-0.030 | 72 / 134 | 0.87(0.55-1.36) | 1.05(0.63-1.74) | |
| High | 0.031-0.090 | 27 / 84 | 0.52(0.30-0.90) | 0.71(0.38-1.33) | |
| α-tocopherol | |||||
| Low | 10.84-24.43 | 44 / 101 | 1.00 | 1.00 | |
| Middle | 24.44-30.61 | 45 / 102 | 1.01(0.62-1.67) | 0.67(0.37-1.20) | |
| High | 30.62-65.43 | 62 / 99 | 1.44(0.89-2.31) | 0.83(0.43-1.59) | |
| Retinol | |||||
| Low | 1.19-2.73 | 50 / 102 | 1.00 | 1.00 | |
| Middle | 2.74-3.35 | 35 / 103 | 0.69(0.42-1.16) | 0.62(0.36-1.01) | |
| High | 3.36-7.01 | 66 / 97 | 1.39(0.88-2.20) | 0.75(0.43-1.31) | |
95% confidence intervals in parenthesis.
a: Odds ratio adjusted for smoking habit, alcohol consumption, systolic blood pressure, BMI and serum levels of total cholesterol, triglyceride and γ-GTP activity.
*: p<0.05, **p<0.01
A significantly lower odds ratio for DM was observed in the high group of β-cryptoxanthin than in the low group. No significant differences in serum other carotenoids, α-tocopherol or retinol were observed between the high and low groups (Table 7).
Table 7. Odds ratios (ORs) and 95% confidence intervals (CIs) of serum levels of carotenoids, α-tocopherol and retinol for diabetes mellitus.
| Items | Range(µmol/l) | No. of DM / Healthy2 |
Crude odds ratio | Adjuste odds ratioa | |
| β-carotene | |||||
| Low | 0.08-0.88 | 50 / 89 | 1.00 | 1.00 | |
| Middle | 0.89-1.74 | 45 / 91 | 0.87(0.53-1.43) | 0.95(0.53-1.69) | |
| High | 1.78-9.18 | 38 / 86 | 0.78(0.46-1.32) | 1.16(0.60-2.25) | |
| α-carotene | |||||
| Low | 0.01-0.09 | 45 / 87 | 1.00 | 1.00 | |
| Middle | 0.10-0.17 | 42 / 93 | 0.87(0.52-1.46) | 1.08(0.61-1.93) | |
| High | 0.18-1.20 | 46 / 86 | 1.03(0.62-1.72) | 1.28(0.70-2.34) | |
| Lycopene | |||||
| Low | 0.01-0.36 | 43 / 92 | 1.00 | 1.00 | |
| Middle | 0.37-0.70 | 43 / 87 | 1.01(0.61-1.69) | 1.15(0.65-2.03) | |
| High | 0.71-2.61 | 47 / 84 | 1.20(0.72-1.98) | 1.30(0.74-2.31) | |
| β-cryptoxanthin | |||||
| Low | 0.02-0.19 | 51 / 89 | 1.00 | 1.00 | |
| Middle | 0.20-0.34 | 52 / 85 | 1.09(0.61-1.69) | 1.07(0.61-1.90) | |
| High | 0.35-1.91 | 30 / 92 | 0.57(0.34-0.98)* | 0.68(0.36-1.30) | |
| Zeaxanthin & Lutein | |||||
| Low | 0.27-0.86 | 49 / 85 | 1.00 | 1.00 | |
| Middle | 0.87-1.26 | 36 / 89 | 0.69(0.41-1.16) | 0.77(0.44-1.37) | |
| High | 1.27-6.67 | 48 / 92 | 0.88(0.54-1.46) | 0.97(0.55-1.70) | |
| Canthaxanthin | |||||
| Low | 0.010-0.019 | 46 / 74 | 1.00 | 1.00 | |
| Middle | 0.020-0.029 | 42 / 98 | 0.66(0.39-1.11) | 0.65(0.37-1.16) | |
| High | 0.030-0.105 | 45 / 94 | 0.74(0.44-1.24) | 0.76(0.42-1.35) | |
| α-tocopherol | |||||
| Low | 10.84-24.82 | 33 / 93 | 1.00 | 1.00 | |
| Middle | 24.83-31.33 | 49 / 90 | 1.49(0.90-2.48) | 1.70(0.92-2.99) | |
| High | 31.34-94.11 | 51 / 81 | 1.75(1.04-2.95) | 1.59(0.79-2.87) | |
| Retinol | |||||
| Low | 0.87-2.58 | 41 / 89 | 1.00 | 1.00 | |
| Middle | 2.59-3.32 | 38 / 90 | 0.91(0.54-1.53) | 0.68(0.38-1.23) | |
| High | 3.33-6.53 | 54 / 87 | 1.37(0.82-2.29) | 0.95(0.52-1.72) | |
95% confidence intervals in parenthesis.
a : Odds ratio adjusted for smoking habit, alcohol consumption, systolic blood pressure, BMI and serum levels of total cholesterol, triglyceride and γ-GTP activity.
*: p<0.05
In addition, the results obtained using the logistic regression between DM and Healthy1 group were also similar to those using Healthy 2 group as controls.
DISCUSSION
Few studies have reported the relationships between serum carotenoids and DM or chronic hyperglycemia. The NHANES III study revealed that serum β-carotene levels in subjects with impaired glucose tolerance and newly diagnosed with diabetes were lower compared to individuals with normal glucose tolerance.13 In addition, serum lycopene levels demonstrated an inverse relationship to glucose intolerance.13 In a nested case-control study in Finland,22 high levels of β-carotene and α-tocopherol were associated with decreased risk of non-insulin dependent DM.
In this study, the High HbA1c group demonstrated significantly lower intakes of carrot and pumpkin and serum levels of serum carotenoids, such as α- and β-carotene, lycopene, β-cryptoxanthin, and zeaxanthin & lutein compared to the Healthy 1 group. Furthermore, significantly low odds ratios for high HbA1c were observed in the high groups of α- and β-carotene, lycopene, β-cryptoxanthin, and zeaxanthin & lutein.
Chronic hyperglycemia leads to auto-oxidation of glucose and causes nonenzymatic glycation of proteins through Maillard’s reaction.1,5 In these processes, reactive oxygen species are produced.1 The reactive oxygen species are well known as important risk factors for cardiovascular disease9,23 and cancer.9,24 The reactive oxygen species cause pancreatic β-cell damage and lead to insulin pathocrinia.5 In an animal experimental study,25 hyperglycemia caused oxidative damage to the pancreatic β-cells of GK rats, a model of Type 2 diabetes.
Antioxidant enzymes such as superoxide dismutase and glutathione peroxidase play protective roles against the reactive oxygen species.1 Chronic hyperglycemia reportedly causes glycation of those enzymes and reduces these functions,26,27 a factor for increased oxidative stress in vivo.
Carotenoids such as β-carotene are known to protect cells from oxidative stress by quenching free radicals.9,10 In addition, there was significantly negative correlation between serum β-carotene and HbA1c levels in this study subjects (r=0.119, p<0.05). Carotenoids are particularly abundant in green-yellow vegetables and fruits.28 Our previous report29 revealed that serum levels of carotenoids such as α- and β-carotene were correlated with intake frequencies of green-yellow vegetables and fruits. Low serum levels of carotenoids such as α- and β-carotene may reflect to low intake of fruit and vegetables in carotenoids. Spearman’s rank correlation coefficient between intake frequency of carrot or pumpkin, however, and serum β-carotene was 0.20-0.25 and those between intake frequency of other vegetables or fruits and serum other carotenoids levels were approximately 0.20.29 There was insufficient validity of questionnaire of dietary food intake in this study, therefore, and clearly results might not be obtained.
We suggested that individuals with high HbA1c values display lower serum carotenoids such as α- and β-carotene, due to both low intake frequencies of fruit and vegetables and increased production of the reactive oxygen species by chronic hyperglycemia.1
Another report showed that serum levels of vitamin C and E, which play as antioxidants, were lower in people with DM and impaired glucose tolerance.27 Furthermore, oral administration of large doses of α-tocopherol to subjects with diabetes and of ascorbic acid to non-diabetic subjects decreased glycosylated hemoglobin concentrations.7,30
Our results that the difference of serum β-carotene levels between in High HbA1c group and in Healthy 1 group were approximately 30%. In the NHANES III study,13 the difference of serum β-carotene levels in subjects with impaired glucose tolerance or in subjects newly diagnosed with diabetes were 13% and 20% lower, respectively, compared to subjects with normal glucose tolerance. This difference might be affected by the difference of case and control subjects. Our case subjects were high HbA1c group that was HbA1c values of 5.6% or more, but subjects in the NHANES III study were classified by an oral glucose tolerance test based on the World Health Organization criteria. Our control group was healthy subjects with no history of illness, but that of NHANES III study was subjects with normal glucose tolerance including those with illness, except DM. Moreover, serum β-carotene levels in our study subjects were higher than in the NHANES III study subjects (geometric mean of the subjects with normal glucose tolerance: 0.425 µmol/l).
In the DM group, intake frequency of oranges rich in β-cryptoxanthin and serum levels of β-cryptoxanthin were lower compared with the Healthy 2 group. Serum levels of other carotenoids such as β-carotene were not significantly different between DM and Healthy 2 groups.
No differences in serum levels of total cholesterol and TBARS were observed between DM and Healthy 2 groups. The percentage of ex-smokers was higher in the DM group than the Healthy 2 group. It is possible that these results reflect the effect of medical treatment, including health education, in medical institutions. Serum levels of lipid peroxidation in DM patients with good control are not elevated.26 We suggest that the DM group in this study included many patients with good control.
There were two limitations in this study subjects. First, the subjects in the High HbA1c group were individuals with newly untreated DM or abnormal carbohydrate tolerance. It is impossible to exclude newly untreated DM from High HbA1c group, because we cannot diagnose diabetes mellitus from only results of health examination. According the Japan Diabetes Society criteria,31 the diagnosis of diabetes can be made by fasting blood glucose levels of 126mg/dl or higher and HbA1c values of 6.5% or higher. There were 12 subjects (7.9%) who fulfilled those criteria (newly DM group) in High HbA1c group of this study. The same results were obtained when those subjects were excluded in analysis. Newly DM group had significantly lower levels of serum β-carotene than Healthy 1 group. Serum other carotenoids levels in newly DM group were lower than Healthy 1 group, but not significant (data not shown).
Second, many subjects had attended these health examinations two or more times in the present study. We therefore suggest that many subjects paid close attention to their health. Subjects in the DM group complied with health guidance and many good control patients were included.
Results from this study suggest that intake of fruit and vegetables rich in carotenoids might be a protective factor against hyperglycemia.
ACKNOWLEDGMENTS
We are grateful to the participants and staff of the health examination program for residents of Y-town, Hokkaido, Japan, and to Dr. Rikio Shinohara, Professor of Biochemistry at Fujita Health University, for measurement of serum thiobarbituric acid-reactive substance.
REFERENCES
- 1.Giugliano D, Ceriello A, Paolisso G. Oxidative stress and diabetic vascular complications. Diabetes Care 1996;19:257-67. [DOI] [PubMed] [Google Scholar]
- 2.Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes, other risk factors and 12-yr cardiovascular mortality for men screened in the Multiple Risk Factor Intervention Trial. Diabetes Care 1993; 16: 434-44. [DOI] [PubMed] [Google Scholar]
- 3.Terao A, Konishi M, Baba S, Mannami T. Prevalence of glucose intolerance, diabetes mellitus, and high serum insulin levels in a Japanese urban population. Jpn J Public Health 1997;44:283-91. (in Japanese) [PubMed] [Google Scholar]
- 4.King H, Aubert RE, Herman WH. Global burden of diabetes, 1995-2025: prevalence, numerical estimates, and projections. Diabetes Care 1998;21:1414-31. [DOI] [PubMed] [Google Scholar]
- 5.Sakurai T, Tsuchiya S. Superoxide production from nonenzymatically glycated protein. FEBS Lett 1988;236:406-10. [DOI] [PubMed] [Google Scholar]
- 6.Gilley P, Monboisse JC, Maquart FX, Borel JP. Glycation of proteins as a source of superoxide. Diabete Metab 1988;14:25-30. [PubMed] [Google Scholar]
- 7.Hunt JV, Dean RT, Wolff SP. Hydroxyl radical production and autoxidative glycosylation. Glucose autoxidation as the cause of protein damage in the experimental glycation model of diabetes mellitus and ageing. Biochem J 1988;256:205-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ceriello A, Giugliano D, Quatraro A, Donzella C, dipalo G, Lefebvre PJ. Vitamin E reduction of protein glycosylation in diabetes. New prospect for prevention of diabetic complications? Diabetes Care 1991;14:68-72. [DOI] [PubMed] [Google Scholar]
- 9.Machlin LJ, Bendich A. Free radical tissue damage: protective role of antioxidant nutrients. FASEB J 1987;1:441-5. [PubMed] [Google Scholar]
- 10.Burton GW. Antioxidant action of carotenoids. J Nutr 1989;119:109-11. [DOI] [PubMed] [Google Scholar]
- 11.Comstock GW, Helzlsouer KJ, Bush TL. Prediagnostic serum levels of carotenoids and vitamin E as related to subsequent cancer in Washington Country, Maryland. Am J Clin Nutr 1991; 53: 260S-264S. [DOI] [PubMed] [Google Scholar]
- 12.Ito Y, Suzuki S, Yagyu K, Sasaki R, Suzuki K, Aoki K. Relationship between serum carotenoid levels and cancer death rates in the residents, living in a rural area of Hokkaido, Japan. J Epidemiol 1997;7:1-8. [DOI] [PubMed] [Google Scholar]
- 13.Ford ES, Will JC, Bowman BA, Narayan KMV. Diabetes mellitus and serum carotenoids: finding from the Third National Health and Nutrition Examination Survey. Am J Epidemiol 1999;149:168-76. [DOI] [PubMed] [Google Scholar]
- 14.The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N Engl J Med 1993;329:977-86. [DOI] [PubMed] [Google Scholar]
- 15.Klenk DC, Hermanson GT, Krohn RI, Fujimoto EK, Mallia AK, Smith PK, et al. Determination of glycosylated hemoglobin by affinity chromatography: comparison with colorimetric and ion-exchange methods, and effects of common interferences. Clin Chem 1982;28:2088-94. [PubMed] [Google Scholar]
- 16.Ito Y, Ochiai J, Sasaki R, Sasaki S, Kusuhara Y, Morimitsu Y, et al. Serum concentrations of carotenoids, retinol, and α-tocopherol in healthy persons determined by high performance liquid chromatography. Clin Chem Acta 1990;194:131-44. [DOI] [PubMed] [Google Scholar]
- 17.Yagi K. A simple fluorometric assay for lipoperoxide in blood plasma. Biochem Med 1976;15:212-6. [DOI] [PubMed] [Google Scholar]
- 18.Ito Y, Shinohara R, Sasaki R, Yagyu S, Mizutani N, Okamoto K, et al. Relationship between serum levels of carotenoids, retinol or α-tocopherol and serum lipid peroxide levels in the inhabitants living in a town of southern Hokkaido. Vitamins 1994;68:569-78. (in Japanese) [Google Scholar]
- 19.Wallstom P, Wirfalt E, Lahmann PH, Gullberg B, Janzon L, Berglund G. Serum concentrations of β-carotene and α-tocopherol are associated with diet, smoking, and general and central adiposity. Am J Clin Nutr 2001;73:777-85. [DOI] [PubMed] [Google Scholar]
- 20.Stryker WS, Kaplan LA, Stein EA, Atampfer MJ, Sober A, Willett WC. The relation of diet, cigarette smoking, and alcohol consumption to plasma beta-carotene and alpha-tocopherol levels. Am J Epidemiol 1988;127:283-96. [DOI] [PubMed] [Google Scholar]
- 21.Goodman DS. Overview of current knowledge of metabolism of vitamin A and carotenoids. J Natl Cancer Inst 1984;73:1375-9. [PubMed] [Google Scholar]
- 22.Reunanen A, Knekt P, Aaran RK, Aromaa A. Serum antioxidants and risk of non-insulin dependent diabetes mellitus. Eur J Clin Nutr 1998;52:89-93. [DOI] [PubMed] [Google Scholar]
- 23.Duthie GG, Wahle KWJ, James WPT. Oxidants, antioxidants and cardiovascular disease. Nutr Res Rev 1989;2:51-62. [DOI] [PubMed] [Google Scholar]
- 24.Certti PA. Prooxidant status and tumor promotion. Science 1985;227:375-81. [DOI] [PubMed] [Google Scholar]
- 25.Ihara Y, Toyokuni S, Uchida K, Odaka H, Tanaka T, Ikeda H, et al. Hyperglycemia causes oxidative stress in pancreatic β-cells of GK rats, a model of type 2 diabetes. Diabetes 1999;48:927-32. [DOI] [PubMed] [Google Scholar]
- 26.Altomare E, Vendemiale G, Chicco D, Procacci V, Cirelli F. Increased lipid peroxidation in type 2 poorly controlled diabetic patients. Diabete Metab 1992;18:264-71. [PubMed] [Google Scholar]
- 27.Vijayalingam S, Parthiban A, Shanmugasundaram KR, Mohan V. Abnormal antioxidant status in impaired glucose tolerance and non-insulin-dependent diabetes mellitus. Diabet Med 1996;13:715-9. [DOI] [PubMed] [Google Scholar]
- 28.Beecher GR, Khachik F. Qualitative relationship of dietary and plasma carotenoids in human beings. Ann N Y Acad Sci 1992;669:320-1. [DOI] [PubMed] [Google Scholar]
- 29.Suzuki K, Ito Y, Otani M, Suzuki S, Aoki K. A study on serum carotenoid levels of people with hyperglycemia who were screened among residents living in a rural area of Hokkaido, Japan. Jpn J Hyg 2000;55:481-8. (in Japanese) [DOI] [PubMed] [Google Scholar]
- 30.Davie SJ, Gould BJ, Yudkin JS. Effect of vitamin C on glycosylation of proteins. Diabetes 1992;41:167-73. [DOI] [PubMed] [Google Scholar]
- 31.Kuzuya T, Nakagawa S, Satoh J, Kanazawa Y, Iwamoto Y, Kobayashi M, et al. Report of the Committee of Japan Diabetes Society on the Classification and Diagnostic Criteria of Diabetes Mellitus The Committee of Japan Diabetes Society for the Diagnostic Criteria of Diabetes Mellitus. J Jpn Diabet Soc 1999;42:385-404. (in Japanese) [Google Scholar]