Association between an Antioxidant-Rich Japanese Diet and Chronic Kidney Disease: The Ohasama Study

Abstract

Aims: Although physiological effects of hydrophilic- (H-) and lipophilic- (L-) antioxidant capacities (AOCs) are suggested to differ, the association of an antioxidant-rich diet and chronic kidney disease (CKD) incidence has not been examined. We therefore explored the association between the H- or L-AOC of a whole Japanese diet and CKD risk in a general population.

Methods: A total of 922 individuals without CKD (69.2% women; mean age, 59.5 years old) from Ohasama Town, Japan, were examined. CKD incidence was defined as the presence of proteinuria and/or an estimated glomerular filtration rate (eGFR) of ＜60 ml/min/1.73 m². Consumption of H-/L-AOC was determined based on the oxygen radical absorbance capacity in a specially developed Japanese food AOC database. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated for new-onset CKD using a Cox proportional hazards model.

Results: During the median follow-up of 9.7 years, 137 CKD incidents were recorded. After adjusting for potential confounding variables, the highest quartile of L-AOC was significantly associated with a 51% reduced CKD risk among only women. An increased L-AOC intake was more effective in preventing eGFR reduction than in preventing proteinuria in women. These associations were not seen for H-AOC intake in both sexes and L-AOC intake in men.

Conclusions: A high intake of lipophilic antioxidants may be associated with a reduced CKD risk. The balance between dietary antioxidant intake and pro-oxidants induced by unhealthy lifestyles may be crucial for preventing future kidney deterioration.

Introduction

Chronic kidney disease (CKD) has been recognized as a major public health issue¹⁾. The renal function declines with age, and CKD is an independent risk factor of cardiovascular disease (CVD)^2-4), stroke⁴⁾, end-stage renal disease⁵⁾, and all-cause mortality²⁾. Therefore, identifying the risk and preventive factors of CKD is important.

Increasing evidence from both experimental and epidemiological studies suggests that oxidative stress plays a major role in CKD pathogenesis^{6, 7)}. Dietary compounds with antioxidant activity may have cumulative or synergistic antioxidant effects. However, data on assessing the association between the intake of antioxidant-rich foods and CKD risk are limited⁶⁾. Few studies have prospectively examined the antioxidant capacity (AOC) and preventive effects on CKD incidence by targeting community-based healthy people.

Previous Japanese studies have referenced the AOC values of foods using databases developed primarily in other countries, not in Japan^{8, 9)}, although geographic location and environmental conditions, such as climate, season, and fertility of soil in which the collected food was grown, are known to affect the AOC of foods¹⁰⁾. Furthermore, foods expected to have a high AOC contribution in the Japanese diet, such as rice and seafood, were barely considered in those studies^{8, 9)}, as those foods were considered to contain very low AOCs or were not regularly consumed in Western countries¹¹⁾. Moreover, the preventive effects on diseases of free radical-scavenging activities in foods are thought to differ between water- and lipid-soluble antioxidants due to different pharmacokinetics in the body^{12, 13)}. However, previous epidemiological studies used a single indicator mixed or solely hydrophilic (H-) and lipophilic (L-) AOC^{6, 8, 9)}. We recently developed a reliable database of H- and L-AOCs reflecting the characteristics of food intake in the Japanese diet¹⁴⁾. Since the Japanese diet is widely known to reduce not only overall mortality but also chronic diseases, such as cardiovascular disease, cancer, and stroke^15-18), it is meaningful to examine the potential effects of the dietary AOC intake in Japanese individuals on CKD incidence.

This longitudinal study was conducted to investigate the association between hydrophilic or lipophilic AOCs in a typical Japanese diet and CKD incidence in the general Japanese population.

Materials and Methods

Study Design

This study was part of the longitudinal community-based observational study in Ohasama, Iwate Prefecture, Japan. The socioeconomic and demographic characteristics in this region and the full details of the project have been described previously^{19, 20)}. This noninvasive observational study complies with the Declaration of Helsinki and was approved by the Institutional Review Board of Teikyo University School of Medicine and by the Department of Health of the Hanamaki (Ohasama) Municipal Government (Approval number: 16-075-7). All participants provided their informed consent. The present research is reported based on the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.

Participants

Fig.1 shows the flow diagram of participant selection. In Japan, annual health checkups for farmers, self-employed people, pensioners, and dependents ≥ 35 years old were conducted. Among the Ohasama residents, 2,719 were eligible for annual health checkups in 1997 (study baseline). Among the 1,831 individuals who participated in checkups that year, 1,629 provided their informed consent and participated in this study. Overall, 433 participants were excluded from follow-up for the following reasons: insufficient urine dipstick and serum creatinine data (n=93), incomplete answers on food intake (n=41), and extreme levels of energy intake (2.5% above or below the range for all participants; n=122). Furthermore, 108 participants with CKD at baseline and 69 who were not fully independent in their basal activities of daily living (ADL) were excluded in order to examine new-onset CKD incidence. Those who did not undergo follow-up health checkups (n=274) were also excluded. Ultimately, 922 individuals (69.2% women; mean age 59.5 years old) were included in this study.

Fig.1.

Flow diagram of study participants in the present analysis of the Ohasama Study, Japan

Compared with participants in the follow-up study, non-participants were more likely to be older (participants vs. non-participants: 59.5 years vs. 64.3 years) and current smokers (23.1% vs. 30.7%) and had a higher hypertension prevalence (36.1% vs. 52.6%) and lower estimated glomerular filtration rate (eGFR; 81.9 vs. 79.4 ml/min/1.73 m²). Other lifestyle factors did not differ markedly between the participants and non-participants (Supplementary Table 1).

Supplementary Table 1.Characteristics of the study participants vs. nonparticipants

	Participants	Nonparticipants
Number of participants	917	274
Men, %	30.8	36.5
Age in years, mean±SD	59.5±10.1	64.3±12.2
BMI in kg/m2, mean±SD	23.6±3.1	23.5±3.5
≥ 25 kg/m2, %	31.3	31.4
Hypertension, %	36.1	52.6
Diabetes, %	7.2	9.1
Hypercholesterolemia, %	26.5	24.8
History of cardiovascular disease, %	7.9	10.2
Current smokers, %	23.1	30.7
Current drinkers, %	39.6	38.3
Total energy intake in kcal/day, mean±SD	1728±541	1670±518
Serum creatinine in mg/dl, mean±SD	0.63±0.13	0.64±0.13
eGFR in ml/min/1.73 m2, mean±SDa	81.9±8.4	79.4±9.2

Abbreviations: BMI, Body Mass Index; eGFR, estimated Glomerular Filtration Rate; SD, Standard Deviation.

^a eGFR was estimated from the serum creatinine (SCr) value using eGFR_CKD-EPI (mL/min/1.73 m²) = 141×min (SCr/κ, 1)^α×max (SCr/κ, 1)^－1.209× 0.993^Age×1.018 (if women)×0.813 (Japanese coefficient). κ: 0.7 in women and 0.9 in men, α; -0.329 in women and -0.411 in men, min indicates the minimum of SCr/κ or 1, and max indicates the maximum of SCr/κ or 1.

Dietary Data

A standardized methodology was used to calculate nutrient and food consumption from the data obtained using the Japanese version of a 141-item food-frequency questionnaire (FFQ) conducted between February 1 and March 28, 1998. The reproducibility and validity of this questionnaire has been previously reported in detail²¹⁾. The FFQ asked questions regarding the average frequency of consuming each food during the previous year, with the nine frequency categories ranging from “no consumption” to “seven or more times per day”. The standard portion size of a single serving was specified for each food, and respondents were asked if their usual portion was larger than (＞1.5 times), similar to, or smaller than (＜0.5) the standard. In this study, alcoholic beverages were not considered. Nutritional supplements were also not considered, as only a few responders used them.

The mean daily food and nutritional intakes were evaluated using a food composition table derived from the 2010 Japanese Standard Food Composition Table²²⁾. Since the table did not contain the antioxidant capacity value of foods, daily average intakes of hydrophilic (H-)/lipophilic (L-) oxygen radical absorbance capacity (ORAC) values were calculated using an AOC database containing 189 food items specialized to Japanese dietary characteristics^{14, 23, 24)}. This database was established based on the 12-day dietary records from 113 Japanese participants (55 men; 58 women) and covered 78.8% of their total food intake on a food-weight base. All AOCs were actual measured values of foods obtained in Japan using validated protocols^25-27). Food types that contributed the most to each antioxidant were tea, soy products, coffee, and rice for H-AOC and soy products, fish and shellfish, vegetables, and seaweed for L-AOC, which covered ≥ 50% of these AOC values in the diet; of note, the latter was more reflective of traditional Japanese foods than the former.

All food and nutrient intakes were adjusted for the total energy intake using the residual method, and separate regression models were established to obtain residuals in men and women. Following this procedure, participants were divided into quartiles according to their H- and L-AOC intakes, with the lowest quartiles used as reference categories.

Measurement of the eGFR and Proteinuria

The Jaffé method was used to measure the serum creatinine level during annual checkups before 2002, with the enzymatic method used thereafter. The renal function was estimated by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation modified for Japanese using the Japanese coefficient based on inulin clearance, as follows: eGFR_CKD-EPI (ml/min/1.73 m²)=141×min (serum creatinine (SCr)/κ, 1)^α×max (SCr/κ, 1)^−1.209×0.993^Age×1.018 (if women)×0.813 (Japanese coefficient). κ: 0.7 in women and 0.9 in men, α: −0.329 in women and −0.411 in men, min indicates the minimum of SCr/κ or 1, and max indicates the maximum of SCr/κ or 1 ^{28, 29)}. The following equation was used to convert the serum creatinine level from the Jaffé method to the enzymatic method: (serum creatinine [enzymatic]=serum creatinine [Jaffé] −0.2)³⁰⁾. Proteinuria was diagnosed using a dipstick test for spot urine (Urohemabonbix 5G08C; Bayer Medical, Tokyo, Japan). Proteinuria was considered for a dipstick result of ≥ 1+, corresponding to a urinary protein level of ＞30 mg/dl. CKD was defined as an eGFR of ＜60 ml/min/1.73 m² and/or positive proteinuria³¹⁾.

Other Variables

Smoking and drinking status and medication use for hypertension, hypercholesterolemia, diabetes mellitus, and the CVD history were verified by reviewing the medical records and using a questionnaire. Blood samples were collected while participants were sitting after resting for approximately 30 min, either between 09:00 and 11:00 or between 13:00 and 15:00; the majority of the participants had not fasted. Serum creatinine, total cholesterol, glucose, and glycosylated hemoglobin (HbA1c) levels were measured. Diabetes mellitus was defined as random blood glucose levels of ≥ 11.1 mmol/l (≥ 200 mg/dl), non-fasting blood glucose level of ≥ 11.1 mmol/l (≥ 200 mg/dl), fasting blood glucose level of ≥ 7.0 mmol/l (≥ 126 mg/dl), and HbA1c level of ≥ 6.5% according to the Japan Diabetes Society, or treatment with insulin, oral hypoglycemic agents, and/or a history of diabetes mellitus. Hypercholesterolemia was defined as total cholesterol levels of ≥ 5.7 mmol/l (≥ 220 mg/dl), medication use, and/or a history of hypercholesterolemia. Nurses at local medical centers measured blood pressure twice using a semi-automatic device (USM-700F; UEDA Electronic Works, Tokyo, Japan) based on the Korotkoff sound technique, while participants were sitting after a rest interval of at least 2 min. The two readings were averaged and used for the analysis. Hypertension was defined as the use of antihypertensive medication and/or blood pressure of ≥ 140/90 mmHg.

Follow-Up and Outcomes

The primary outcome was new-onset CKD diagnosed during annual checkups between 2002 and 2010. If the outcome occurred more than once during follow-up, only the first outcome was used for the analysis. The incidence date was defined as the midpoint between the examination date before the CKD onset and the initial onset of CKD symptoms. The observation period was from baseline to the incidence date in participants with CKD and to the date of final checkup in non-CKD participants.

Statistical Analyses

The multivariable-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for the associations for CKD incidences across each H- or L-AOC quartile were calculated using a Cox regression model. The proportional hazard assumption was assessed by calculating scaled Schoenfeld residuals, which did not reveal any violation. Potential risk factors for adjustment included the age, body mass index (BMI), smoking status, drinking status, hypertension, hypercholesterolemia, diabetes mellitus, history of CVD, basal ADL, energy intake (kcal), and number of follow-up visits, as well as the baseline eGFR. We conducted tests for trends across increasing quartiles of AOC intake by assigning a median value to each category to create a single variable. Sensitivity analyses were also performed a) to exclude age from our results, as the age includes the calculation of the eGFR,and b) to examine when the incidence date was defined as the examination date of the initial onset of CKD symptoms.

Possible interactions were tested for by introducing a multiplicative term into the main-effect model. We then performed additional subgroup analyses according to age (cut-off, 65 years), lifestyle factors (including current smoking and drinking; no/yes), overweight (cut-off BMI, 25 kg/m²), and hypertension (absence/presence) to evaluate the varied effects of H- or L-AOC intake on CKD risk related to potential risk factors. Although we found no significant evidence of heterogeneity in H- and L-AOC intake according to sex (p=0.3120 for H-AOC; p=0.2123 for L-AOC), all analyses were performed based on sex due to different distributions in dietary intakes and lifestyle characteristics by sex. Missing BMI values (n=10) were interpolated from the regression slope on age according to sex.

The SAS software program (ver. 9.4; SAS Institute Inc., Cary, NC, USA) was used for all analyses. A two-tailed p＜0.05 was considered statistically significant.

Results

During a median follow-up of 9.7 (range, 2.2–13.0) years, 137 participants (92 men and 45 women) developed CKD; 100 were diagnosed according to the eGFR alone, 24 according to proteinuria alone, and 13 according to both criteria simultaneously. Women were more likely to have healthier lifestyles than men, as suggested by lower smoking and drinking rates and a higher plant-based food intake; the BMI did not differ markedly between sexes (Supplementary Table 2).

Supplementary Table 2.Characteristics of the study participants by sex

	Total	By sex
		Women	Men
Number of participants	922	638	284
Men, %	30.8
Age in years, mean±SD	59.5±10.1	59.0±9.9	60.7±10.4
BMI in kg/m2, mean±SD	23.6±3.1	23.7±3.1	23.4±2.9
≥ 25 kg/m2, %	31.3	32.3	29.2
Hypertension, %	36.1	33.2	42.6
Diabetes, %	7.2	7.4	6.7
Hypercholesterolemia, %	26.5	29.9	18.7
History of cardiovascular disease, %	7.9	6.6	10.9
Current smokers, %	23.1	2.8	68.7
Current drinkers, %	39.6	22.7	77.5
Total energy and nutrient intakea unit/day, mean±SD
Total energy in kcal	1728±541	1644±502	1917±577
Protein in g	37±11	37±9	36±14
Sodium in mg	4444±2313	4509±2323	4300±2287
Potassium in mg	2441±757	2501±714	2308±833
H-AOC in μmol TE, median (IQR)	16787 (11072, 25644)	15432 (10734, 23707)	20179 (12255, 31516)
L-AOC in μmol TE, median (IQR)	2637 (1879, 3998)	2514 (1846, 3674)	2964 (1982, 4600)
Food intake in g/dayc, median (IQR)
Rice, bread, and noodles	489.6 (437.7, 620.1)	471.0 (408.6, 500.3)	678.0 (585.5, 870.2)
Beans	24.8 (13.6, 34.3)	26.3 (16.5, 35.1)	19.4 (7.7, 32.6)
Vegetables	185.3 (141.5, 237.1)	187.6 (149.9, 238.1)	179.8 (111.0, 236.7)
Fruits	51.3 (31.3, 77.3)	58.0 (37.9, 84.3)	37.6 (20.8, 60.7)
Mushroom	4.0 (2.8, 5.3)	4.3 (3.0, 5.6)	3.5 (2.1, 4.7)
Seaweeds	24.1 (13.5, 41.4)	22.8 (13.7, 37.1)	26.2 (12.4, 57.3)
Fish and shellfish	71.3 (51.1, 99.7)	69.6 (50.5, 96.0)	74.2 (52.9, 103.4)
Meats	37.5 (23.9, 55.3)	37.6 (26.6, 51.8)	36.9 (12.1, 63.2)
Beverages	178.8 (95.6, 307.3)	185.6 (105.3, 307.5)	157.1 (75.5, 306.8)
Seasonings and spices	56.3 (43.6, 74.2)	54.7 (44.5, 67.9)	62.0 (38.6, 85.7)
Serum creatinine in mg/dl, mean±SD	0.63±0.13	0.57±0.10	0.76±0.11
eGFR in ml/min/1.73 m2, mean±SDb	81.9±8.4	82.6±8.3	80.5±8.5
Basal ADL score, mean±SDc	5.02±1.17	4.89±1.24	5.30±0.93
Number of follow-up exams, mean±SD	2.75±1.20	2.77±1.19	2.70±1.23
Follow-up period in years, mean±SD	8.99±2.93	9.01±2.88	8.93±3.05

Abbreviations: ADL, Activity of Daily Living; BMI, Body Mass Index; eGFR, estimated Glomerular Filtration Rate; H-AOC, Hydrophilic Antioxidant Capacity; IQR, interquartile range; L-ORAC, Lipophilic Antioxidant Capacity; SD, Standard Deviation; TE, Trolox Equivalent

^a Data were adjusted for total energy using the residual method.

^b eGFR was estimated from the serum creatinine (SCr) value using eGFR_CKD-EPI (mL/min/1.73 m²) = 141×min (SCr/κ, 1)^α×max (SCr/κ, 1)^－1.209×0.993^Age×1.018 (if women)×0.813 (Japanese coefficient). κ: 0.7 in women and 0.9 in men, α; -0.329 in women and -0.411 in men, min indicates the minimum of SCr/κ or 1, and max indicates the maximum of SCr/κ or 1.

^c Basal ADL is measured using MOS with a range of 2-6.

The baseline characteristics across quartiles of each H- and L-AOC are summarized in Table 1 for women and Table 2 for men. Participants in the highest H-AOC and L-AOC quartiles tended to consume more vitamins and minerals and rice, whereas several differences in dietary intake between AOCs were observed, especially in women. Women with the highest H-AOC tended to eat more vegetables, seaweed, and beverages than others, while those with the highest L-AOC tended to eat more main dishes, such as beans, fish, and meat. In men, no remarkable dietary differences by quartiles were seen, with several exceptions, such as grain, beverages and seaweed intake among those with H-AOC intake. Unhealthy lifestyles, such as a current smoking or drinking habit; the prevalence of hypertension, diabetes, and hypercholesterolemia; a history of cardiovascular disease; mean values of serum creatinine, eGFR, and basal ADL score; number of follow-up exams; and follow-up period did not largely differ among categories. The percentage agreement and Spearman’s correlation coefficients between the H- and L-AOC quartiles were 64.9% and 0.8102 for women and 64.1% and 0.8141 for men, respectively; when restricted to participants with the highest L-AOC quartile, 79.9% of women and 74.7% of men were classified in the highest H-AOC quartile (not tabulated).

Table 1.Baseline characteristics of the study participants across quartiles of H- and L-AOC intake among women in the Ohasama study, Japan, 1998

	H-AOC				L-AOC
	Lowest	Second	Third	Highest	Lowest	Second	Third	Highest
AOC in μmol TE, median (IQR)	8428	12695	18678	30647	1532	2136	2966	4711
	(6335, 9589)	(11595, 14340)	(17087, 21247)	(26069, 37453)	(1225, 1712)	(1986, 2302)	(2797, 3272)	(4279, 5909)
Number of participants	159	160	160	159	159	160	160	159
Age in years, mean±SD	60.8±8.9	58.7±9.6	57.5±10.6	59.2±10.1	58.7±10.0	59.1±9.8	58.0±9.9	60.2±9.8
BMI in kg/m2, mean±SD	23.4±3.4	24.1±3.4	23.8±2.6	23.6±3.1	23.7±3.4	23.7±3.2	23.8±2.9	23.6±3.1
≥ 25 kg/m2, %	24.5	38.1	34.4	32.1	28.9	33.8	34.4	32.1
Hypertension, %	39.0	35.0	25.0	34.0	37.1	35.6	26.9	33.3
Diabetes, %	6.9	8.1	7.5	6.9	5.7	6.3	9.4	8.2
Hypercholesterolemia, %	24.5	33.8	31.9	29.6	26.4	32.5	30.0	30.8
History of cardiovascular disease, %	6.3	8.1	5.6	6.3	7.6	6.9	5.6	6.3
Current smokers, %	1.3	3.8	3.1	3.1	1.3	2.5	5.0	2.5
Current drinkers, %	13.8	21.9	26.9	28.3	19.5	19.4	25.0	27.0
Total energy and nutrient intakea unit/ day, mean±SD
Total energy in kcal	1769±506	1541±501	1553±432	1714±531	1688±508	1558±479	1627±477	1703±535
Protein in g	35±11	37±7	38±8	38±10	33±10	38±7	38±9	39±10
Sodium in mg	4213±2106	4542±2044	4434±1748	4846±3135	3728±1643	4499±1923	4581±2013	5226±3181
Potassium in mg	2270±702	2464±608	2562±623	2707±836	2107±685	2436±452	2621±689	2839±782
Folate in μg	265±126	293±97	322±126	372±143	243±123	285±79	330±132	394±129
Ascorbic acid in mg	61±35	72±30	81±33	99±42	57±36	69±25	82±35	105±38
Food intake in g/day, median (IQR)
Rice, bread, and noodles	491.8	471.4	468.9	448.2	488.0	473.5	468.4	438.6
	(454.0, 579.0)	(415.4, 499.4)	(384.6, 490.0)	(367.5, 485.0)	(458.0, 588.6)	(417.6, 500.0)	(390.5, 490.7)	(358.6, 485.0)
Beans	25.0	28.1	26.3	25.2	22.8	27.6	26.6	25.3
	(10.7, 32.6)	(19.1, 34.6)	(20.1, 36.8)	(15.9, 36.4)	(10.4, 32.4)	(20.8, 35.0)	(21.0, 36.0)	(16.4, 37.0)
Vegetables	173.0	186.0	195.6	204.0	169.2	188.1	206.1	205.9
	(135.2, 221.8)	(155.5, 234.4)	(158.3, 243.8)	(156.6, 266.3)	(125.0, 208.0)	(155.5, 230.7)	(163.0, 250.4)	(163.7, 274.3)
Fruits	52.3	60.0	63.4	56.3	52.3	57.9	60.0	63.5
	(32.1, 75.5)	(38.0, 90.4)	(42.5, 88.0)	(38.7, 86.8)	(31.3, 76.3)	(39.7, 83.4)	(38.4, 92.1)	(40.7, 84.9)
Mushroom	4.0 (2.3, 5.4)	4.3 (3.4, 5.6)	4.4 (3.5, 5.7)	4.1 (3.1, 5.8)	4.0 (2.5, 5.0)	4.4 (3.4, 5.7)	4.3 (3.3, 5.7)	4.5 (3.4, 5.9)
Seaweeds	15.9	24.7	26.1	31.5	20.1	26.9	22.8	23.8
	(5.3, 24.6)	(16.0, 34.9)	(14.9, 42.0)	(18.2, 92.7)	(9.3, 28.4)	(14.9, 39.3)	(14.1, 39.3)	(15.1, 46.4)
Fish and shellfish	57.6	65.3	73.2	79.5	52.2	67.7	78.2	83.5
	(39.3, 87.6)	(50.8, 94.5)	(55.8, 98.6)	(58.3, 106.3)	(36.5, 66.9)	(49.7, 94.4)	(55.8, 107.3)	(65.3, 116.1)
Meats	37.9	38.7	38.4	35.6	34.9	41.3	36.6	36.5
	(25.7, 60.3)	(26.5, 51.6)	(28.4, 50.6)	(23.1, 49.0)	(23.7, 49.0)	(30.4, 55.2)	(26.8, 52.9)	(26.0, 51.6)
Beverages	122.7	192.0	228.2	231.4	146.8	179.4	204.6	202.3
	(63.5, 210.8)	(117.7, 287.0)	(128.2, 417.8)	(127.3, 427.3)	(85.0, 285.7)	(114.9, 296.3)	(112.0, 318.8)	(111.0, 392.0)
Seasonings and spices	46.8	54.3	55.5	67.9	48.0	56.1	56.7	63.8
	(31.4, 55.4)	(45.2, 63.2)	(46.1, 68.2)	(53.1, 92.4)	(31.7, 54.8)	(45.8, 64.0)	(46.6, 81.5)	(51.1, 83.0)
Serum creatinine in mg/dl, mean±SD	0.57±0.10	0.57±0.09	0.57±0.10	0.58±0.10	0.58±0.10	0.57±0.09	0.57±0.10	0.58±0.10
eGFR in ml/min/1.73 m2, mean±SDb	81.6±8.2	82.9±7.8	83.6±8.9	82.2±8.4	82.6±8.6	82.9±8.0	83.4±8.6	81.4±8.0
Basal ADL score, mean±SDc	4.63±1.41	4.95±1.14	5.09±1.08	4.91±1.26	4.79±1.32	4.86±1.25	4.99±1.13	4.94±1.25
Number of follow-up exams, mean±SD	2.71±1.13	2.74±1.22	2.85±1.20	2.77±1.22	2.64±1.18	2.84±1.24	2.83±1.18	2.75±1.17
Follow-up period in years, mean±SD	8.88±2.93	8.94±2.86	9.15±2.93	9.08±2.80	8.81±2.90	9.21±2.87	9.03±2.93	8.99±2.82

Abbreviations: ADL, Activity of Daily Living; BMI, Body Mass Index; eGFR, estimated Glomerular Filtration Rate; H-AOC, Hydrophilic Antioxidant Capacity; IQR, interquartile range; L-AOC, Lipophilic Antioxidant Capacity; SD, Standard Deviation; TE, Trolox Equivalent

^a Data were adjusted for total energy using the residual method.

^b eGFR was estimated from the serum creatinine (SCr) value using eGFR_CKD-EPI (mL/min/1.73 m²) = 141×min (SCr/κ, 1)^α×max (SCr/κ, 1)^－1.209×0.993^Age×1.018 (if women)×0.813 (Japanese coefficient). κ: 0.7 in women and 0.9 in men, α; -0.329 in women and -0.411 in men, min indicates the minimum of SCr/κ or 1, and max indicates the maximum of SCr/κ or 1.

^c Basal ADL is measured using MOS with a range of 2-6.

Table 2.Baseline characteristics of the study participants across quartiles of H- and L-AOC intake among men in the Ohasama study, Japan, 1998

	H-AOC				L-AOC
	Lowest	Second	Third	Highest	Lowest	Second	Third	Highest
AOC in μmol TE, median (IQR)	7884	16448	25393	42806	1395	2431	3772	5624
	(5630, 11308)	(14291, 17449)	(22452, 28253)	(36330, 53593)	(884, 1746)	(2199, 2730)	(3291, 4088)	(5074, 7749)
Number of participants	71	71	71	71	71	71	71	71
Age in years, mean±SD	61.7±10.5	61.8±10.3	62.3±9.8	57.3±10.6	60.5±10.3	60.1±11.2	62.5±10.2	59.9±10.1
BMI in kg/m2, mean±SD	23.4±2.7	23.0±3.0	23.9±3.1	23.5±2.7	23.3±2.8	23.0±2.9	23.8±2.6	23.6±3.1
≥ 25 kg/m2, %	26.8	23.9	35.2	31.0	28.2	21.1	33.8	33.8
Hypertension, %	43.7	39.4	47.9	39.4	42.3	38	46.5	43.7
Diabetes, %	5.6	8.5	5.6	7.0	7.0	2.8	5.6	11.3
Hypercholesterolemia, %	19.7	15.5	15.5	23.9	14.1	16.9	21.1	22.5
History of cardiovascular disease, %	7.0	14.1	12.7	9.9	7.0	14.1	12.7	9.9
Current smokers, %	59.2	67.6	71.8	76.1	54.9	73.2	73.2	73.2
Current drinkers, %	81.7	73.2	81.7	73.2	74.7	80.3	83.1	71.8
Total energy and nutrient intakea unit/ day, mean±SD
Total energy in kcal	2068±632	1740±513	1920±633	1941±478	2172±561	1774±557	1855±578	1867±542
Protein in g	31±14	38±15	38±15	35±12	27±12	39±12	40±14	38±15
Sodium in mg	3715±1944	4746±2728	4367±2326	4371±1991	2919±1406	4651±2029	4884±2863	4744±2064
Potassium in mg	1985±725	2416±965	2476±846	2354±696	1684±610	2386±681	2621±871	2539±809
Folate in μg	236±108	290±130	317±120	334±123	206±104	286±111	323±115	362±115
Ascorbic acid in mg	46±25	64±37	72±31	85±36	39±23	61±29	75±34	91±32
Food intake in g/day, median (IQR)
Rice, bread, and noodles	798.0	635.0	651.2	689.1	883.0	639.1	609.3	643.8
	(665.1, 932.7)	(529.5, 851.3)	(572.1, 844.3)	(588.2, 793.3)	(743.0, 948.5)	(593.4, 751.4)	(506.2, 781.2)	(562.0, 810.6)
Beans	14.5	24.6	23.8	17.6	10.2	25.6	23.1	21.0
	(3.3, 29.0)	(9.7, 35.4)	(8.2, 40.3)	(8.5, 27.8)	(0.0, 19.6)	(12.2, 34.2)	(9.7, 39.0)	(9.1, 29.2)
Vegetables	138.1	186.1	201.5	182.6	119.5	200.0	214.8	187.1
	(100.3, 219.7)	(111.0, 236.8)	(118.7, 270.6)	(137.0, 225.4)	(55.6, 175.0)	(126.8, 243.2)	(133.7, 267.4)	(137.0, 250.1)
Fruits	32.1	36.7	42.2	39.1	28.2	36.0	42.7	43.2
	(12.3, 54.6)	(24.3, 62.8)	(22.0, 72.8)	(24.6, 58.6)	(10.8, 49.1)	(24.3, 56.8)	(22.0, 82.3)	(27.9, 60.8)
Mushroom	3.5 (1.7, 4.9)	3.3 (2.1, 4.5)	4.0 (2.6, 5.3)	3.1 (2.1, 4.3)	2.5 (1.5, 4.2)	3.6 (2.6, 4.7)	3.8 (2.7, 5.1)	3.3 (2.2, 4.8)
Seaweeds	12.0	31.6	31.1	68.5	12.5	33.1	32.1	31.7
	(2.7, 23.1)	(19.6, 48.6)	(15.5, 83.7)	(21.3, 123.7)	(2.7, 27.4)	(22.3, 78.6)	(18.5, 82.0)	(15.2, 76.4)
Fish and shellfish	56.0	79.5	80.3	78.8	46.1	74.4	84.9	88.5
	(35.1, 84.5)	(55.0, 101.9)	(53.4, 118.2)	(64.5, 107.9)	(29.4, 73.3)	(54.5, 93.2)	(59.4, 123.8)	(67.3, 121.9)
Meats	28.3	48.0	43.1	35.9	24.0	48.0	45.3	35.9
	(5.0, 51.1)	(16.5, 65.7)	(15.0, 72.6)	(12.3, 57.3)	(0.0, 44.2)	(21.2, 68.6)	(15.7, 70.0)	(15.0, 64.3)
Beverages	83.5	141.6	192.4	247.6	107.1	151.3	192.9	221.4
	(36.4, 166.1)	(84.4, 276.4)	(93.4, 344.9)	(113.4, 487.7)	(39.1, 180.4)	(77.8, 278.6)	(94.7, 337.2)	(84.8, 400.3)
Seasonings and spices	39.6	63.7	74.7	80.1	36.1	71.0	76.5	68.9
	(21.9, 58.8)	(42.9, 76.0)	(45.5, 97.0)	(49.5, 132.0)	(21.3, 50.2)	(55.6, 90.2)	(53.5, 103.3)	(46.0, 98.5)
Serum creatinine in mg/dl, mean±SD	0.74±0.11	0.76±0.10	0.78±0.11	0.75±0.11	0.76±0.10	0.74±0.10	0.76±0.11	0.77±0.12
eGFR in ml/min/1.73 m2, mean±SDb	80.8±8.0	79.7±7.2	78.7±8.6	82.8±9.5	80.4±7.8	81.9±8.4	79.3±8.1	80.4±9.4
Basal ADL score, mean±SDc	5.17±1.03	5.20±1.06	5.39±0.85	5.45±0.71	5.23±0.93	5.34±1.00	5.21±1.03	5.44±0.73
Number of follow-up exams, mean±SD	2.54±1.30	2.66±1.21	2.82±1.23	2.80±1.17	2.77±1.26	2.59±1.25	2.59±1.15	2.86±1.25
Follow-up period in years, mean±SD	8.77±2.91	8.46±3.20	8.91±3.17	9.58±2.85	8.99±2.90	8.86±3.17	8.82±2.84	9.05±3.33

Abbreviations: ADL, Activity of Daily Living; BMI, Body Mass Index; eGFR, estimated Glomerular Filtration Rate; H-AOC, Hydrophilic Antioxidant Capacity; IQR, interquartile range; L-AOC, Lipophilic Antioxidant Capacity; SD, Standard Deviation; TE, Trolox Equivalent

^a Data were adjusted for total energy using the residual method.

^b eGFR was estimated from the serum creatinine (SCr) value using eGFR_CKD-EPI (mL/min/1.73 m²) = 141×min (SCr/κ, 1)^α×max (SCr/κ, 1)^－1.209×0.993^Age×1.018 (if women)×0.813 (Japanese coefficient). κ: 0.7 in women and 0.9 in men, α; -0.329 in women and -0.411 in men, min indicates the minimum of SCr/κ or 1, and max indicates the maximum of SCr/κ or 1.

c Basal ADL is measured using MOS with a range of 2-6.

HRs on the association of each H- and L-AOC intake for CKD incidence are shown separately in women (Fig.2) and men (Fig.3). In the multivariable-adjusted analyses, women in the highest L-AOC quartile had a significantly lower CKD risk those in the lowest quartile (HRs, 95% CI compared with the lowest L-AOC intake quartile: 0.49, 0.27-0.90). The association did not differ even after further adjusting for the H-AOC, sodium, potassium, folate, and vitamin C intake. The highest L-AOC intake was inversely and significantly associated with the diagnosis according to an eGFR of ＜60 ml/min/1.73 m² but was not significantly associated with the occurrence of proteinuria (0.49, 0.25–0.97 for eGFR of ＜60 ml/min/1.73 m²; 0.35, 0.10–1.19 for positive proteinuria). The highest L-AOC intake in men was not associated with a reduced CKD risk. The H-AOC intake was not associated with new-onset CKD in either men or women. Sensitivity analyses which excluded age as a variable did not alter our results (vs. lowest L-AOC intake: 0.59, 0.37–0.95). An additional analysis with the incidence date was defined as the examination date of the initial onset of CKD symptoms instead of the midpoint between the examination date before the CKD onset and the initial onset of CKD symptoms. Using either analysis method, a significant association was found between the L-AOC intake and the development of CKD in women, thus suggesting that our results were robust (data not shown).

Fig.2. Adjusted hazard ratios and 95% confidence intervals assessing the association between incident chronic kidney disease and the hydrophilic- and lipophilic-antioxidant capacity (AOC) intake among women in the Ohasama Study, 1997-2010

Abbreviations: ADL, Activity of Daily Living; CKD_CKD-EPI, Chronic Kidney Disease; TE, Trolox Equivalent

Filled squares represent point estimates, and horizontal lines denote the 95% confidence intervals.

CKD was defined as an estimated glomerular filtration rate (eGFR) of ＜60 ml/min/1.73 m² and/or positive proteinuria.

Adjusted for age, body mass index, ever smoking, ever drinking, hypertension, hypercholesterolemia, diabetes mellitus, history of cardiovascular disease, basal activity of daily living, energy intake, number of follow-up visits, and baseline eGFR.

^† AOC was adjusted for total energy using the residual method.

Fig.3. Adjusted hazard ratios and 95% confidence intervals assessing the association between incident chronic kidney disease and the hydrophilic- and lipophilic-antioxidant capacity (AOC) intake among men in the Ohasama Study, 1997-2010

Abbreviations: ADL, Activity of Daily Living; CKD_CKD-EPI, Chronic Kidney Disease; TE, Trolox Equivalent

Filled squares represent point estimates, and horizontal lines denote the 95% confidence intervals.

CKD was defined as an estimated glomerular filtration rate (eGFR) of ＜60 ml/min/1.73 m² and/or positive proteinuria.

Adjusted for age, body mass index, ever smoking, ever drinking, hypertension, hypercholesterolemia, diabetes mellitus, history of cardiovascular disease, basal activity of daily living, energy intake, number of follow-up visits, and baseline eGFR.

^† AOC was adjusted for total energy using the residual method.

Stratification analyses according to age, lifestyle, and disease status were also performed, as shown in Supplementary Table 3 for women and 4 for men. There were no interactions among H-/L-AOC, age and lifestyle variables, except for hypertension (hypertension×H-AOC, p=0.08 in men; p=0.06 in women).

Supplementary Table 3.Summary of stratification results from hazard ratios^a (95% confidence interval) for incidence of chronic kidney disease according to quartile of H- and L-AOC intake^b in women, Ohasama study, 1997-2010

	Event/ Participants	Quartile				P for interaction
		Lowest	Second	Third	Highest
H-AOC
Age						0.86
＜65 years	73/441	Reference	0.75 (0.30–1.88)	0.68 (0.26–1.74)	0.75 (0.30–1.87)
≥ 65 years	57/198	Reference	0.78 (0.34–1.83)	0.92 (0.40–2.10)	0.80 (0.37–1.74)
BMI						0.70
＜25 kg/m2	90/428	Reference	0.78 (0.35–1.74)	1.06 (0.52–2.17)	0.71 (0.35–1.47)
≥ 25 kg/m2	40/211	Reference	0.73 (0.25–2.14)	0.41 (0.11–1.50)	0.60 (0.19–1.90)
Hypertension						0.06
Absence	71/424	Reference	0.93 (0.37–2.29)	0.83 (0.33–2.10)	1.32 (0.58–3.01)
Presence	59/215	Reference	0.52 (0.22–1.25)	0.64 (0.27–1.56)	0.39 (0.16–0.93)
Smoking status						0.15
No	123/620	Reference	0.80 (0.43–1.49)	0.74 (0.40–1.38)	0.68 (0.38–1.24)
Yes	7/19	Reference	ne.	ne.	ne.
Drinking status						0.68
No	101/491	Reference	0.62 (0.32–1.22)	0.71 (0.37–1.38)	0.73 (0.39–1.34)
Yes	29/148	Reference	1.53 (0.15–15.74)	1.17 (0.10–13.16)	0.49 (0.04–6.09)
L-AOC
Age						0.96
＜65 years	73/441	Reference	0.48 (0.18–1.25)	0.90 (0.40–2.03)	0.45 (0.16–1.22)
≥ 65 years	57/198	Reference	0.54 (0.23–1.24)	0.41 (0.17–1.001)	0.48 (0.22–1.06)
BMI						0.94
＜25 kg/m2	90/428	Reference	0.50 (0.23–1.07)	0.60 (0.28–1.26)	0.41 (0.20–0.86)
≥ 25 kg/m2	40/211	Reference	0.64 (0.21–1.92)	0.89 (0.30–2.67)	0.42 (0.12–1.44)
Hypertension						0.13
Absence	71/424	Reference	0.66 (0.26–1.69)	1.26 (0.56–2.84)	0.71 (0.29–1.75)
Presence	59/215	Reference	0.48 (0.21–1.08)	0.24 (0.09–0.67)	0.33 (0.13–0.79)
Smoking status						0.15
No	123/620	Reference	0.49 (0.26–0.93)	0.71 (0.39–1.27)	0.48 (0.26–0.89)
Yes	7/19	Reference	ne.	ne.	ne.
Drinking status						0.55
No	101/491	Reference	0.50 (0.26–0.97)	0.62 (0.32–1.18)	0.51 (0.27–0.97)
Yes	29/148	Reference	1.42 (0.18–11.28)	0.57 (0.07–4.80)	0.32 (0.04–2.50)

Abbreviations: BMI, Body Mass Index; H-AOC, Hydrophilic Antioxidant Capacity; L-AOC, Lipophilic Antioxidant Capacity.

CKD was defined as an eGFR of ＜60 ml/min/1.73 m² and/or positive proteinuria.

^a Adjusted for age, BMI, ever smoking, ever drinking, hypertension, hypercholesterolemia, diabetes mellitus, history of cardiovascular disease, basal

activity of daily living, energy intake, number of follow-up visits, and baseline eGFR.

^b AOC was adjusted for total energy using the residual method.

ne. not able to estimate

Supplementary Table 4.Summary of stratification results from hazard ratios^a (95% confidence interval) for incidence of chronic kidney disease according to quartile of H- and L-AOC intake^b in men, Ohasama study, 1997-2010

	Event/ Participants	Quartile				P for interaction
		Lowest	Second	Third	Highest
H-AOC
Age						0.24
＜65 years	24/162	Reference	4.24 (0.38–47.82)	1.73 (0.13–23.42)	3.05 (0.30–31.12)
≥ 65 years	32/122	Reference	4.31 (1.04–17.87)	1.50 (0.37–6.12)	2.47 (0.53–11.55)
BMI						0.45
＜25 kg/m2	36/202	Reference	7.42 (1.50–36.70)	2.30 (0.48–11.10)	3.38 (0.70–16.26)
≥ 25 kg/m2	20/82	Reference	2.60 (0.35–19.43)	0.70 (0.08–5.83)	2.43 (0.28–20.75)
Hypertension						0.08
Absence	28/163	Reference	1.32 (0.29–5.99)	1.09 (0.21–5.74)	0.44 (0.07–2.61)
Presence	28/121	Reference	8.59 (1.14–64.57)	2.66 (0.36–19.88)	7.61 (1.12–51.66)
Smoking status						0.89
No	15/86	Reference	4.12 (0.18–94.70)	0.13 (0.00–15.25)	3.74 (0.23–61.35)
Yes	41/198	Reference	4.56 (1.11–18.73)	2.49 (0.60–10.36)	2.65 (0.55–12.74)
Drinking status						0.12
No	14/61	Reference	3.03 (0.48–18.99)	0.23 (0.01–3.73)	0.05 (0.01–0.75)
Yes	42/223	Reference	5.73 (1.09–30.19)	2.48 (0.44–13.88)	5.15 (0.95–27.98)
L-AOC
Age						0.28
＜65 years	24/162	Reference	6.96 (0.74–65.29)	0.85 (0.04–16.11)	1.005 (0.13–7.85)
≥ 65 years	32/122	Reference	0.99 (0.28–3.53)	1.20 (0.35–4.12)	0.80 (0.19–3.35)
BMI						0.36
＜25 kg/m2	36/202	Reference	1.99 (0.54–7.32)	1.11 (0.25–4.88)	1.59 (0.40–6.34)
≥ 25 kg/m2	20/82	Reference	0.31 (0.02–4.21)	0.46 (0.06–3.48)	0.50 (0.05–4.96)
Hypertension						0.20
Absence	28/163	Reference	0.70 (0.16–3.06)	0.74 (0.15–3.65)	0.33 (0.06–1.76)
Presence	28/121	Reference	1.35 (0.24–7.71)	1.71 (0.34–8.64)	1.81 (0.39–8.53)
Smoking status						0.70
No	15/86	Reference	3.08 (0.07–129.04)	0.41 (0.02–9.95)	1.70 (0.23–12.56)
Yes	41/198	Reference	1.30 (0.38–4.40)	1.40 (0.40–4.96)	0.93 (0.22–3.83)
Drinking status						0.10
No	14/61	Reference	1.30 (0.38–4.40)	1.40 (0.40–4.96)	0.93 (0.22–3.83)
Yes	42/223	Reference	3.83 (0.68–21.48)	2.63 (0.49–14.31)	3.03 (0.50–18.42)

Abbreviations: BMI, Body Mass Index; H-AOC, Hydrophilic Antioxidant Capacity; L-AOC, Lipophilic Antioxidant Capacity. CKD was defined as an eGFR of ＜60 ml/min/1.73 m² and/or positive proteinuria.

^a Adjusted for age, BMI, ever smoking, ever drinking, hypertension, hypercholesterolemia, diabetes mellitus, history of cardiovascular disease, basal

activity of daily living, energy intake, number of follow-up visits, and baseline eGFR.

^b AOC was adjusted for total energy using the residual method.

Discussion

In the prospective population-based study in Japanese adults, an increased intake of L-AOC was significantly associated with a reduced CKD risk among women. These associations were not seen for the intakes of H-AOC in either sex or for L-AOC in men.

This study has several strengths. It was the first study to examine the association between the dietary AOC intake and CKD prevention while considering AOCs from major food items in the Japanese diet, an aspect that has not been adequately evaluated in previous studies^{8, 9)}. The sampling protocol in the AOC database used for the calculations attempted to consider the potential variations that might exist in the Japanese market and reflect the typical diet of Japanese consumers, such as rice and seafood, which has not been measured in other countries¹⁴⁾. Our study was also the first study to examine the relationship of the disease outcome with each intake of hydrophilic and lipophilic AOCs, separately.

Only one previous study reported findings consistent with our own⁶⁾; that study followed Iranian dysglycemia patients without CKD for three years and found an association between the dietary AOC intake and a reduced CKD risk. Several potential mechanisms underlying the association between dietary AOC and CKD have been identified. First, the protective function of dietary antioxidants against oxidative stress toward endothelial dysfunction can be considered. Oxidative stress has been widely reported as a cause of endothelial dysfunction that is greatly associated with the progression of an impaired renal function^{7, 32)}. In this pathway, antioxidants are hypothesized to inhibit the atherosclerotic process by reducing oxidation or scavenging reactive oxygen species³³⁾. Other potential protective mechanisms by which dietary antioxidants act against the atherosclerotic process include reducing platelet aggregation and blood pressure as well as exerting anti-inflammatory effects^{34, 35)}. Indeed, previous prospective studies have confirmed that dietary AOCs were associated with a lower risk of stroke³⁶⁾, CVD³⁷⁾, and type 2 diabetes³⁸⁾, and these diseases have also been implicated in endothelial dysfunction through oxidative stress. Although endothelial dysfunction and disorders related to atherosclerosis may be potential mediators of AOC and CKD, our study design could not identify a temporal relationship between AOC intake and the present disease. Further research that includes an analysis of factors that mediate the disease onset during the follow-up period is expected to elucidate the underlying mechanism.

Although the present study examined the association between the AOC intake from the overall Japanese diet and CKD, preclinical and clinical studies to examine the effect of a single or combined antioxidant intake, such as vitamin C and beta-carotene, on disease prevention or mortality have shown inconsistent results indicating either positive or negative associations^{35, 39, 40)}. Antioxidants exert their antioxidant properties by being oxidized themselves; therefore, antioxidants can exert pro-oxidant effects^{41, 42)}. Our findings that the AOC intake of different types of antioxidants from the whole diet was inversely associated with CKD incidence might suggest that the beneficial effects of additive synergies from consuming various antioxidants led to a reduced disease risk⁴³⁾.

We further revealed in our study that the intakes of hydrophilic and lipophilic antioxidants have differing effects on the prevention of CKD. These AOC values were frequently used as a single indicator without distinguishing between H-AOC and L-AOC in previous epidemiological studies^{8, 9)}. In addition, no previous reports have focused on the difference between each water- or lipid-soluble antioxidant capacity and the disease outcome. In general, water-soluble antioxidants suppress oxidation in the cytoplasmic substrate and plasma, whereas lipid-soluble antioxidants prevent lipid peroxidation in the cell membranes. Given that AOC intake does not always linearly correlate with the free radical scavenging capacity¹²⁾, the efficacy of antioxidants may depend on concentrations, localizations, physiological mobilities, and interactions of the oxidants and/or antioxidants. Although the L-AOC intake was approximately 1/10 of the H-AOC intake, lipophilic components have been confirmed as having different functions and/or metabolic pathways in the body because of differences in the physicochemical properties of hydrophilic components^{12, 13)}. A recent report demonstrated that L-AOC has a high biological activity even at low concentrations but is more effective above a certain concentration⁴⁴⁾. Although the correlation coefficient between H- and L-AOC was high in the present study, the characteristics of foods with a high L-AOC did not necessarily coincide with those with a high H-AOC. Selecting food items based on L-AOC may aid in CKD prevention. Our previous study reported that the most commonly consumed types of L-AOC food groups from a Japanese diet were soybean products, fish and shellfish, vegetables, and seaweed, which have not been measured in most studies from Western countries¹⁴⁾. Western diets are generally different from Japanese diets in terms of food intake characteristics, with fewer consuming soybeans, fish, and shellfish and many not eating seaweed at all¹¹⁾, so Western diets, except for in some regions, such as the Mediterranean, may have a relatively low L-AOC intake. Further research is expected concerning the relationship between the H- and L-AOC intake from a whole diet and disease outcomes in other countries.

In the present study, the significant preventive effect of an increased L-AOC intake was only seen in women, with these associations not noted among men, although no significant interaction was seen between sex and the AOC intake. Preventive effects of antioxidant intake are also affected by pro-oxidant factors, such as aging, diabetes, obesity, and smoking^{45, 46)}. As individuals with an increased risk of CKD at baseline were excluded, healthy and relatively young women were more likely to be followed up than others. Even in the presence of an existing disease, dietary AOC may have further preventative effects among women who maintain a healthy lifestyle, whereas the protective effect in men may be weakened by underlying diseases and unhealthy lifestyle factors related to pro-oxidation, such as smoking and drinking. However, the results in men showed opposite trend point estimates. In particular, the H-AOC intake in men was positively associated with the incidence of CKD. However, the possibility of selection bias due to the small sample size and the survival effects of aging makes it difficult to conclude that the risk effects for CKD from dietary AOC intake are the opposite for men and women. Indeed, in comparison to the other quartiles of H-AOC in men, the second H-AOC quartile in men had a higher prevalence of diabetes mellitus and CVD history, lower energy intake, and higher sodium and meat intake. These variables may have been associated with the absence—with several exceptions—of remarkable interquartile dietary characteristics in men. Given the lack of interaction between sex and AOC intake, the effect of inter-individual lifestyle differences or unknown confounding factors may be greater in men than in women. More detailed studies of factors influencing the antioxidant and pro-oxidant balance are needed to prevent future deterioration of CKD.

Several limitations associated with the present study warrant mention. First, since dietary estimates were only performed in a single dietary recall at baseline, information regarding dietary intake modification during the follow-up period was not considered. Due to the prospective and population-based design and the exclusion of CKD at baseline, the number of participants who changed their diet due to illness is estimated to have been small. Although the validity of this method has been well established and widely accepted in epidemiological studies, the misclassification of exposure due to a limited number of items and minimal information regarding portion sizes is also possible. Second, this study estimated the dietary intake information in participants from 1997 to 1998. Our study participants lived in rural areas, and there are still no shops for luxury foods, fast-food restaurants, and convenience store in their local environment; the current findings may therefore only be applicable to rural areas in Japan. The limited generalizability remains evident, especially for large cities with a large sample size. Marked differences also exist in the epidemiology of dietary intakes and lifestyles between Japan and Western countries, especially in aging societies. Further prospective studies with more detailed information on aging-associated factors in other ethnic and cultural populations will be required to confirm the generalizability of our findings. Finally, the ORAC method evaluated only one aspect of the antioxidant potential of foods, and its functional efficacy on the prevention of CKD was shown by that method⁴⁷⁾. AOC values therefore cannot be directly compared using other methods, such as the Ferric Ion Reducing Antioxidant Power and Trolox Equivalent Antioxidant Capacity⁴⁸⁾, because of different mechanisms and radical or oxidant sources. Although several compounds responsible for ORAC have been identified⁴⁹⁾, other compounds responsible for high AOC values are still largely unknown. The utilization of more than one method to determine AOC values in conjunction with the elucidation of compounds contained in AOC might enable a more accurate measurement of AOC in the diet.

Conclusion

The CKD risk may be reduced by an increased intake of lipophilic antioxidants found in the Japanese diet. Along with healthy lifestyle habits, consuming a healthy diet rich in antioxidants may prevent future CKD occurrence. Of note, our study only showed the usefulness of antioxidant from the regular diet, with no suggestion to expand the antioxidant supplement intake.

Acknowledgements

The authors are grateful to the residents of Ohasama Town, all related investigators and study staff, and the staff members of the Hanamaki City Government, Iwate Prefectural Central Hospital Attachment Ohasama Regional Clinical Center (the former Ohasama Hospital), and Iwate Prefectural Stroke Registry for their valuable support on this project.

Notice of Grant Support

This study was supported by Grants for Scientific Research, Ministry of Education, Culture, Sports, Science and Technology, Japan (18K09674, 18K09904, 18K17396, 19K19325, 19K19466, 19H03908, 19K10662, 20K08612, 20K18819, 21K10452, 21K10478, 21H04854, 21K17313, 21K19670, 22H03358, 22K10070, 23K07690 and 23K09698); The internal research grants from Keio University; the Japan Arteriosclerosis Prevention Fund; Grant–in–aid from the Ministry of Health, Labor, and Welfare, Japan (H29–Junkankitou–Ippan–003 and 20FA1002); ACRO Incubation Grants of Teikyo University; The Academic Contributions from Pfizer Japan Inc. and Bayer Yakuhin, Ltd; Scholarship donations from Daiichi Sankyo Co.,Ltd.; Research Support from Astellas Pharma Inc. and Takeda Pharmaceutical Co.,Ltd.; Health Science Center Research Grant; Takeda Science Foundation.

Conflicts of Interest

H.M., K.A., Y.I., and T.O. concurrently held the position of director of the Tohoku Institute for the Management of Blood Pressure, supported by Omron Healthcare Co., Ltd. MS received a scholarship donation (Academic support program) from Bayer Yakuhin Co., Ltd. K.A. received honoraria from Takeda Pharmaceutical Co., Ltd. K.A. and T.O. received a joint research grant from Omron Healthcare Co., Ltd. There is no COI for other authors.

Author Contributions

M.T-U.: Conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, supervision, validation, visualization, and writing of the original draft. M.S., J.W., J.T., T.Oki., Y.T., K.A., T.M., T.H., H.M., A.H., K.N. and Y.I.: funding acquisition, investigation, methodology, project administration, review, and editing. M.K.: data curation, funding acquisition, investigation, methodology, project administration, review, and editing. T.Ohkubo.: Principal investigator of the Ohasama study, conceptualization, funding acquisition, investigation, methodology, project administration, visualization, review, and editing. All authors contributed to this scientific work and approved the final version of the manuscript.

References

• 1)  Imai E, Horio M, Watanabe T, Iseki K, Yamagata K, Hara S, Ura N, Kiyohara Y, Moriyama T, Ando Y, Fujimoto S, Konta T, Yokoyama H, Makino H, Hishida A and Matsuo S: Prevalence of chronic kidney disease in the Japanese general population. Clin Exp Nephrol, 2009; 13: 621-630
• 2)  Go AS, Chertow GM, Fan D, McCulloch CE and Hsu CY: Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med, 2004; 351: 1296-1305
• 3)  Ninomiya T, Kiyohara Y, Tokuda Y, Doi Y, Arima H, Harada A, Ohashi Y, Ueshima H and Japan Arteriosclerosis Longitudinal Study G: Impact of kidney disease and blood pressure on the development of cardiovascular disease: an overview from the Japan Arteriosclerosis Longitudinal Study. Circulation, 2008; 118: 2694-2701
• 4)  Nakayama M, Metoki H, Terawaki H, Ohkubo T, Kikuya M, Sato T, Nakayama K, Asayama K, Inoue R, Hashimoto J, Totsune K, Hoshi H, Ito S and Imai Y: Kidney dysfunction as a risk factor for first symptomatic stroke events in a general Japanese population--the Ohasama study. Nephrol Dial Transplant, 2007; 22: 1910-1915
• 5)  Gansevoort RT, Matsushita K, van der Velde M, Astor BC, Woodward M, Levey AS, de Jong PE, Coresh J and Chronic Kidney Disease Prognosis C: Lower estimated GFR and higher albuminuria are associated with adverse kidney outcomes. A collaborative meta-analysis of general and high-risk population cohorts. Kidney Int, 2011; 80: 93-104
• 6)  Asghari G, Yuzbashian E, Shahemi S, Gaeini Z, Mirmiran P and Azizi F: Dietary total antioxidant capacity and incidence of chronic kidney disease in subjects with dysglycemia: Tehran Lipid and Glucose Study. Eur J Nutr, 2018; 57: 2377-2385
• 7)  Nezu M, Souma T, Yu L, Suzuki T, Saigusa D, Ito S, Suzuki N and Yamamoto M: Transcription factor Nrf2 hyperactivation in early-phase renal ischemia-reperfusion injury prevents tubular damage progression. Kidney Int, 2017; 91: 387-401
• 8)  Kobayashi S, Murakami K, Sasaki S, Uenishi K, Yamasaki M, Hayabuchi H, Goda T, Oka J, Baba K, Ohki K, Watanabe R and Sugiyamama Y: Dietary total antioxidant capacity from different assays in relation to serum C-reactive protein among young Japanese women. Nutr J, 2012; 11: 91
• 9)  Tatsumi Y, Ishihara J, Morimoto A, Ohno Y and Watanabe S: Seasonal differences in total antioxidant capacity intake from foods consumed by a Japanese population. Eur J Clin Nutr, 2014; 68: 799-803
• 10)  Wakagi M, Watanabe J and Takano-Ishikawa Y: Effects of producing area and heavest season on antioxidant capacities of spinach, komatsuna, tomato, and cucumber. Report of National Food Research Institute, 2014; 78: 65-71
• 11)  FAOSTAT Data [on-line]. available at https: //www.fao.org/faostat/en/#data/SCL Accessed Dec. 12, 2022;
• 12)  Niki E, Omata Y, Fukuhara A, Saito Y and Yoshida Y: Assessment of radical scavenging capacity and lipid peroxidation inhibiting capacity of antioxidant. J Agric Food Chem, 2008; 56: 8255-8260
• 13)  Prior RL, Gu L, Wu X, Jacob RA, Sotoudeh G, Kader AA and Cook RA: Plasma antioxidant capacity changes following a meal as a measure of the ability of a food to alter in vivo antioxidant status. J Am Coll Nutr, 2007; 26: 170-181
• 14)  Tsubota-Utsugi M, Watanabe J, Takebayashi J, Oki T, Tsubono Y and Ohkubo T: The major source of antioxidants intake from typical diet among the rural farmer in north-eastern Japan in 1990s. J Epidemiol, 2021; 31: 101-108
• 15)  Nakamura Y, Ueshima H, Okamura T, Kadowaki T, Hayakawa T, Kita Y, Abbott RD, Okayama A, National Integrated Project for Prospective Observation of Non-Communicable D and its Trends in the Aged RG: A Japanese diet and 19-year mortality: national integrated project for prospective observation of non-communicable diseases and its trends in the aged, 1980. Br J Nutr, 2009; 101: 1696-1705
• 16)  Shimazu T, Kuriyama S, Hozawa A, Ohmori K, Sato Y, Nakaya N, Nishino Y, Tsubono Y and Tsuji I: Dietary patterns and cardiovascular disease mortality in Japan: a prospective cohort study. Int J Epidemiol, 2007; 36: 600-609
• 17)  Okada E, Nakamura K, Ukawa S, Wakai K, Date C, Iso H and Tamakoshi A: The Japanese food score and risk of all-cause, CVD and cancer mortality: the Japan Collaborative Cohort Study. Br J Nutr, 2018; 120: 464-471
• 18)  Nanri A, Mizoue T, Shimazu T, Ishihara J, Takachi R, Noda M, Iso H, Sasazuki S, Sawada N, Tsugane S and Japan Public Health Center-Based Prospective Study G: Dietary patterns and all-cause, cancer, and cardiovascular disease mortality in Japanese men and women: The Japan public health center-based prospective study. PLoS One, 2017; 12: e0174848
• 19)  Ohkubo T, Asayama K and Imai Y: The value of self-measured home blood pressure in predicting stroke. Expert Rev Neurother, 2006; 6: 163-173
• 20)  Tsuji I, Imai Y, Nagai K, Ohkubo T, Watanabe N, Minami N, Itoh O, Bando T, Sakuma M, Fukao A, Satoh H, Hisamichi S and Abe K: Proposal of reference values for home blood pressure measurement: prognostic criteria based on a prospective observation of the general population in Ohasama, Japan. Am J Hypertens, 1997; 10: 409-418
• 21)  Tsubono Y, Ogawa K, Watanabe Y, Nishino Y, Tsuji I, Watanabe T, Nakatsuka H, Takahashi N, Kawamura M and Hisamichi S: Food frequency questionnaire and a screening test. Nutr Cancer, 2001; 39: 78-84
• 22)  Ministry of Education C, Sports, Science and Technology: Standard Tables of Food Composition in Japan 2010., Ministry of Education, Culture, Sports, Science and Technology, Tokyo, 2010
• 23)  Takebayashi J, Oki T, Chen J, Sato M, Matsumoto T, Taku K, Tsubota-Utsugi M, Watanabe J and Ishimi Y: Estimated average daily intake of antioxidants from typical vegetables consumed in Japan: a preliminary study. Biosci Biotechnol Biochem, 2010; 74: 2137-2140
• 24)  Takebayashi J, Oki T, Watanabe J, Yamasaki K, Chen J, Sato-Furukawa M, Tsubota-Utsugi M, Taku K, Goto K, Matsumoto T and Ishimi Y: Hydrophilic antioxidant capacities of vegetables and fruits commonly consumed in Japan and estimated average daily intake of hydrophilic antioxidants from these foods. J Food Compos Anal, 2013; 29: 25-31
• 25)  Watanabe J, Oki T, Takebayashi J and Takano-Ishikawa Y: Extraction efficiency of hydrophilic and lipophilic antioxidants from lyophilized foods using pressurized liquid extraction and manual extraction. J Food Sci, 2014; 79: C1665-1671
• 26)  Watanabe J, Oki T, Takebayashi J, Yada H, Wagaki M, Takano-Ishikawa Y and Yasui A: Improvement and Interlaboratory Validation of the Lipophilic Oxygen Radical Absorbance Capacity: Determination of Antioxidant Capacities of Lipophilic Antioxidant Solutions and Food Extracts. Anal Sci, 2016; 32: 171-175
• 27)  Watanabe J, Oki T, Takebayashi J, Yamasaki K, Takano-Ishikawa Y, Hino A and Yasui A: Method validation by interlaboratory studies of improved hydrophilic oxygen radical absorbance capacity methods for the determination of antioxidant capacities of antioxidant solutions and food extracts. Anal Sci, 2012; 28: 159-165
• 28)  Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J and Ckd EPI: A new equation to estimate glomerular filtration rate. Ann Intern Med, 2009; 150: 604-612
• 29)  Horio M, Imai E, Yasuda Y, Watanabe T and Matsuo S: Modification of the CKD epidemiology collaboration (CKD-EPI) equation for Japanese: accuracy and use for population estimates. Am J Kidney Dis, 2010; 56: 32-38
• 30)  Horio M and Orita Y: Comparison of Jaffe rate assay and enzymatic method for the measurement of creatinine clearance. Nihon Jinzo Gakkai Shi, 1996; 38: 296-299
• 31)  Ando Y, Ito S, Uemura O, Kato T, Kimura G, Nakao T, Hattori M, Fukagawa M, Horio M, Mitarai T and Japanese Society of N: CKD Clinical Practice Guidebook. The essence of treatment for CKD patients. Clin Exp Nephrol, 2009; 13: 191-248
• 32)  Annuk M, Zilmer M and Fellstrom B: Endothelium-dependent vasodilation and oxidative stress in chronic renal failure: impact on cardiovascular disease. Kidney Int Suppl, 2003; S50-53
• 33)  Madamanchi NR, Vendrov A and Runge MS: Oxidative stress and vascular disease. Arterioscler Thromb Vasc Biol, 2005; 25: 29-38
• 34)  Grassi D, Desideri G, Croce G, Tiberti S, Aggio A and Ferri C: Flavonoids, vascular function and cardiovascular protection. Curr Pharm Des, 2009; 15: 1072-1084
• 35)  Valtuena S, Pellegrini N, Franzini L, Bianchi MA, Ardigo D, Del Rio D, Piatti P, Scazzina F, Zavaroni I and Brighenti F: Food selection based on total antioxidant capacity can modify antioxidant intake, systemic inflammation, and liver function without altering markers of oxidative stress. Am J Clin Nutr, 2008; 87: 1290-1297
• 36)  Rautiainen S, Larsson S, Virtamo J and Wolk A: Total antioxidant capacity of diet and risk of stroke: a population-based prospective cohort of women. Stroke, 2012; 43: 335-340
• 37)  Bastide N, Dartois L, Dyevre V, Dossus L, Fagherazzi G, Serafini M and Boutron-Ruault MC: Dietary antioxidant capacity and all-cause and cause-specific mortality in the E3N/EPIC cohort study. Eur J Nutr, 2017; 56: 1233-1243
• 38)  Mancini FR, Affret A, Dow C, Balkau B, Bonnet F, Boutron-Ruault MC and Fagherazzi G: Dietary antioxidant capacity and risk of type 2 diabetes in the large prospective E3N-EPIC cohort. Diabetologia, 2018; 61: 308-316
• 39)  Bjelakovic G, Nikolova D, Gluud LL, Simonetti RG and Gluud C: Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases. Cochrane Database Syst Rev, 2012; 2012: CD007176
• 40)  Del Rio D, Agnoli C, Pellegrini N, Krogh V, Brighenti F, Mazzeo T, Masala G, Bendinelli B, Berrino F, Sieri S, Tumino R, Rollo PC, Gallo V, Sacerdote C, Mattiello A, Chiodini P and Panico S: Total antioxidant capacity of the diet is associated with lower risk of ischemic stroke in a large Italian cohort. J Nutr, 2011; 141: 118-123
• 41)  Halliwell B: The antioxidant paradox. Lancet, 2000; 355: 1179-1180
• 42)  Huang HY and Appel LJ: Supplementation of diets with alpha-tocopherol reduces serum concentrations of gamma- and delta-tocopherol in humans. J Nutr, 2003; 133: 3137-3140
• 43)  Vertuani S, Angusti A and Manfredini S: The antioxidants and pro-antioxidants network: an overview. Curr Pharm Des, 2004; 10: 1677-1694
• 44)  Tovmasyan A, Reboucas JS and Benov L: Simple biological systems for assessing the activity of superoxide dismutase mimics. Antioxid Redox Signal, 2014; 20: 2416-2436
• 45)  Block G, Dietrich M, Norkus EP, Morrow JD, Hudes M, Caan B and Packer L: Factors associated with oxidative stress in human populations. Am J Epidemiol, 2002; 156: 274-285
• 46)  Pappas RS: Toxic elements in tobacco and in cigarette smoke: inflammation and sensitization. Metallomics, 2011; 3: 1181-1198
• 47)  Pompella A, Sies H, Wacker R, Brouns F, Grune T, Biesalski HK and Frank J: The use of total antioxidant capacity as surrogate marker for food quality and its effect on health is to be discouraged. Nutrition, 2014; 30: 791-793
• 48)  Aruoma OI: Methodological considerations for characterizing potential antioxidant actions of bioactive components in plant foods. Mutat Res, 2003; 523-524: 9-20
• 49)  Wu X, Beecher GR, Holden JM, Haytowitz DB, Gebhardt SE and Prior RL: Lipophilic and hydrophilic antioxidant capacities of common foods in the United States. J Agric Food Chem, 2004; 52: 4026-4037