Association between oxidative balance score and inflammatory markers in middle‐aged and older Japanese people

This study aimed to investigate the association between oxidative balance score (OBS), wherein higher OBSs indicate lower oxidative stress, and high‐sensitivity C‐reactive protein (hs‐CRP), as well as inflammatory scores, in a large cohort of Japanese adults.


| INTRODUCTION
Chronic inflammation plays a central role in the development of cardiovascular diseases (CVDs), the primary cause of death worldwide (Lazou et al., 2020).C-reactive protein (CRP), a plasma protein produced after stimulation by inflammatory cytokines, is used as a biomarker for inflammatory reaction because it reflects the systemic response to inflammation (Pepys & Hirschfield, 2003).The plasma concentration of high-sensitivity CRP (hs-CRP) is a marker of chronic and low-grade inflammation linked to elevated risk of CVDs (Ridker, 2003).Additionally, it has been reported that the variation in hs-CRP is influenced by various factors, such as age and lifestyle (Inoue et al., 2016;McDade et al., 2011).
Oxidative stress, defined as an imbalance between the production and accumulation of reactive oxygen species (ROS) in cells and tissues and the ability of a biological system to detoxify these reactive products, causes damage to important biomolecules and cells (Reuter et al., 2010).Evidence suggests that sustained oxidative stress can cause cellular injury, triggering an inflammatory response that further exacerbates oxidative stress and creates a detrimental cell cycle (Mittal et al., 2014;Pizzino et al., 2017).A substantial body of evidence from population-level studies has extensively supported the impact of individual lifestyle factors on oxidative stress (Eastwood, 1999;Wright et al., 2004).
Some studies have reported that a higher dietary intake of antioxidant-rich (e.g., carotene, vitamin C, and vitamin E) foods and beverages may offer protection against oxidative stress.Conversely, prooxidant factors, including smoking and higher dietary iron intake, enhance the production of reactive oxygen and nitrogen species, thereby accelerating the cellular damage associated with oxidative stress.More recently, complex combinations of lifestyle factors, including diet, have been reported to be more strongly associated with disease risk than when individual factors are considered (Goodman et al., 2007;Van Hoydonck et al., 2002).
To address this issue, the oxidative balance score (OBS) was developed to encapsulate the cumulative impact of multiple lifestyle factors on oxidative stress, with a higher OBS indicating a greater predominance of antioxidants over prooxidant exposure (Goodman et al., 2007;Van Hoydonck et al., 2002).Epidemiological evidence regarding the association between OBS and serum hs-CRP levels has been reported previously (Kong et al., 2014;Lakkur et al., 2015;Lee & Park, 2017).A study of US patients with sporadic colorectal adenomas revealed an inverse association between OBS and serum hs-CRP levels (Kong et al., 2014).US participants from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study found a negative association with high CRP levels (Lakkur et al., 2015).Additionally, a study based on the Korea Association Resource, which included 6414 Korean participants, demonstrated that the group with the highest OBS exhibited lower CRP levels than the lowest group (Lee & Park, 2017).
However, several issues remain unresolved.First, evidence on the association between OBS and hs-CRP levels is limited (Kong et al., 2014;Lakkur et al., 2015).Only one study was conducted in Asia (Lee & Park, 2017), with an inadequately stratified analysis by sex.Given that several significant differences in hs-CRP levels by sex have been reported in large cohort studies (Khera et al., 2005;Lakoski et al., 2006), it is necessary to clarify the relationship between oxidative stress scores and hs-CRP in Asian populations, considering sex differences.Given the distinct inflammatory profile of Japanese adults compared to Western populations and the potential influence of cultural factors on inflammatory responses (Coe et al., 2011(Coe et al., , 2020)), it is crucial to explore these associations within the Japanese population.Second, absorption of iron varies across different food sources, and nonheme iron, particularly that found in plant foods and fish, may exert a lower impact on oxidative stress due to its lower absorption rate (Galaris et al., 2008).Therefore, considering body iron stores as a component of the OBS may be more appropriate than considering dietary iron stores alone.However, only a few studies have considered body iron stores as an element of OBS.Third, given that integrating multiple inflammatory biomarkers can offer a more comprehensive evaluation of inflammatory status (Toriola et al., 2013), inflammatory scores may be more robust than individual biomarkers in the assessment of inflammation.However, to our knowledge, no studies have examined the association between OBS and inflammation scores.
Hence, this study aimed to investigate the association between OBS and hs-CRP levels stratified by sex in a substantial Japanese adult cohort.Additionally, the association between OBS and inflammatory z-score was explored using data from 2200 patients, with measurements of serum cytokines being available.

| Study participants
The participants included in the present study were derived from the Japan Multi-Institutional Collaborative Cohort Study conducted in the Saga region (Hara et al., 2010(Hara et al., , 2015)).This cohort consisted of 12 068 middle-aged Japanese individuals aged 40-69 years who participated in a baseline survey conducted between 2005 and 2007 (Hara et al., 2016).On-site assessments were conducted using a self-report questionnaire to gather information on lifestyle factors, anthropometric measurements, and venous blood samples.The research protocol was approved by the ethics committees of the Saga University Graduate School of Medicine (approval no.17-11), Nagoya University Graduate School of Medicine (approval no.253), and the National Institute of Biomedical Innovation, Health and Nutrition (NIBIOHN-135-1).The purpose, content, and conditions of the study were explained in writing and orally; written informed consent was obtained from the participants.
For the analysis investigating the association between OBS and hs-CRP levels, participants were selected based on the following criteria: from the initial cohort of 12 068 individuals, 902 were excluded for the following reasons: (i) missing data on hs-CRP levels or exceeding 3000 ng/ mL (n = 802); (ii) missing data on individual components of OBS (n = 157); (iii) presence of any history of the inflammation-related disease, such as CVD, cancer, liver disease, or chronic renal failure (n = 1399); (iv) extremely low or high dietary energy intake (n = 4), defined as dietary energy intake <800 or ≥4500 kcal/day for men, and <500 or ≥3500 kcal/day for women.The final analysis comprised 9703 eligible participants, consisting of 3902 men and 5801 women (all subjects).Serum levels of inflammatory cytokines were measured using samples collected from a subset of baseline participants.The association between OBS and the inflammatory z-score was examined specifically within a subgroup of 1652 individuals (subgroup subjects).There were no significant differences in subject characteristics between data for all subjects and data for subgroup subjects in our analysis (Table S1).

| Dietary survey
A validated self-administered food frequency questionnaire (FFQ) (Goto et al., 2006;Imaeda et al., 2021;Tokudome et al., 2004Tokudome et al., , 2005) ) was employed to collect dietary information.In this study, a baseline survey was conducted to evaluate the dietary intake of key nutrients implicated in oxidative stress, specifically carotene, vitamin E, vitamin C, saturated fatty acids (SFA), n-3 polyunsaturated fatty acids (PUFA), and n-6 PUFA.The FFQ required participants to report their frequency of consumption of 47 food and beverage items over the past year to determine average intake.The program was developed at the Department of Public Health, Nagoya City University School of Medicine (Goto et al., 2006;Imaeda et al., 2021;Tokudome et al., 2004Tokudome et al., , 2005)), utilizing the standard tables of food consumption in Japan (5th revised ed.) (Science and Technology Agency of Japan, 2000) to calculate daily nutritional intake.To assess alcohol consumption, the questionnaire addressed the frequency and quantity of six different alcoholic beverages.Ethanol consumption per day among current drinkers was estimated based on age-specific ethanol concentrations.Fatty acid and vitamin intake from supplements were not considered in the analyses.All nutritional covariates, excluding alcohol consumption, were adjusted for total energy intake using the residual method.

| Data collection
Height and weight measurements were obtained with a precision of 0.1 cm and 0.1 kg, respectively, and body mass index (BMI) was subsequently calculated.A selfadministered questionnaire that included questions on smoking, dietary habits, current medications, and disease history was sent to the participants beforehand (Hara et al., 2016).Regarding smoking habits, the participants were first asked about their current smoking status (and cessation time for former smokers).Current and former smokers reported their usual cigarette consumption (cigarettes per day) and the age at which they started smoking.Physical activity level (PAL) was assessed using a singleaxis accelerometer (Kenz Lifecorder EX; Suzuken Co., Ltd., Nagoya, Japan) on either side of the hip, except when sleeping or bathing, for 10 days after the baseline survey.
The PAL was calculated by dividing the total energy expenditure (kcal/day) by the basal metabolic rate (kcal/ day).The former was estimated from the accelerometer as the average daily (excluding the first 3 days) energy expenditure, and the latter was defined as the basal metabolic standard (Ministry of Health and Welfare, 1994) Â body surface area (Fujimoto et al., 1968) Â 24 h.

| Calculation of oxidative balance score
OBS was derived by amalgamating data from 11 predetermined prooxidant and antioxidant factors, including carotene, vitamin E, vitamin C, serum ferritin, n-3 and n-6 PUFA, SFA, PAL, alcohol consumption, smoking status, and regular use of nonsteroidal anti-inflammatory drugs (NSAIDs) (Table 1).Continuous variables representing prooxidant (serum ferritin, n-3 and n-6 PUFA, and SFA) and antioxidant (carotene, vitamin E, vitamin C, and PAL) factors were categorized into low, medium, and high categories based on the tertile values for each exposure.For antioxidants, scores of 0-2 were assigned to the first through third tertiles, whereas the reverse assignment was made for prooxidants (0 points for the highest tertile and 2 points for the lowest tertile).A similar scoring methodology was applied to categorical prooxidant and antioxidant variables.Regarding alcohol consumption, nondrinkers, moderate drinkers (>0-<23 g of ethanol/day), and heavy drinkers (≥23 g of ethanol/ day) were allocated 2, 1, and 0 points, respectively.The participants were categorized based on their smoking status as never smokers (2 points), former smokers (1 point), and current smokers (0 points).Participants who did not regularly use NSAIDs received 0 points, while those who regularly used NSAIDs were assigned two points.OBS score was computed by summing the assigned points for each participant, with a higher score indicating a predominance of antioxidant over prooxidant exposure.

| Statistical analysis
Separate analyses were conducted for men and women owing to variations in participant characteristics between the sexes.Prior to analysis, serum ferritin and inflammatory markers were subjected to natural log transformation to achieve an approximately normal distribution.Linear regression analysis was used to evaluate the selected characteristics of participants across OBS quartiles for continuous variables, whereas the Mantel-Haenszel test was used for categorical variables.Adjusted geometric means of serum hs-CRP levels and their 95% confidence intervals (CIs) were computed using a general linear model according to OBS quartiles.Linear trends were tested by assigning ordinal numbers (0-3) to the OBS quartile categories.We adjusted for age (years, continuous), total energy intake (kcal, continuous), BMI (kg/m 2 , continuous), hypertension (category), diabetes mellitus (category), dyslipidemia (category), and menopausal status (women only, category).Iron storage and smoking status are potent prooxidants and significant risk factors for inflammation; additional analyses were conducted, excluding serum ferritin or smoking status from the calculation of OBS but accounting for them in the model.An inflammatory cytokine z-score was constructed, computed as the sum of z-scores for hs-CRP, IL-6, IL-8, IL-15, and TNF, and analyzed in a manner similar to that for individual inflammatory cytokines.Furthermore, the association between OBS and the inflammation z-score, as well as each marker, was assessed using the general linear model.All statistical analyses were performed using the SAS statistical software (Ver.9.4; SAS Institute, Cary, NC, USA).

| RESULTS
Table 2 presents the characteristics of the study participants based on OBS quartiles.Participants of both sexes with higher OBSs had lower BMI and total energy intake.Among men, those with a higher OBS tended to be older.Among women, higher OBSs were associated with a lower prevalence of hypertension and postmenopausal status.
Table 3 shows the adjusted geometric means and 95% CIs for serum hs-CRP levels according to OBS quartiles in men and women.A significant association was observed between the total OBS and serum hs-CRP levels in both sexes, even after accounting for all covariates (p trend <.01).Similar results were obtained when both sexes excluded serum ferritin levels or smoking status from the OBS analysis.

| DISCUSSION
In this large population-based cross-sectional study, we found a negative association between OBS, which comprises dietary and lifestyle prooxidant and antioxidant components, and hs-CRP concentrations in both sexes after adjustment for multiple confounders.These results remained unaltered when the powerful prooxidants, serum ferritin, and smoking were excluded from the OBS analysis.OBS was also associated with a lower inflammatory z-score, indicating reduced overall systemic inflammation, particularly for hs-CRP.To the best of our knowledge, this is the first study to report a sex-specific association between OBS and hs-CPR levels.
A cross-sectional study conducted in the USA showed an inverse association between OBS and hs-CRP concentrations (OR high vs. low OBS = 0.50; 95% CI:0.38-0.66); the OBS analyzed in this study by Lakkur et al. (2015) comprised 13 components including measurements of serum ferritin and antioxidant nutrients, as well as lifestyle factors including physical activity, smoking and alcohol consumption, and medication use.According to a study by Lee and Park (2017), elevated OBSs were also associated with lower CRP levels (β = À0.28 for high vs. low OBS quartiles; pvalue = <.01);however, the OBS components were limited to only seven factors.Similar findings were reported by Kong et al. (2014), who reported a negative association between OBS based on a questionnaire and nutrient biomarkers and lower CRP levels (OR high vs. low OBS = 0.21; 95% CI:0.09-0.49); the OBS analysis included several biomarker nutrient components, except α-tocopherol, and the same dietary (vitamin C and PUFA intake), lifestyle (smoking, alcohol, Se supplemental use), and medication components (NSAID and aspirin) as in the present study.Our findings align with previous studies, suggesting a robust association between OBS and inflammatory markers regardless of race, cultural background, and sex.
Increased ROS generation attributable to increased oxidative stress induces the oxidation of biomolecules, enzymatic protein alterations, genetic modifications, and intricate signaling pathways, culminating in the onset of inflammatory disorders.Activating transcription factors and subsequent induction of inflammatory genes by ROS initiates an inflammatory response (Mittal et al., 2014).The OBS used in this study included dietary and lifestyle factors along with antioxidant components.Serum ferritin, which induces oxidative stress, showed the strongest association with oxidative stress markers, and its value was stronger than when OBS was generated using dietary iron (data not shown).When serum ferritin level was excluded from the OBS analysis, the trend between OBS and hs-CRP levels remained statistically significant.This suggests that creating an OBS with multiple factors may be useful for assessing the roles of oxidative stressrelated lifestyle factors.
This study had some limitations.First, it used a crosssectional design, and reverse causation may have accounted for the observed associations.Second, we used self-reported intake to assess prooxidant and antioxidant exposure.It has long been acknowledged that dietary questionnaires may not capture all the possible sources of each nutrient, do not account for bioavailability, and are subject to recall bias.However, the validity and reliability of the FFQ used in our study have been extensively evaluated, and any misclassification is expected to be nondifferential.In addition, there were no data on specific cancers in this study.Third, the OBS in our study was limited to dietary/lifestyle exposure and did not include any endogenous factors that influence cellular antioxidant defense, DNA damage and repair, cell growth, or cell death, all of which contribute to the survival of individuals.
In conclusion, the results of this study suggest that a higher OBS, reflecting a greater balance between antioxidant and prooxidant lifestyle exposures, is associated with lower hs-CRP levels and reduced systemic inflammation, regardless of sex.These findings confirm the results of previous studies and suggest that OBS might be useful for evaluating the role of oxidative stress-related lifestyle factors, including diet, as determinants of inflammation.
T A B L E 4 Multiple regression analysis of the total oxidative balance score and inflammatory markers.Men (n = 668) aWomen (n = 984) b Characteristics of the study participants by oxidative balance score quartile.
T A B L E 2Note: p-values for linear trends across quartiles (assigned ordinal numbers 0-3) of the oxidative balance score were based on linear regression analysis for continuous variables and the Mantel test for categorical variables.Values are presented as mean (standard deviation) for continuous variables and number (percentage) for categorical variables.
T A B L E 3 Adjusted geometric means (95% CI) of inflammatory biomarkers according to oxidative balance score quartiles by sex.