Association of dietary inflammatory index, composite dietary antioxidant index, and frailty in elderly adults with diabetes

Abstract

Background

We aimed to examine the relationship of 2 dietary scores [dietary inflammatory index (DII) and composite dietary antioxidant index (CDAI)] with frailty in elderly adults with diabetes.

Methods

Data were gathered from the National Health and Nutrition Examination Survey (NHANES) between 2007 and 2018. The frailty index was calculated using 49 deficits across various systems to define frailty. To examine the relationship of 2 dietary scores (DII and CDAI) with frailty in elderly adults with diabetes, multiple logistic regression analyses were performed. In logistic regression model, DII and CDAI were calculated as both continuous and tertiles. Subgroup analyses were performed to demonstrate stability of results. Restricted cubic splines were utilized to examine the non-linear correlations.

Results

A total of 2,795 elderly adults with diabetes were included in this study. In the multivariate logistic regression model, the odds ratio (OR) of DII for risk of frailty was 1.08 (95% CI 1.02–1.15) and the OR of CDAI for risk of frailty was 0.96 (95% CI 0.93–0.99). The ORs of DII for risk of frailty were 1.36 (95% CI 1.09–1.70) and 1.33 (95% CI 1.04–1.70) for tertiles 2 and 3, respectively (p for trend 0.027). The ORs of CDAI for risk of frailty were 0.94 (95% CI 0.75–1.17) and 0.75 (95% CI 0.58–0.98) for tertiles 2 and 3, respectively (p for trend 0.036). The subgroup analysis demonstrated reliable and enduring connections between 2 dietary scores and frailty (all p for interaction > 0.05). In the restricted cubic spline analyses, we discovered the non-linear relationship between DII and frailty (P for nonlinearity = 0.045) and linear relationship between CDAI and frailty (P for nonlinearity = 0.769).

Conclusion

The research showed connections between 2 dietary scores (DII and CDAI) and frailty as measured by frailty index in elderly adults with diabetes.

Introduction

Frailty, a condition of physiological decline associated with aging, impacts around 10% of elderly adults living in the community [1]. Frailty is defined by reduced capacity for normal function, resulting in increased susceptibility to stress and negative health effects, such as falls, dementia and Alzheimer [2], acute illness, disability, and mortality [3,4,5]. The frailty in elderly populations is a significant concern for public health, leading to increased dependency and greater medical needs among certain groups [6, 7]. Additionally, frailty poses challenges for healthcare systems in terms of service management, provision, and costs [8]. Individuals with diabetes tend to be frailer than elderly adults without diabetes [9]. In a comparable study, Zeng et al. [10] conducted a national cross-sectional survey in China involving 18,010 elderly adults with diabetes, revealing a higher incidence of frailty and pre-frailty in this population compared to elderly adults without diabetes. The possible mechanism is that chronic hyperglycemia may cause insulin resistance, inflammation, and mitochondrial dysfunction in skeletal muscle, leading to muscle weakness, loss of muscle mass, and frailty [11].

Previous research has identified inflammation and oxidative stress as potential contributors to frailty, with systemic inflammation posited as a significant mechanism influencing the development of this condition. [12]. A buildup of inflammation throughout aging can impair physical function [13]. Diet, as a significant potential source of pro-inflammatory and anti-inflammatory compounds, serves as a crucial modulator of inflammation. [14]. The dietary inflammatory index (DII) and composite dietary antioxidant index (CDAI), which are calculated by considering the inflammatory and antioxidant properties of nutrients in the diet, are useful tools for evaluating how diet impacts inflammation and oxidative stress, and have shown to be effective in predicting relevant biomarkers [15,16,17]. To date, a growing body of research has documented an association between higher DII or lower CDAI values and adverse health outcomes, as well as an increased risk of chronic diseases, including frailty [18,19,20], hypertension [21, 22], diabetes [23, 24], atherosclerosis cardiovascular disease [25, 26], mortality [27, 28], and so on. Therefore, it is plausible to hypothesize that DII and CDAI may interact synergistically to influence the frailty in elderly people with diabetes.

Nevertheless, to the best of our knowledge, no one has studied the effects of diets on frailty in elderly people with diabetes. Therefore, we aimed to examine the relationship of 2 dietary scores (DII and CDAI) with frailty in older adults with diabetes using the National Health and Nutrition Examination Survey (NHANES) database.

Methods

Study population

The NHANES conducted by the Center for Disease Control and Prevention was a large-scale, ongoing, and nationally representative health survey of noninstitutionalized American. Data were gathered from six NHANES cycles (2007–2018) and were extracted from January 2024 to February 2024. The participants with diabetes from NHANES (2007–2018) were initial included. The participants with diabetes who had at least one of the following conditions were excluded: (1) aged < 60 years; (2) missing data for diagnosis of frailty; (3) implausible energy intake; (4) missing dietary data; and (5) missing other covariates. Dietary energy intake below 800 or higher than 5000 kcal/day was defined as an implausible energy intake. Following a rigorous selection process, we included 2795 elderly individuals (Fig. 1).

Fig. 1

The flowchart of study participants

Definition of diabetes

According to the American Diabetes Association criteria [29], diabetes was diagnosed when one of the following conditions was met: self-reported diabetes, or fasting plasma glucose ≥ 7.0 mmol/l, or glycated hemoglobin (HbA1c) ≥ 6.5%, or two-hour oral glucose tolerance test (OGTT) blood glucose ≥ 11.1 mmol/l, or rand plasma glucose ≥ 11.1 mmol/l, or using insulin or diabetes medication.

Definition of frailty

We constructed frailty developed using the standard method [30, 31]. In order for characteristics to be incorporated into the index, they must be health impairments that become more prevalent with advancing age, they must be prevalent across multiple systems. A score ranging from 0 to 1 was given based on the seriousness of the shortfall. The frailty index consisted of 49 deficits across various systems. Only deficits with at least 80% recorded are considered valid. The frailty index is calculated by dividing the total number of deficits acquired by the participant by the total number of possible deficits. Supplementary Table 1 contains the variables for the frailty index along with their corresponding scores. A frailty index cut-off score of 0.21 was designated for use in this study [30].

Dietary information

All NHANES participants were eligible for two 24-h dietary recall interviews. The first dietary recall was collected in person at a mobile inspection center with relatively little data missing. The second interview was conducted 3–10 days later by telephone consultation with relatively much data missing. Given the potential significant statistical impact of excessive data deletion, only the initial 24-h dietary recall was utilized for analysis.

Definition of DII

DII is determined based on Hébert's method, which assessed the inflammatory impact of 45 nutrients and was commonly utilized to evaluate dietary inflammation [16]. Twenty-eight out of forty-five dietary components were selected for the calculation of DII. Research has demonstrated that the utilization of these particular foods also maintains the reliability of predictive capabilities [26]. For more in-depth details about the nutritional components used to determine DII, please refer to Supplementary Table 2.

Definition of CDAI

To assess the cumulative exposure from dietary antioxidant consumption, we applied a revised version of the CDAI developed by Maugeri et al. [32]. The total dietary intake of six antioxidants—namely, vitamin A, vitamin C, vitamin E, zinc, selenium, and carotenoids—was quantified by subtracting the mean intake from each individual intake value and subsequently dividing by the standard deviation (SD). The formula was provided below

$${\sum }_{i=1}^{n=6}\frac{\text{Individual} \text{Intake}-\text{Mean}}{\text{SD}}.$$

Covariates

The potential confounders accounted for in the current study included age, sex, race, education, body mass index (BMI), hypertension, cardiovascular disease (CVD), smoking and drinking status, HbA1c levels, and total energy intake. Hypertension was operationally defined as the use of antihypertensive medications, diastolic blood pressure of ≥ 90 mmHg, or systolic blood pressure of ≥ 140 mmHg. CVD was based on self-report of clinical records.

Statistical analysis

All characteristics were reported as mean (SD) for continuous variables and as proportions (percentages) for categorical variables. Categorical variables were analyzed utilizing the Chi-square test, whereas normally distributed continuous variables were assessed using one-way analysis of variance. BMI was categorical, and HbA1c and total energy intake were continuous. The DII and the CDAI were categorized into tertiles. The relationships between 2 dietary scores (DII and CDAI) and frailty were analyzed using logistic regression model. Subgroup analyses were conducted while accounting for confounding variables. Restricted cubic splines were used to investigate the non-linear correlations. Two-sided P < 0.05 was determined statistical significance, and all analyses performed by R (4.3.1).

Results

Table 1 shows the baseline characteristics of participants. The proportion of frailty in elderly diabetic population is 49.8%. The average age of all participants was 69.76 years, and the majority of them were non-Hispanic White (41.43%). Additionally, frail participants were more prone to being older, having a lower education level, smoking, drinking, having hypertension, having longer diabetes duration, having higher BMI, having higher DII scores, and having lower CDAI scores.

Table 1 Baseline characteristics of participants, from NHANES 2007–2018

Table 2 shows the logistic regression analysis on the association between 2 dietary scores (DII and CDAI) and frailty. In the crude model, the odds ratio (OR) of DII was 1.14 [95% confidence interval (CI), 1.09–1.18] and the OR of CDAI was 0.96 [95% CI 0.94–0.98]. Participants in the uppermost DII tertiles compared to those the lowest tertiles had an increased susceptibility to frailty [OR 1.64 (95% CI 1.36–1.97)]. Participants in the uppermost CDAI tertiles compared to those the lowest tertiles were insusceptible to be frail [OR 0.71 (95% CI 0.59–0.85)]. The relationship between 2 dietary scores (DII and CDAI) and frailty remained consistent across various models and the trend was robust. In the model 2, after adjusting for all variables, the OR of DII was 1.08 (95% CI 1.02–1.15) and the OR of CDAI was 0.96 (95% CI 0.93–0.99). The ORs of DII were 1.36 (95% CI 1.09–1.70) and 1.33 (95% CI 1.04–1.70) for tertiles 2 and 3, respectively (p for trend 0.027). The ORs of CDAI were 0.94 (95% CI 0.75–1.17) and 0.75 (95% CI 0.58–0.98) for tertiles 2 and 3, respectively (p for trend 0.036).

Table 2 Logistic regression analysis on the association between DII and CDAI on frailty in elderly adults with diabetes

Table 3 presents the results of the subgroup analysis. The findings from this analysis indicate a consistent association between the two dietary scores (DII and CDAI) and frailty across various subgroups. Importantly, no significant interactions were detected for sex, race, smoking status, drinking status, hypertension, education, CVD, duration of diabetes, and BMI, suggesting that the observed association is independent of these variables (all p values for interaction > 0.05).

Table 3 Subgroup analyses of the association between DII and CDAI on frailty in elderly adults with diabetes stratified by sex, race, smoking status, education, drinking status, hypertension, CVD, diabetes duration, and BMI

Figure 2 shows the restricted cubic spline of OR and 95% CI for the association between 2 dietary scores (DII and CDAI) and frailty. We discovered the non-linear relationship between DII and frailty (P for nonlinearity = 0.045) and linear relationship between CDAI and frailty (P for nonlinearity = 0.769).

Fig. 2

Restricted cubic spline of OR and 95% CI for the association between frailty and (A) DII and (B) CDAI

Discussion

In this study, we investigated the connections of 2 dietary scores (DII and CDAI) with frailty measured by frailty index. In the restricted cubic spline analysis, we discovered the non-linear relationship between DII and frailty and linear relationship between CDAI and frailty. We demonstrated that higher DII or lower CDAI was associated with higher frailty risk. The findings from subgroup analysis demonstrated reliable and enduring connections between 2 dietary scores (DII and CDAI) and frailty across various subgroups.

Various frailty models have been created using various conceptual definitions. The 2 main models are the Fried’s phenotype model [1] and the frailty index model. The phenotype model defines frailty as meeting at least 3 out of 5 physical criteria [1]. Rockwood et al. created frailty index model conducted by evaluating a wide range of factors, such as chronic illnesses, psychosocial elements, cognitive impairments, and other signs of aging, a thorough geriatric assessment to tally up all deficits [33]. Frailty is detailed as a predictor of negative health results in a cumulative manner. The index is shown by comparing the total deficits acquired to all possible deficits in the index. While both tools have their own merits, the frailty index has demonstrated superior predictive value for mortality and other adverse outcomes [34, 35]. Therefore, we used frailty index included 49 deficits for this study.

Although frailty has been receiving more attention, the pathophysiological factors that play a role in frailty remain poorly understood [18]. Studies to date have implicated systemic inflammation as a key mechanism to developing frailty [36]. As aging, the buildup of reactive oxygen species can cause inflammation and hinder physical abilities [13]. Ideally, molecules possessing antioxidant properties can counteract reactive oxygen species [13]. Nevertheless, this procedure could be overwhelmed in the elderly due to the fact that aging and illness frequently coincide with increased inflammation levels. As a result, the buildup of inflammatory cytokines may contribute to muscle breakdown, potentially resulting in decreased physical movement [37]. Elevated IL-6 levels and C-reactive protein (CRP) [38, 39] have been associated with reductions in muscle mass and motor function, which is a characteristic of frailty [40]. Interestingly, IL-6 and CRP are highly modifiable by diet.

In this study, we verified the previous hypothesis that two different measures of dietary inflammation and anti-inflammatory were associated with frail indicators in elderly diabetic populations. The DII was created to assist in determining how much an individual diet can impact systemic inflammation level. The non-linear relationship of DII indicated that the influence of DII on fatigue had a threshold effect, and the influence on fatigue was not increase after a certain value. Frailty is caused by a variety of factors. Inflammation is only one aspect, and an anti-inflammatory diet can only eliminate the impact of inflammation on frailty. Therefore, frailty will not be reduced to disappear with the increase of anti-inflammatory substances and there will be a threshold effect. CDAI is calculated from 6 types of anti-inflammatory foods. Compared to the DII calculated from 28 types of inflammatory and anti-inflammatory foods, CDAI only represents a partial anti-inflammatory effect and does not provide a comprehensive assessment of anti-inflammatory properties. Therefore, CDAI may not reach the threshold necessary to assess the influence of inflammation on frailty, suggesting that the relationship between CDAI and frailty is linear. Previous studies have linked DII with frailty in the elderly [18]. However, they used phenotypic frailty to evaluate frailty rather than frailty index. The CDAI, as formulated by Wright et al. [41], serves as a comprehensive measure encompassing various dietary antioxidants, thereby reflecting an individual overall dietary antioxidant intake profile. Unlike DII, the CDAI contains only six anti-inflammatory elements, which makes the CDAI easier to calculate. However, the disadvantage is that the results may vary considerably depending on the sample size, and may be more suitable for studies with large samples.

Dietary changes, a straightforward and effective approach, have been empirically shown to offer numerous health advantages [42]. Proper nutrition is crucial for managing inflammation and oxidative stress [43]. A balanced diet can help decrease inflammation and combat oxidative stress by incorporating anti-inflammatory and antioxidant compounds, maintaining optimal fatty acid levels, increasing dietary fiber intake, and consuming a variety of fruits, vegetables, and high-quality proteins [44, 45]. Unhealthy eating habits, like consuming foods rich in oil and salt, sugary drinks, too much alcohol, excessive red meat consumption, and cooking at high temperatures, can increase inflammation and oxidative stress levels, which are linked to disease and mortality [46,47,48].

This research analyzed the extensive NHANES data that are publicly accessible and based on a large population, with a strict and well-managed protocol. Nutritional information was consistently documented and computed. Various potential confounding factors were taken into account in all analyses. However, it is important to note that there are certain restrictions to consider when analyzing these findings. NHANES utilizes a cross-sectional design, making it difficult to directly assess the causal relationship between frailty and nutrition due to the lack of ability to evaluate temporal relationships in the data. Dietary information was collected through 24-h recall, which may not accurately reflect participants’ long-term eating habits and nutrient intake, particularly micronutrients, due to potential changes in food intake during the study period and the inability to measure day-to-day variation. Because dietary data are based on participants’ recall, cognitively impaired people may not be included in the study, because they cannot recall what they ate. However, people who are cognitively impaired are likely to be frail, which could lead to bias in the study. Additionally, the dietary scores were calculated using self-reported dietary information, without taking into account supplement consumption. The absence of waist circumference, an anthropometric measurement closely related to inflammation, also be a limitation of the study. Additionally, there is a possibility of reverse causation, where changes in dietary intake and nutritional status occur when individuals begin to experience symptoms of illness. Due to the demographic makeup of our sample, which is mainly made up of people from the United States, it is uncertain if the findings of our study can be generalized to larger populations. Additional randomized-controlled trials in the future are needed to verify the connection between dietary score and frailty.

Conclusions

The research showed connections between 2 dietary scores (DII and CDAI) and frailty as measured by frailty index in elderly adults with diabetes. It emphasized the significance of maintaining proper nutrition in older age and including measures of nutritional health in regular evaluations and surveys of elderly individuals. In future longitudinal research on the connection between diet and frailty, it is recommended to utilize the frailty index due to its ability to capture a wide range of geriatric conditions and its responsiveness to changes in clinical and laboratory evaluations.