Carbohydrate intake and fasting blood glucose in relation to frailty among Chinese adults: two separate cross-sectional analyses from CHNS and CHARLS

Technical Summary: Carbohydrate Intake, Fasting Blood Glucose, and Frailty in Chinese Adults

Problem Statement
Frailty is a multidimensional geriatric condition characterized by reduced physiological reserve and increased vulnerability to adverse health outcomes. While nutrition and glycometabolic status are theoretically linked to frailty, comprehensive evidence is often limited by the unavailability of detailed dietary data and fasting glucose measurements within the same population dataset. Furthermore, existing studies on carbohydrate intake and frailty have yielded inconsistent results, and evidence specific to Chinese populations using a deficit-accumulation frailty index remains scarce. This study addresses the gap by examining two separate cross-sectional associations in large Chinese population-based cohorts: the relationship between carbohydrate intake and frailty, and the relationship between fasting blood glucose and frailty.

Methodology
The study utilized two distinct, publicly available datasets, analyzing them as independent cross-sectional studies due to the lack of overlapping exposure and outcome variables within a single cohort.

• Data Sources:
  • CHNS (China Health and Nutrition Survey): Pooled data from 2004, 2006, 2009, and 2011 waves ($N=5,904$) were used to analyze carbohydrate intake. Dietary intake was assessed via three consecutive 24-hour recalls.
  • CHARLS (China Health and Retirement Longitudinal Study): Baseline data from 2011 ($N=8,687$) were used to analyze fasting blood glucose. Venous blood samples were collected after an overnight fast.
• Frailty Assessment: Frailty was operationalized using a Frailty Index (FI) based on the deficit-accumulation approach.
  • CHNS utilized 32 health deficit indicators; CHARLS utilized 29.
  • Participants were classified as non-frail ($FI \le 0.10$), pre-frail ($0.10 < FI \le 0.21$), or frail ($FI > 0.21$).
  • The primary outcome was a binary classification: non-frailty versus pre-frailty/frailty combined.
• Statistical Analysis:
  • Primary Analysis: Multivariable logistic regression was employed to estimate Odds Ratios (ORs).
    • CHNS: ORs reported per 50 g/day increase in carbohydrate intake.
    • CHARLS: ORs reported per 10 mg/dL increase in fasting blood glucose.
    • Models were adjusted for demographics, socioeconomic status, lifestyle factors, and health-related covariates (e.g., fat/protein intake in CHNS; lipids/uric acid in CHARLS).
  • Dose-Response: Restricted cubic spline models were used to assess linearity and nonlinearity.
  • Supplementary Analyses: Subgroup analyses were conducted to test for effect modification. LASSO regression was used as an exploratory variable-screening method. Exploratory nomograms were generated to visualize fitted associations.

Key Results

• Carbohydrate Intake (CHNS): Higher carbohydrate intake was significantly associated with lower odds of frailty. In the fully adjusted model, each 50 g/day increase in carbohydrate intake corresponded to an OR of 0.872 (95% CI: 0.827–0.920, $P < 0.001$). Restricted cubic spline analysis confirmed a significant inverse association without evidence of nonlinearity ($P_{\text{nonlinearity}} = 0.158$).
• Fasting Blood Glucose (CHARLS): Higher fasting blood glucose was significantly associated with higher odds of frailty. In the fully adjusted model, each 10 mg/dL increase in fasting blood glucose corresponded to an OR of 1.033 (95% CI: 1.019–1.048, $P < 0.001$). Spline analysis showed a generally positive, monotonic association without significant nonlinearity ($P_{\text{nonlinearity}} = 0.068$).
• Subgroups: The directions of association remained generally consistent across most demographic, lifestyle, and clinical subgroups in both cohorts. Nominal evidence of interaction was observed only for hypertension status in CHNS and triglyceride levels in CHARLS.

Key Contributions

1. Dual-Cohort Evidence: The study provides parallel evidence from two major Chinese cohorts, linking lower carbohydrate intake (via CHNS) and higher fasting glucose (via CHARLS) to frailty status, despite the exposures being measured in separate datasets.
2. Deficit-Accumulation Approach: By utilizing a continuous Frailty Index rather than a physical frailty phenotype alone, the study captures a broader spectrum of health deficits relevant to the Chinese aging population.
3. Methodological Rigor: The study employs advanced statistical techniques, including restricted cubic splines for dose-response modeling and LASSO for exploratory variable selection, while explicitly acknowledging the limitations of cross-sectional design and potential construct overlap between covariates and the frailty index.
4. Contextual Nuance: The analysis distinguishes between total energy intake and specific macronutrients by adjusting for fat and protein intake in the CHNS model, suggesting the association is not merely a proxy for total caloric sufficiency.

Significance and Claims
The authors emphasize that these findings represent two independent cross-sectional associations rather than evidence of a single causal pathway linking carbohydrate intake, glucose metabolism, and frailty.

• Interpretation of Carbohydrate Intake: The inverse association in CHNS is not claimed to prove that higher carbohydrate intake causes reduced frailty. Instead, it is interpreted as reflecting broader dietary adequacy, habitual food patterns, and appetite in older adults. The authors note that lower intake may signal underlying health issues or socioeconomic constraints.
• Interpretation of Glucose: The positive association in CHARLS positions fasting blood glucose as a metabolic marker associated with frailty, rather than a direct cause. The cross-sectional nature precludes determining whether hyperglycemia precedes frailty or if frailty-related physiological changes (e.g., reduced muscle mass, inactivity) drive glucose dysregulation.
• Public Health Implication: The study suggests that frailty assessment should consider both dietary adequacy (including carbohydrate sources and quality) and glycometabolic status. However, the authors explicitly state that their results do not establish that modifying carbohydrate intake or fasting glucose alone would reduce frailty.
• Future Directions: The paper concludes that longitudinal studies with repeated measures of diet, metabolism, and frailty in the same individuals are necessary to clarify temporal relationships and causal mechanisms.

Limitations Acknowledged
The study explicitly notes limitations including the cross-sectional design (precluding causal inference), the separation of exposure and outcome data across different cohorts, potential recall bias in self-reported dietary data, the lack of distinction between carbohydrate types (e.g., refined vs. whole grains), and the combination of pre-frailty and frailty into a single outcome category.