Background The relationship between diet and cognitive function is likely multifaceted and may involve bidirectional influences. However, most studies on this topic have been cross-sectional and are therefore limited in their ability to establish temporal relationships. Therefore, this study aimed to employ cross-lagged panel network analysis to clarify the bidirectional temporal dynamics between culturally specific dietary patterns and cognitive function in a longitudinal cohort of Chinese older adults. Methods A total of 2440 participants were recruited from the Chinese Longitudinal Healthy Longevity Survey spanning from 2008 to 2018. Chinese-specific dietary patterns were categorized as milk–egg–sugar pattern, carnivorous pattern, healthy pattern and northeastern pattern. Cognitive function was divided into five domains including orientation, registration, attention and calculation, recall, and language by using the validated Chinese version of the Mini Mental State Examination. Four contemporaneous networks and three cross-lagged panel networks were estimated based on dietary patterns and cognitive function, with age, sex, educational level, and household income included as covariates in all analyses to control for potential confounding effects. All data analyses were conducted in R version 4.3.1. Results The contemporaneous networks showed that the strongest bridge edge between Chinese-specific dietary patterns and cognitive function was “CGF3: attention and calculation”—“PTN2: carnivorous pattern” in the T1 (r = 0.082), T2 (r = 0.085), and T3 (r = 0.077) networks and “CGF3: attention and calculation”—“PTN3: healthy pattern” in the T4 (r = 0.123) network. Cross-lagged panel networks revealed that the strongest bridge edges connecting the Chinese-specific dietary patterns and cognitive function were “PTN3: healthy pattern” → “CGF3: attention and calculation” (bridge weight = 0.101) in the T1 → T2 network, “PTN2: carnivorous pattern” → “CGF3: attention and calculation” (bridge weight = 0.113) in the T2 → T3 network, and “CGF4: recall” → “PTN1: milk–egg–sugar pattern” (bridge weight = − 0.096) in the T3 → T4 network. Conclusion Attention and calculation, the central cognitive domain identified, showed strong bidirectional links with diet over time. These findings support life stage-specific dietary interventions to maintain cognitive health.

Chinese-specific dietary patterns and cognitive function in Chinese older adults: a cross-lagged panel network analysis

期刊论文
发表日期：

2026年3月31日

查看全文

Zhizhuo Wang,
Shiyi Wang

摘要

Background

The relationship between diet and cognitive function is likely multifaceted and may involve bidirectional influences. However, most studies on this topic have been cross-sectional and are therefore limited in their ability to establish temporal relationships. Therefore, this study aimed to employ cross-lagged panel network analysis to clarify the bidirectional temporal dynamics between culturally specific dietary patterns and cognitive function in a longitudinal cohort of Chinese older adults.

Methods

A total of 2440 participants were recruited from the Chinese Longitudinal Healthy Longevity Survey spanning from 2008 to 2018. Chinese-specific dietary patterns were categorized as milk–egg–sugar pattern, carnivorous pattern, healthy pattern and northeastern pattern. Cognitive function was divided into five domains including orientation, registration, attention and calculation, recall, and language by using the validated Chinese version of the Mini Mental State Examination. Four contemporaneous networks and three cross-lagged panel networks were estimated based on dietary patterns and cognitive function, with age, sex, educational level, and household income included as covariates in all analyses to control for potential confounding effects. All data analyses were conducted in R version 4.3.1.

Results

The contemporaneous networks showed that the strongest bridge edge between Chinese-specific dietary patterns and cognitive function was “CGF3: attention and calculation”—“PTN2: carnivorous pattern” in the T1 (r = 0.082), T2 (r = 0.085), and T3 (r = 0.077) networks and “CGF3: attention and calculation”—“PTN3: healthy pattern” in the T4 (r = 0.123) network. Cross-lagged panel networks revealed that the strongest bridge edges connecting the Chinese-specific dietary patterns and cognitive function were “PTN3: healthy pattern” → “CGF3: attention and calculation” (bridge weight = 0.101) in the T1 → T2 network, “PTN2: carnivorous pattern” → “CGF3: attention and calculation” (bridge weight = 0.113) in the T2 → T3 network, and “CGF4: recall” → “PTN1: milk–egg–sugar pattern” (bridge weight = − 0.096) in the T3 → T4 network.

Conclusion

Attention and calculation, the central cognitive domain identified, showed strong bidirectional links with diet over time. These findings support life stage-specific dietary interventions to maintain cognitive health.