Brain responses to intermittent fasting and the healthy living diet in older adults

Abstract

Diet may promote brain health in metabolically impaired older individuals. In an 8-week randomized clinical trial involving 40 cognitively intact older adults with insulin resistance, we examined the effects of 5:2 intermittent fasting and the healthy living diet on brain health. Although intermittent fasting induced greater weight loss, the two diets had comparable effects in improving insulin signaling biomarkers in neuron-derived extracellular vesicles, decreasing the brain-age-gap estimate (reflecting the pace of biological aging of the brain) on magnetic resonance imaging, reducing brain glucose on magnetic resonance spectroscopy, and improving blood biomarkers of carbohydrate and lipid metabolism, with minimal changes in cerebrospinal fluid biomarkers for Alzheimer's disease. Intermittent fasting and healthy living improved executive function and memory, with intermittent fasting benefiting more certain cognitive measures. In exploratory analyses, sex, body mass index, and apolipoprotein E and SLC16A7 genotypes modulated diet effects. The study provides a blueprint for assessing brain effects of dietary interventions and motivates further research on intermittent fasting and continuous diets for brain health optimization. For further information, please see ClinicalTrials.gov registration: NCT02460783.

Keywords: Alzheimer’s disease; biomarkers; brain aging; insulin resistance; intermittent fasting; randomized clinical trial.

Figure 1.. A, Study design. B, Flow diagram.

Figure 2.. Brain health measures.

A, NDEV-associated pS312-IRS-1 and pY-IRS-1 B, Significant BrainAGE change within an anterior cingulate/ventromedial prefrontal cortex cluster. C, Glucose concentration in posteromedial cortex (PMC) by ¹H J-PRESS MRS. D, CSF biomarkers for AD/ADRD. Across all figures, data points represent model-predicted values, boxes demonstrate variance, horizontal lines depict estimated marginal means and whiskers depict standard errors. Brackets indicate P values for F tests, as applicable: F_time (effects of time across groups), F_IF*time (effects of time in the IF group), F_HL*time (effects of time in the HL group). * P < 0.05, ** P < 0.01, *** P < 0.001, # P = 0.05 – 0.10, ns = not significant.

Figure 3.. Cognition and physical activity.

A, Executive Function Composite, Fluency Factor, Dimensional Set Shifting by NIH EXAMINER. B, Sex modulation of diet effects on the Executive Function Composite, Fluency Factor, Dimensional Set Shifting. C, CVLT domains of short and long delay cued recall. D, Total number of sedentary bouts on actigraphy. Across all figures, data points represent model-predicted values, boxes demonstrate variance, horizontal lines depict estimated marginal means and whiskers depict standard errors. Brackets indicate P values for F tests, as applicable: F_time (effects of time across groups), F_IF*time (effects of time in the IF group), F_HL*time (effects of time in the HL group), F_time*sex (effects of time across groups by sex), F_IF*time*sex (effects of time in the IF group by sex), F_HL*time*sex (effects of time in the HL group by sex). * P < 0.05, ** P < 0.01, *** P < 0.001, # P = 0.05 – 0.10, ns = not significant.

Figure 4.. Systemic and peripheral metabolism measures.

A, Anthropometrics: body weight, BMI, and waist circumference. B, Systemic IR and glycemia: fasting insulin, HOMA2-IR, and HbA1c. C, Blood ketones and lipids: β-Hb, AcAc, total and LDL cholesterol. Across all figures, data points represent model-predicted values, boxes demonstrate variance, horizontal lines depict estimated marginal means and whiskers depict standard errors. Brackets indicate P values for F tests, as applicable: F_time (effects of time across groups), F_IF*time (effects of time in the IF group), F_HL*time (effects of time in the HL group). * P < 0.05, ** P < 0.01, *** P < 0.001, # P = 0.05 – 0.10, ns = not significant.