Skip to main content
NCBI home page
Alzheimer's Research & Therapy logoLink to Alzheimer's Research & Therapy
. 2026 May 20;18:160. doi: 10.1186/s13195-026-02074-9

Metabolic parameter variability, brain structure, perfusion, and cognition: a population-based study

Xiaoshuai Li 1,#, Xuejia Li 2,#, Mengmeng Bai 3, Ling Yang 1, Jing Chen 4, Jiacheng Fan 5, Huijing Shi 6, Rui Li 1, Xiaoliang Liang 5, Shun Zhang 5, Pengfei Zhao 1, Jing Li 7, Han Lv 1, Shuohua Chen 8, Zhenghan Yang 1, Zhenjian Yu 5, Yuntao Wu 8, Ying Hui 9,✉, Shouling Wu 8,✉, Zhenchang Wang 1,✉
PMCID: PMC13360039  PMID: 42157314

Abstract

Background

Metabolic instability can affect cognitive function, but the mechanisms are largely unknown. To address this, we investigated the association between metabolic parameter variability, brain volume, cerebral blood flow (CBF), and cognitive performance, and evaluated whether CBF and brain volume mediate this relationship.

Methods

Participants were prospectively included from the Kailuan study. Between 2006 and 2020, the variability in metabolic parameters such as systolic blood pressure, fasting blood glucose, low-density lipoprotein cholesterol, and body mass index was evaluated using the coefficient of variation (CV). Starting in 2020, brain MRI and the Montreal Cognitive Assessment (MoCA) were performed continuously as part of the seventh follow-up visit and subsequent assessments. Generalized linear regression models were used to analyze the associations between metabolic variability, CBF, brain volume, and cognitive performance. Mediation analysis was performed to evaluate the mediation effects of CBF or brain volume.

Results

A total of 1894 participants (mean age, 55.4 ± 10.9 years; 51.8% male) were included. High variability scores of metabolic parameters were associated with lower CBF in total brain (β [95% confidence interval]: -2.02 [-3.73, -0.30]), total white matter (WM) (-3.04 [-4.72, -1.37]), temporal lobe (-1.82 [-3.60, -0.03]), and hippocampus (-1.91 [-3.65, -0.17]), as well as decreased volume in both total gray matter (GM) (-5.64 [-11.24, -0.04]) and temporal lobe (-1.70 [-3.16, -0.25]). High variability in metabolic parameters was associated with reduced MoCA scores (-0.98 [-1.75, -0.20]). After multiple comparison correction, the association with total WM CBF remained significant. Decreased brain volume and CBF were associated with reduced MoCA scores (P < 0.05). Mediation analysis revealed that the association between variability in metabolic parameters and MoCA scores was mediated by total WM CBF, total GM volume, and temporal lobe volume, accounting for 7.518%, 5.619%, and 6.864% of the effect, respectively.

Conclusion

Reduced total WM CBF, GM volume, and temporal lobe volume mediated the association between elevated metabolic variability and cognitive decline, with total WM CBF showing the most robust mediating effect.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13195-026-02074-9.

Keywords: Metabolic parameter variability, Coefficient of variation, Cerebral blood flow, Brain volume, Cognitive function, Mediation analyses

Background

As the global population ages, the prevalence of cognitive decline and dementia continues to increase, with the global burden of dementia expected to increase from the current 50 million to 150 million worldwide by 2050 [1]. Therefore, identifying modifiable risk factors for cognitive impairment is critical for delaying or reducing this public health burden [1, 2]. Metabolic homeostasis is crucial for preserving normal brain function [3]. Variability in metabolic parameters is an important indicator of metabolic instability. Notably, the effects of metabolic parameter variability on brain structure and cognitive decline may be independent of baseline levels of metabolic parameters [4, 5]. Emerging evidence indicates that high variability in metabolic parameters is not only a risk factor for cardiovascular disease and mortality [6–8] but also strongly associated with structural brain alterations and cognitive decline [3, 9–11].

Most existing studies on the association between metabolic parameters and the brain have primarily focused on either one single metabolic parameter, such as blood pressure (BP), fasting blood glucose (FBG), body mass index (BMI), or low-density lipoprotein cholesterol (LDL-C) [12–15], or the variability in a single metabolic parameter [11, 16–19]. Kim analyzed the associations of variability of SBP, FBG, BMI, and cholesterol concentration with mortality and cardiovascular events in the general population. Currently, limited research has explored the associations among the variability in multiple metabolic parameters, brain volume, perfusion, or cognitive function [20, 21]. We hypothesized that metabolic instability may lead to cognitive decline through changes in brain structure or perfusion. To test this hypothesis, we systematically analyzed the associations among metabolic parameter variability, brain volume, perfusion, and cognitive function using multimodal neuroimaging data from the Kailuan study. Furthermore, brain structural or perfusion alterations were evaluated as potential mediators associating variability in metabolic parameter to cognitive decline.

Methods

Study population

The Kailuan Study is a prospective cohort study conducted in the Kailuan community in Tangshan. The detailed protocols of the study design and procedures have been described previously [22]. From July 2006 to October 2007, 101,510 adults (81,110 men; 20,400 women) were enrolled from 11 hospitals in the Kailuan community. The participants underwent baseline assessments, including questionnaires, clinical examinations, and laboratory tests in one day, with biennial follow-ups thereafter. Timeline of the study is presented in Fig. 1. The seventh follow-up visit began in December 2020, and a subset of participants who were followed up at Kailuan General Hospital voluntarily completed brain MRI and cognitive assessment on the same day at Kailuan General Hospital. Between December 2020 and May 2024, 2350 participants completed the seventh follow-up examination. The inclusion criteria in this study were as follows: (a) attended the seventh follow-up and underwent simultaneous brain MRI and cognitive function assessment; (b) attended at least 2 physical examinations from 2006 to 2018. Exclusion criteria included a history of stroke, brain tumor, and incomplete or poor-quality imaging data. Based on these criteria, we enrolled a total of 1,894 participants in this study (Supplementary Fig. 1). The study protocol was approved by the Institutional Ethics Committee of Kailuan General Hospital and was conducted in accordance with the Declaration of Helsinki. All participants signed a written informed consent form.

Fig. 1.

Fig. 1

Open in a new tab

Timeline of the study

Measurement and definitions

Data on demographics (age, sex), lifestyle factors (smoking, physical activity, drinking), medical history, and medication use were collected via standardized questionnaires in the final follow-up. All procedures were performed by trained professionals. After the participants rested in a quiet room for at least 5 min, systolic BP (SBP) and diastolic BP (DBP) were measured three times at 1–2 min intervals on the left upper arm using a validated digital sphygmomanometer (Omron HBP-1100U, Kyoto, Japan), and the average of the 3 BP measurements was used for analysis. Blood samples were collected from the antecubital vein of the participants after 8 to 12 h of fasting and analyzed using an automated analyzer (Hitachi 747; Hitachi, Tokyo, Japan). FBG was measured using the hexokinase/glucose-6-phosphate dehydrogenase method. LDL-C was quantified via the enzyme colorimetric method (Mind Bioengineering Co., Ltd., Shanghai, China). BMI was calculated using weight and height, each measured three times and averaged.

Quantification of metabolic parameter variability and scoring

For each participant, the within-individual variability in SBP, FBG, LDL-C, and BMI was evaluated using the coefficient of variation (CV), calculated as follows: CV = [standard deviation (SD)/mean] × 100%. The CV of each metabolic parameter for a given participant was dichotomized at the population median; values above the median were assigned a score of 1, and those at or below the median a score of 0. A composite metabolic parameter variability score (range: 0–4) was calculated by summing the dichotomized scores across all four parameters [6]. A composite metabolic parameter variability score of 4 indicated that an individual exhibited high variability in metabolic parameters.

MRI data acquisition and processing

Brain MRI was acquired using a 3.0T MRI scanner (General Electric 750 W, Milwaukee, WI, USA). 3D pseudo-continuous arterial spin labeling (PCASL) images were acquired with the following parameters: repetition time (TR)/echo time (TE) = 5313/10.7 ms; FOV = 256 mm × 256 mm; in-plane resolution = 3.37 mm × 3.37 mm; slice thickness = 4 mm; flip angle = 111°; and post-labeling delay (PLD) = 2525 ms. The brain structural data were acquired using a 3D T1-weighted brain volume (BRAVO) sequence with the following parameters: TR/TE = 6.7/2.6 ms; FOV = 256 mm × 256 mm; flip angle = 15°; slice thickness = 1 mm; layer spacing = 1 mm; and number of slices = 170.

The 3D-PCASL images were processed using SPM12 as previously described [23], and 3D ASL Functool software (AW 4.6 Workstation, GE Healthcare, Milwaukee, WI, USA) was used to automatically generate the cerebral blood flow (CBF) maps. The CBF maps were coregistered to the Montreal Neurological Institute (MNI) space. Finally, according to the specific mask, the absolute CBF values of each region of interest (ROI) in participants were extracted, including the global region (total brain, total white matter (WM), total gray matter (GM)), and the ROIs related to cognitive decline (frontal lobe, parietal lobe, temporal lobe, and hippocampus) in previous studies [24–26]. The ROIs were selected and partitioned in a manner consistent with previous studies [27] and were selected using the Anatomical Automatic Labeling Template in the SPM12 software (Fig. 2, Supplementary Fig. 2). Two experienced doctors performed rigorous visual quality checks to exclude images with registration errors.

Fig. 2.

Fig. 2

Open in a new tab

Automatic Segmentation of the Cerebral Lobes. Cerebral Blood Flow (CBF) and brain volume of regions of interests (frontal lobe, parietal lobe, temporal lobe, and hippocampus) were extracted based on anatomical automatic labeling template

T1-weighted structural images were processed in SPM12 for reorientation, segmentation, and normalization, with methods as described in previous studies [25]. The total intracranial volume (TIV), total GM volume, and total WM volume were subsequently calculated. The total brain volume was defined as the sum of the GM volume and WM volume [28]. Absolute values of volume measurements in ROIs, including the frontal lobe, parietal lobe, temporal lobe, and hippocampus, which are associated with cognitive impairment, were obtained via the anatomical automatic labeling template in the software [24–26].

Assessment of cognitive function

Cognitive function was assessed with the Montreal Cognitive Assessment (MoCA), a validated and widely used cognitive screening tool. One psychiatrist administered the MoCA in a face-to-face setting, completing the 30-item questionnaire for assessment of different cognitive domains, within approximately 10–15 min. The scale consists of eight sections: visuospatial or executive function (maximum score of 5), naming (maximum score of 3), attention (maximum score of 6), language (maximum score of 2), abstraction (maximum score of 2), delayed recall (maximum score of 5), orienting (maximum score of 6), and memory (unscored). Participants with ≤ 12 years of education received a 1-point adjustment if their total score was < 30. Total scores range from 0 to 30, with lower scores indicating cognitive decline.

Statistical analysis

Statistical analyses were performed using SAS 9.4 (SAS Institute, Cary, NC, USA). The baseline characteristics of the participants were grouped according to the number of metabolic parameters with high variability, with data presented as mean ± SD or frequency (%). Differences across groups for continuous variables were assessed with one-way ANOVA for normally distributed data and the Kruskal–Wallis test otherwise. Categorical variables were compared between groups using the chi-square test. We applied generalized linear regression models to assess the associations between metabolic parameter variability, brain volume, CBF, and cognitive function, and multiple comparisons were corrected using the false discovery rate (FDR). The results are expressed as p values, regression coefficients β, and 95% confidence intervals (CIs). Three models were constructed. Model 1 was adjusted for sex, age, and TIV. Model 2 included covariates from Model 1 and was further adjusted for drinking, smoking, and physical activity assessed at the seventh follow-up visit. Model 3 extended Model 2 by additionally adjusting for SBP, FBG, LDL-C, BMI, and medication use (antihypertensive, hypoglycemic, hypolipidemic) at the seventh follow-up.

We used mediation analyses to assess whether brain volume or CBF changes mediated the associations between metabolic parameter variability and cognitive decline. To comprehensively explore the associations among metabolic parameter variability, brain MRI measures, and cognition, and to maximize the identification of potential mediators, we included in the mediation analyses all brain MRI measures that showed nominal associations (p < 0.05, uncorrected for multiple comparisons) with both metabolic parameter variability and cognition. Mediation analyses were conducted using SAS PROC CAUSALMED to estimate natural indirect and direct effects within the potential outcomes framework, adjusting for baseline covariates: sex, age, smoking, alcohol consumption, physical activity, systolic blood pressure (SBP), fasting blood glucose (FBG), low-density lipoprotein cholesterol (LDL-C), body mass index (BMI), and use of antihypertensive, hypoglycemic, and hypolipidemic medications at the final follow-up. Point estimates were obtained via maximum likelihood estimation, and 95% CIs were derived using the delta method under asymptotic normality. Statistical significance was defined as P < 0.05.

Results

Demographic characteristics

The study enrolled 1,894 participants, including 981 males (51.8%), with a mean age of 55.38 ± 10.88 years (range: 29.7 to 83.9 years) at the MRI examination. The average interval between the first health assessment and MRI examination was 14.6 years, and the mean number of physical examinations for the participants was 6. Table 1 shows demographic, clinical, and neuroimaging features across participant subgroups defined by the count of metabolic parameters exhibiting high variability. Participants with higher variability scores of metabolic parameters were older; had higher levels of SBP, FBG, BMI, and LDL-C; had a higher prevalence of smoking; were more likely to be users of antihypertensive, hypoglycemic, and lipid-lowering medication; and had lower global and regional CBF, lower brain volume, and lower MoCA scores.

Table 1.

Baseline characteristics of participants in the final follow-up

Variables Number of High Variability in the Metabolic Parameters
Total 0 1 2 3 4 P-value
N 1894 126 500 661 462 145
Age, year 55.38 ± 10.88 52.46 ± 10.60 53.84 ± 10.67 56.09 ± 10.97 56.31 ± 10.48 57.06 ± 11.67 < 0.001a
Male, n% 981 (51.8) 68 (54.0) 270 (54.0) 342 (51.7) 223 (48.3) 78 (53.8) 0.442b
Smoking, n% 602 (31.8) 38 (30.2) 147 (29.4) 214 (32.4) 153 (33.1) 50 (34.5) 0.656b
Drinking, n% 779 (41.1) 50 (39.7) 216 (43.2) 269 (40.7) 193 (41.8) 51 (35.2) 0.517b
Physical exercise, n% 1102 (58.2) 66 (52.4) 296 (59.2) 374 (56.6) 274 (59.3) 92 (63.4) 0.337b
BMI, kg/m2 25.33 ± 3.46 24.26 ± 3.39 25.27 ± 3.27 25.25 ± 3.48 25.54 ± 3.48 26.09 ± 3.79 < 0.001a
SBP, mm Hg 133.78 ± 19.28 125.50 ± 16.90 130.03 ± 17.57 134.61 ± 19.53 136.06 ± 18.89 142.89 ± 21.62 < 0.001a
FBG, mmol/L 5.70 ± 1.49 5.12 ± 0.80 5.44 ± 0.99 5.68 ± 1.55 6.01 ± 1.77 6.25 ± 1.69 < 0.001a
LDL-C, mmol/L 3.14 ± 0.78 3.00 ± 0.72 3.12 ± 0.75 3.12 ± 0.75 3.20 ± 0.86 3.21 ± 0.80 0.066a
Antihypertensive treatment, n% 677 (35.7) 22 (17.5) 164 (32.8) 243 (36.8) 174 (37.7) 74 (51.0) < 0.001b
Hypoglycemic treatment, n% 181 (0.10) 0 (0) 29 (0.06) 59 (0.09) 68 (0.15) 25 (0.17) < 0.001b
Lipid-lowering treatment, n% 283 (14.9) 10 (7.9) 66 (13.2) 99 (15.0) 82 (17.7) 26 (17.9) 0.041b
MoCA score 24.26 ± 3.62 25.34 ± 3.15 24.79 ± 3.36 24.18 ± 3.56 23.75 ± 3.77 23.46 ± 4.19 < 0.001a
CBF, ml/100 g/min
 Total brain 46.18 ± 7.36 47.61 ± 6.83 46.98 ± 7.31 45.87 ± 7.10 45.99 ± 7.56 44.17 ± 8.00 < 0.001a
 Total GM 49.05 ± 7.81 50.34 ± 7.31 49.89 ± 7.77 48.78 ± 7.51 48.83 ± 8.06 46.91 ± 8.41 < 0.001a
 Total WM 42.12 ± 7.02 43.89 ± 6.48 42.82 ± 7.06 41.81 ± 6.77 41.90 ± 7.14 40.21 ± 7.53 < 0.001a
 Frontal lobe 48.17 ± 7.75 49.35 ± 7.40 49.10 ± 7.84 41.81 ± 6.77 41.90 ± 7.14 40.21 ± 7.53 < 0.001a
 Parietal lobe 49.23 ± 8.59 50.63 ± 7.94 50.17 ± 8.64 47.94 ± 7.37 47.87 ± 7.98 46.01 ± 8.13 < 0.001a
 Temporal lobe 48.33 ± 7.62 49.56 ± 7.10 49.16 ± 7.49 48.08 ± 7.36 48.13 ± 7.88 46.21 ± 8.39 < 0.001a
 Hippocampus 43.93 ± 7.37 45.58 ± 7.86 44.42 ± 7.06 43.81 ± 7.22 43.63 ± 7.49 42.34 ± 7.90 0.002a
Brain volume, cm3
 Total brain 1091.15 ± 108.04 1122.93 ± 102.34 1108.52 ± 107.00 1084.51 ± 107.57 1076.18 ± 103.54 1081.62 ± 119.92 < 0.001a
 Total GM 593.08 ± 53.80 610.33 ± 47.44 602.01 ± 53.40 589.07 ± 53.64 586.61 ± 52.31 586.20 ± 58.95 < 0.001a
 Total WM 498.07 ± 59.17 512.60 ± 59.58 506.51 ± 58.79 495.43 ± 58.93 489.57 ± 56.06 495.43 ± 65.49 < 0.001a
 Frontal lobe 349.06–742.67 345.57–687.56 352.02–662.13 165.41 ± 16.92 171.25 ± 15.70 167.51 ± 16.76 < 0.001a
 Parietal lobe 82.64 ± 7.83 84.82 ± 7.25 83.86 ± 7.98 82.27 ± 7.79 81.65 ± 7.55 81.29 ± 8.07 < 0.001a
 Temporal lobe 116.28 ± 11.88 120.02 ± 10.42 118.42 ± 11.89 115.49 ± 11.77 114.79 ± 11.45 114.06 ± 13.15 < 0.001a
 Hippocampus 7.48 ± 0.83 7.63 ± 0.74 7.63 ± 0.80 7.42 ± 0.80 7.43 ± 0.85 7.28 ± 0.99 < 0.001a
Open in a new tab

Data are presented as the mean ± standard deviation, median (interquartile range), or number (percentage)

CBF  cerebral blood flow, SBP  systolic blood pressure, FBG  fasting blood glucose, LDL-C  low-density lipoprotein cholesterol, BMI  body mass index, GM  gray matter, WM  white matter

a One-way analysis of variance

b Chi-squared test

Metabolic parameter variability, CBF, and brain volume

Table 2 illustrates the associations of metabolic parameter variability with total and regional CBF. After adjustment for potential confounders (sex, age, smoking, drinking, physical activity, TIV, SBP, FBG, BMI, and LDL-C, and use of antihypertensive, lipid-lowering, and hypoglycemic medications), compared to the participants with lowest variability scores of metabolic parameters, the participants with highest variability scores of metabolic parameters exhibited lower CBF in total brain (β [95% CI]: -2.02 [-3.73, -0.30]), total WM (β [95% CI]: -3.04 [-4.72, -1.37]), temporal lobe (β [95% CI]: -1.82 [-3.60, -0.03]), and hippocampus (β [95% CI]: -1.91 [-3.65, -0.17]). Table 3 summarizes the associations of metabolic parameter variability with brain volume. Compared to the participants with lowest variability scores of metabolic parameters, the participants with highest variability scores of metabolic parameters demonstrated lower total GM volume (β [95% CI]: -5.64 [ -11.24, -0.04]) and temporal lobe volume (β [95% CI]: -1.70 [-3.16, -0.25]). After multiple comparison correction, the association between metabolic parameter variability and total WM CBF remained consistently significant.

Table 2.

Associations between metabolic parameter variability and CBF

CBF Model 1 Model 2 Model 3
Variable β (95% CI) P-value P FDR β (95% CI) P-value P FDR β (95% CI) P-value P FDR
Total brain
0 Ref Ref Ref
1 -0.57 (-1.97, 0.83) 0.424 0.495 -0.55 (-1.95, 0.85) 0.439 0.512 -0.05 (-1.42, 1.33) 0.946 0.946
2 -1.67 (-3.03, -0.30) 0.017 0.040 -1.69 (-3.06, -0.32) 0.016 0.037 -1.01 (-2.36, 0.34) 0.142 0.355
3 -1.65 (-3.07, -0.23) 0.023 0.050 -1.7 (-3.11, -0.28) 0.019 0.041 -0.75 (-2.16, 0.66) 0.296 0.592
4 -3.25 (-4.96, -1.53) < 0.001 0.003 -3.3 (-5.01, -1.59) < 0.001 0.003 -2.02 (-3.73, -0.30) 0.022 0.154
Total GM
0 Ref Ref Ref
1 -0.37 (-1.85, 1.11) 0.625 0.686 -0.35 (-1.83, 1.13) 0.646 0.701 0.21 (-1.25, 1.66) 0.782 0.811
2 -1.42 (-2.87, 0.03) 0.054 0.094 -1.44 (-2.89, 0.01) 0.051 0.087 -0.70 (-2.14, 0.73) 0.337 0.629
3 -1.48 (-2.98, 0.02) 0.053 0.094 -1.52 (-3.02, -0.02) 0.047 0.087 -0.50 (-2.00, 0.99) 0.510 0.685
4 -3.16 (-4.97, -1.34) 0.001 0.004 -3.22 (-5.03, -1.40) 0.001 0.003 -1.81 (-3.64, 0.01) 0.051 0.204
Total WM
0 Ref Ref Ref
1 -1.11 (-2.46, 0.25) 0.110 0.141 -1.11 (-2.46, 0.25) 0.110 0.140 -0.84 (-2.18, 0.51) 0.223 0.480
2 -2.20 (-3.52, -0.87) 0.001 0.004 -2.24 (-3.56, -0.91) 0.001 0.003 -1.84 (-3.17, -0.52) 0.006 0.084
3 -2.19 (-3.56, -0.81) 0.002 0.006 -2.26 (-3.64, -0.89) 0.001 0.003 -1.68 (-3.06, -0.30) 0.017 0.154
4 -3.76 (-5.42, -2.10) < 0.001 0.003 -3.81 (-5.47, -2.15) < 0.001 0.003 -3.04 (-4.72, -1.37) < 0.001 0.011
Frontal lobe
0 Ref Ref Ref
1 -0.13 (-1.6, 1.34) 0.858 0.858 -0.12 (-1.59, 1.35) 0.868 0.868 0.43 (-1.02, 1.88) 0.563 0.685
2 -1.17 (-2.61, 0.27) 0.111 0.141 -1.19 (-2.63, 0.25) 0.105 0.140 -0.44 (-1.87, 0.98) 0.541 0.685
3 -1.32 (-2.81, 0.17) 0.081 0.113 -1.36 (-2.85, 0.13) 0.073 0.102 -0.37 (-1.85, 1.12) 0.628 0.733
4 -2.95 (-4.75, -1.15) 0.001 0.004 -3.01 (-4.81, -1.21) 0.001 0.003 -1.61 (-3.41, 0.20) 0.082 0.255
Parietal lobe
0 Ref Ref Ref
1 -0.39 (-2.02, 1.25) 0.644 0.686 -0.38 (-2.01, 1.25) 0.651 0.701 0.26 (-1.35, 1.86) 0.755 0.811
2 -1.55 (-3.15, 0.05) 0.057 0.094 -1.58 (-3.17, 0.02) 0.053 0.087 -0.70 (-2.28, 0.87) 0.382 0.629
3 -1.70 (-3.36, -0.05) 0.043 0.086 -1.76 (-3.41, -0.11) 0.037 0.074 -0.57 (-2.22, 1.08) 0.497 0.685
4 -3.49 (-5.49, -1.49) 0.001 0.004 -3.56 (-5.56, -1.56) 0.001 0.003 -1.89 (-3.90, 0.11) 0.064 0.224
Temporal lobe
0 Ref Ref Ref
1 -0.32 (-1.77, 1.12) 0.661 0.686 -0.29 (-1.74, 1.16) 0.697 0.723 0.25 (-1.18, 1.68) 0.731 0.811
2 -1.34 (-2.76, 0.08) 0.064 0.094 -1.35 (-2.77, 0.06) 0.061 0.090 -0.65 (-2.06, 0.75) 0.361 0.629
3 -1.39 (-2.85, 0.08) 0.064 0.094 -1.41 (-2.87, 0.06) 0.061 0.090 -0.45 (-1.91, 1.02) 0.550 0.685
4 -3.08 (-4.86, -1.31) 0.001 0.004 -3.13 (-4.90, -1.35) 0.001 0.003 -1.82 (-3.60, -0.03) 0.046 0.204
Hippocampus
0 Ref Ref Ref
1 -1.13 (-2.55, 0.29) 0.120 0.146 -1.07 (-2.49, 0.35) 0.139 0.169 -0.58 (-1.98, 0.81) 0.413 0.642
2 -1.73 (-3.12, -0.34) 0.015 0.038 -1.74 (-3.12, -0.35) 0.014 0.036 -1.08 (-2.45, 0.30) 0.124 0.347
3 -2.00 (-3.44, -0.56) 0.007 0.020 -2.00 (-3.44, -0.56) 0.007 0.020 -1.05 (-2.48, 0.39) 0.152 0.355
4 -3.12 (-4.86, -1.38) < 0.001 0.004 -3.16 (-4.90, -1.42) < 0.001 0.003 -1.91 (-3.65, -0.17) 0.032 0.179
Open in a new tab

Model 1 was adjusted for sex, age, and TIV

Model 2 included covariates in model 1 and was further adjusted for smoking, drinking, and physical activity

Model 3 included covariates in model 2 and was further adjusted for total intracranial volume, SBP, FBG, LDL-C, BMI, and use of antihypertensive medications, use of hypoglycemic medications, and use of hypolipidemic medications in the final follow-up. The statistically significant results have been bolded.

Variable: The numbers of metabolic parameters (SBP, FBG, BMI, and LDL-C) with high variability

TIV  total intracranial volume, CI  confidence interval, CBF  cerebral blood flow, SBP  systolic blood pressure, FBG  fasting blood glucose, LDL-C  low-density lipoprotein cholesterol, BMI  body mass index, GM  gray matter, WM  white matter

Table 3.

Associations between metabolic parameter variability and brain volume

Model 1 Model 2 Model 3
Brain volume Variable β (95% CI) P-value P FDR β (95% CI) P-value P FDR β (95% CI) P-value P FDR
Total brain
0 Ref Ref Ref
1 -3.74 (-11.03, 3.55) 0.315 0.420 -4.18 (-11.46, 3.11) 0.261 0.348 -3.08 (-10.33, 4.18) 0.406 0.601
2 -8.41 (-15.55, -1.28) 0.021 0.065 -8.51 (-15.64, -1.39) 0.019 0.059 -6.88 (-14.01, 0.24) 0.058 0.208
3 -6.92 (-14.31, 0.47) 0.066 0.154 -7.19 (-14.58, 0.19) 0.056 0.130 -4.51 (-11.95, 2.93) 0.234 0.504
4 -11.25 (-20.19, -2.31) 0.014 0.049 -11.44 (-20.37, -2.51) 0.012 0.046 -7.99 (-17.04, 1.07) 0.084 0.235
Total GM
0 Ref Ref Ref
1 -2.84 (-7.35, 1.66) 0.216 0.329 -2.93 (-7.44, 1.58) 0.202 0.321 -1.85 (-6.34, 2.63) 0.418 0.601
2 -5.98 (-10.39, -1.58) 0.008 0.037 -5.96 (-10.37, -1.55) 0.008 0.037 -4.56 (-8.97, -0.16) 0.042 0.208
3 -3.89 (-8.45, 0.68) 0.095 0.205 -3.86 (-8.43, 0.71) 0.098 0.196 -1.69 (-6.29, 2.91) 0.471 0.601
4 -8.40 (-13.92, -2.88) 0.003 0.037 -8.42 (-13.95, -2.90) 0.003 0.037 -5.64 (-11.24, -0.04) 0.048 0.208
Total WM
0 Ref Ref Ref
1 -0.90 (-5.99, 4.19) 0.729 0.759 -1.24 (-6.32, 3.83) 0.631 0.654 -1.22 (-6.31, 3.87) 0.637 0.713
2 -2.43 (-7.41, 2.55) 0.338 0.430 -2.55 (-7.52, 2.41) 0.313 0.398 -2.32 (-7.32, 2.68) 0.363 0.601
3 -3.03 (-8.18, 2.12) 0.249 0.349 -3.33 (-8.48, 1.81) 0.204 0.321 -2.82 (-8.04, 2.40) 0.290 0.580
4 -2.85 (-9.09, 3.38) 0.370 0.442 -3.02 (-9.24, 3.20) 0.342 0.416 -2.34 (-8.70, 4.01) 0.470 0.601
Frontal lobe
0 Ref Ref Ref
1 -2.12 (-3.80, -0.43) 0.014 0.049 -2.13 (-3.82, -0.44) 0.013 0.046 -1.78 (-3.47, -0.08) 0.040 0.208
2 -2.32 (-3.97, -0.67) 0.006 0.037 -2.30 (-3.96, -0.65) 0.006 0.037 -1.86 (-3.53, -0.20) 0.028 0.208
3 -1.88 (-3.59, -0.17) 0.031 0.086 -1.85 (-3.56, -0.14) 0.034 0.095 -1.22 (-2.96, 0.51) 0.167 0.390
4 -2.87 (-4.94, -0.81) 0.007 0.037 -2.88 (-4.95, -0.81) 0.007 0.037 -2.03 (-4.15, 0.08) 0.059 0.208
Parietal lobe
0 Ref Ref Ref
1 -0.23 (-1.11, 0.64) 0.601 0.647 -0.28 (-1.15, 0.60) 0.535 0.576 -0.11 (-0.99, 0.76) 0.800 0.800
2 -0.53 (-1.39, 0.32) 0.222 0.329 -0.54 (-1.39, 0.32) 0.218 0.321 -0.33 (-1.19, 0.53) 0.455 0.601
3 -0.55 (-1.44, 0.33) 0.223 0.329 -0.57 (-1.45, 0.32) 0.210 0.321 -0.24 (-1.13, 0.66) 0.603 0.713
4 -1.45 (-2.53, -0.38) 0.008 0.037 -1.44 (-2.52, -0.37) 0.008 0.037 -1.02 (-2.11, 0.07) 0.067 0.208
Temporal lobe
0 Ref Ref Ref
1 -0.39 (-1.56, 0.77) 0.507 0.568 -0.45 (-1.61, 0.72) 0.451 0.516 -0.19 (-1.35, 0.97) 0.751 0.800
2 -1.17 (-2.31, -0.03) 0.045 0.115 -1.17 (-2.31, -0.03) 0.044 0.112 -0.81 (-1.95, 0.34) 0.166 0.390
3 -0.93 (-2.11, 0.25) 0.122 0.228 -0.95 (-2.13, 0.23) 0.115 0.215 -0.44 (-1.63, 0.75) 0.472 0.601
4 -2.42 (-3.85, -0.99) 0.001 0.028 -2.43 (-3.86, -1.00) 0.001 0.028 -1.70 (-3.16, -0.25) 0.022 0.208
Hippocampus
0 Ref Ref Ref
1 0.08 (-0.04, 0.20) 0.184 0.322 0.07 (-0.05, 0.19) 0.232 0.325 0.05 (-0.06, 0.17) 0.360 0.601
2 0.00 (-0.11, 0.11) 0.992 0.992 0.00 (-0.12, 0.11) 0.957 0.957 -0.02 (-0.13, 0.10) 0.794 0.800
3 0.05 (-0.07, 0.17) 0.379 0.442 0.04 (-0.07, 0.16) 0.461 0.516 0.03 (-0.09, 0.15) 0.628 0.713
4 -0.12 (-0.26, 0.03) 0.107 0.214 -0.12 (-0.26, 0.02) 0.096 0.196 -0.14 (-0.28, 0.01) 0.067 0.208
Open in a new tab

Model 1 was adjusted for sex, age, and TIV

Model 2 included covariates in model 1 and was further adjusted for smoking, drinking, and physical activity

Model 3 included covariates in model 2 and was further adjusted for total intracranial volume, SBP, FBG, LDL-C, BMI, and use of antihypertensive medications, use of hypoglycemic medications, and use of hypolipidemic medications in the final follow-up. The statistically significant results have been bolded.

Variable: The numbers of metabolic parameters (SBP, FBG, BMI, and LDL-C) with high variability

TIV  total intracranial volume, CI  confidence interval, SBP  systolic blood pressure, FBG  fasting blood glucose, LDL-C  low-density lipoprotein cholesterol, BMI  body mass index, GM  gray matter, WM  white matter

Metabolic parameter variability and cognitive function

Figure 3 and Table S1 present the associations between the variability in metabolic parameters and MoCA scores. In fully adjusted models, compared to the participants with lowest variability scores of metabolic parameters, the participants with highest variability in metabolic parameters demonstrated lower MoCA scores (β [95% CI]: -0.98 [1.75, -0.20]).

Fig. 3.

Fig. 3

Open in a new tab

The association between variability in metabolic parameters and cognitive function in model 3. Model 3 was adjusted for sex, age, smoking, drinking, physical activity, SBP, FBG, BMI, LDL-C, and use of antihypertensive medications, use of hypoglycemic medications, and use of hypolipidemic medications in the final follow-up. Variable: The numbers of metabolic parameters (SBP, FBG, BMI, and LDL-C) with high variability; MoCA, Montreal Cognitive Assessment; CI, confidence interval. SBP, systolic blood pressure; FBG, fasting blood glucose; LDL-C, low-density lipoprotein cholesterol; BMI, body mass index

CBF, brain volume, and cognitive function

Figure 4 and Table S2 show the associations between CBF, brain volume, and cognitive function. In fully adjusted models, a one SD decrease of CBF in the total brain, WM, parietal lobe, temporal lobe, and hippocampus was associated with 0.18-, 0.27-, 0.18-, 0.15-, and 0.15- point lower MoCA scores, respectively. A one SD decrease in volume of the total brain, GM, WM, frontal lobe, parietal lobe, temporal lobe, and hippocampus was associated with 0.38-, 0.42-, 0.29-, 0.33-, 0.28-, 0.43-, and 0.39-point lower MoCA scores, respectively, in fully adjusted models. After multiple comparison correction, CBF in the temporal lobe and hippocampus was not significantly associated with MoCA scores.

Fig. 4.

Fig. 4

Open in a new tab

The association between CBF, brain volume, and cognitive function in model 3. Model 3 was adjusted for sex, age, smoking, drinking, physical activity, SBP, FBG, BMI, LDL-C, and use of antihypertensive medications, use of hypoglycemic medications, and use of hypolipidemic medications in the final follow-up. MoCA, Montreal Cognitive Assessment; CI, confidence interval; Variable: CBF and brain volume in the ROIs; CBF, cerebral blood flow; GM, gray matter; WM, white matter

Mediation analysis

In fully adjusted models, mediation analysis showed that the total WM CBF, total GM volume, and temporal lobe volume mediated the negative association between variability score of metabolic parameters and MoCA scores, respectively (total WM CBF, direct effect: -0.259 [-0.402, -0.117]; indirect effect: -0.021 [-0.038, -0.005]; proportion:7.518%; total GM volume, direct effect: -0.265 [-0.407, -0.123]; indirect effect: -0.016 [-0.031, -0.0001]; proportion: 5.619%; temporal lobe volume, direct effect − 0.019 [-0.036, -0.003]; indirect effect − 0.261 [-0.403, -0.119]; proportion: 6.864%) (Fig. 5). The remaining brain volume and CBF measures did not significantly mediate the association between metabolic parameter variability and cognitive decline.

Fig. 5.

Fig. 5

Open in a new tab

Mediation effects of CBF and brain volume in the association between metabolic parameters variability and cognitive function, with total WM CBF (a), total GM volume (b), and temporal volume (c) as mediators. DE, direct effect; IE, indirect effect; WM, white matter; GM, gray matter; CBF, cerebral blood flow; MoCA, Montreal Cognitive Assessment

Discussion

Our findings indicate that higher variability score of metabolic parameters is not only associated with reduced CBF and brain volume but also serves as an independent risk factor for cognitive decline. Meanwhile, decreased CBF and brain volume were correlated with lower MoCA scores. Moreover, mediation analyses revealed that decreases in total WM CBF, total GM volume, and temporal lobe volume mediated the associations between metabolic parameter variability and cognitive decline.

This study revealed that high variability scores of metabolic parameters were negatively associated with total and regional CBF and brain volume. Individuals with higher variability scores of metabolic parameters exhibited lower WM CBF than those with the lowest variability after multiple comparison correction. In addition, these associations were independent of SBP, FBG, LDL-C, and BMI. Previous studies have separately analyzed the associations between SBP, FBG, LDL-C, BMI, and brain structural integrity and CBF, and have consistently confirmed the negative associations [11, 16–19, 29]. However, these studies were based on a single metabolic parameter and could not effectively reflect the overall impact of metabolic status on brain structural integrity. Subsequent studies have analyzed the combined effects of multiple metabolic parameters on brain structure and cognitive function [20, 21, 30], supporting our findings. Lane CA et al.20 constructed Framingham Heart study–cardiovascular risk scores (FHS-CVS) based on multiple cardiovascular risk factors and reported that higher FHS-CVS values in individuals at different ages (36, 53, and 69 years) were associated with smaller total brain volume in later life, with the strongest association observed in the younger age group. Another study [21] reported the associations of the Framingham General Cardiovascular Risk Score (FGCRS) with brain structure and found that FGCRS scores were associated with a smaller volume of GM and hippocampus, as well as greater WM hyperintensities volume. However, these prior studies utilized single-time point measurements and could not reflect the true metabolic status of individuals over the long term. In contrast, the present study calculated the long-term variability in metabolic parameters using SBP, FBG, LDL, and BMI based on 14.6 years of clinical follow-up data and confirmed the negative association between the variability in metabolic parameters and brain health. Although the underlying mechanism remains uncertain, existing evidence suggests that increased metabolic parameter variability may contribute to endothelial and smooth muscle cell dysfunction [31], leading to cerebral microvascular damage and subsequent reductions in CBF. Both GM and the temporal lobe are the key brain regions related to Alzheimer’s disease (AD) [32–34]. GM is composed mainly of neurons. Ischemia and hypoxia due to microvascular injury may accelerate neuronal damage and loss in cognitively relevant brain regions, further leading to GM volume atrophy [31, 35, 36].

In this study, we found that high variability scores of metabolic parameters were significantly associated with cognitive decline. Previous studies have explored the association between a single metabolic parameter and cognitive function [37–39]. For example, Li et al. [16] investigated the association between BP variability and cognitive decline and found that for each increase of SD in SBP variability, the MoCA score decreased by 0.24 points. Other studies have also demonstrated that elevated FBG, LDL-C, and BMI variability are respectively associated with cognitive decline [39–41]. However, the aforementioned studies each used a single metabolic parameter, with some focusing on specific age groups or symptomatic populations. Subsequently, Sherlock et al. [38] demonstrated that individuals with high variability in four or more biological measures (including SBP, heart rate, body weight, FBG, cholesterol, and triglycerides) exhibited significantly poorer cognitive function. However, this study only included dysglycemic patients aged 50 years or older, which may limit the generalizability of its findings. In the present study, based on a general community-dwelling population (age range: 29.7–83.9 years), we confirmed the negative association between high variability scores of metabolic parameters and cognitive decline, thereby enhancing the generalizability and applicability of our findings. The potential mechanism might be that long-term fluctuations in metabolic parameters can cause oxidative stress and chronic inflammation in the vascular wall [42–44]. These pathological changes can lead to dysfunction of the neurovascular unit [36, 45] and disruption of the homeostasis of the blood-brain barrier, resulting in cognitive decline [35].

This study demonstrated that reductions in brain volume and CBF were associated with cognitive decline. More importantly, the mediation analysis indicated that total WM CBF, total GM volume, and temporal lobe volume played a mediating role between high metabolic parameter variability and cognitive decline, with mediation percentages accounting for 7.518%, 5.619%, and 6.864%, respectively. Since only WM CBF remained the most consistent mediator associated with metabolic parameter variability and cognitive decline after the multiple comparison correction, its mediating effect was considered the most robust. The distribution of blood vessels in the WM is lower than that in the GM, thus, the WM is prone to ischemic injury [46]. The GM, which is mainly composed of neurons [16], is the main unit that performs neural functional activities, and ischemia and hypoxia can lead to GM atrophy [47]. GM atrophy, particularly in AD-vulnerable temporal regions, such as the temporal lobe, likely reflects neuronal loss from chronic hypoperfusion injury [47]. Previous studies have shown that high variability in single metabolic parameters is associated with a reduction of total GM volume and temporal lobe volume [29, 48–52]. Yu et al. [49] reported that each SD increase in night SBP variability reduced global and temporal GM volumes by 0.168 and 0.039 ml, respectively. High metabolic parameter variability may induce microvascular dysfunction [19], concurrently disrupting the neurovascular unit [36] and ultimately contributing to cognitive decline. Although there are no previous similar studies of humans, a rat model with high BP variability (BPV) revealed that elevated BPV predisposes cardiomyocytes to extensive damage and even death through inflammatory responses, followed by the replacement of damaged cardiomyocytes with fibrosis [53]. Additional animal studies [54, 55] have shown that high BPV can induce characteristic arteriosclerosis (afferent arterioles and interlobular arteries), leading to glomerular sclerosis, tubular atrophy, and perivascular fibrosis. The above studies suggest that increased variability in metabolic parameters may contribute to a potential risk of functional unit damage in target organs, which supports the biological plausibility of our findings.

Several limitations of this study should be acknowledged. First, all participants were from Asia. Thus, our findings may not be fully applicable to other populations. Second, as a cross-sectional study, the mechanism underlying the associations among metabolic parameter variability, brain changes, and cognitive decline requires further investigation. Third, this study solely used the MoCA scale to assess cognitive function. This limitation can be addressed in future studies via multi-dimensional and multi-standard assessments of cognitive function to improve the accuracy of our findings. Fourth, this study assumes that the variability of each metabolic parameter has the same effect on brain structure and function. In the future, we will investigate the relative importance of different parameters in influencing brain health.

Conclusions

In summary, our results indicate that high variability in metabolic parameters is not only associated with reduced CBF and brain volume but also a risk factor for cognitive decline. Reduced global and regional CBF and brain volume are associated with cognitive decline. The association between high variability in metabolic parameters and cognitive decline was partially mediated by lower WM CBF, reduced volume of global GM and temporal lobe, positioning metabolic instability as a promising early indicator of cognitive risk. Notably, total WM CBF showed the most robust mediating effect between metabolic parameter variability and cognitive decline. These findings support targeting variability in metabolic parameters not only for early risk detection but also as a modifiable factor whose control may help preserve brain structure and delay cognitive decline.

Supplementary Information

Supplementary Material 1. (1.2MB, docx)

Acknowledgements

We sincerely express our gratitude to all participants in the Kailuan Study, as well as members of Kailuan General Hospital and its affiliated hospitals.

Abbreviations

AD

Alzheimer’s disease

GM

Gray matter

BMI

Body mass index

BP

Blood pressure

BPV

Blood pressure variability

CBF

Cerebral blood flow

CI

Confidence interval

CV

Coefficient of variation

DBP

Diastolic blood pressure

FBG

Fasting blood glucose

LDL-C

Low-density lipoprotein cholesterol

MNI

Montreal neurological institute

MoCA

Montreal cognitive assessment

PCASL

Pseudo-continuous arterial spin labeling

PLD

Post-labeling delay

ROI

Region of interest

SBP

Systolic blood pressure

SD

Standard deviation

TE

Echo time

TIV

Total intracranial volume

TR

Repetition time

WM

White matter

Authors’ contributions

XSL, XJL, and ZCW conceptualized and designed the study. XJL, MMB, LY, JC, and JCF participated in the statistical analysis. LXJ led the model development, natural history calibration, and was responsible for the collation and integration of data into the model, performing model analysis. HJS, RL, XLL, SZ, PFZ, JL, HL, SHC, ZHY, ZJY, and YTW participated in the acquisition of data and analysis, and interpretation of data. XSL, XJL, and ZCW drafted the manuscript. YH, SLW, and ZCW supervised the analysis output.

Funding

This research was supported by the National Natural Science Foundation of China (Grant/Award Numbers: 82202109), Beijing Municipal Natural Science Foundation (Grant/Award Number:7232335), Beijing Scholar 2015 (Grant/Award Number: 2015 − 160), Beijing key Clinical Discipline Funding (Grant/Award Number: 2021 − 135) and Beijing Natural Science Foundation Youth Science Fund Key Project (Grant/Award Number: JR25018).

Data availability

Access to the deidentified participant data and code for the statistical analysis will be granted based on reasonable requests to the designated authors (18810833234@163.com; [cjr.wzhch@vip.163.com](mailto: cjr.wzhch@vip.163.com) ).

Declarations

Ethics approval and consent to participate

The Ethics Committee of Beijing Friendship Hospital and Kailuan General Hospital approved the study. This study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all eligible participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Xiaoshuai Li and Xuejia Li contributed equally to this work.

Contributor Information

Ying Hui, Email: huiyingct@163.com.

Shouling Wu, Email: drwusl@163.com.

Zhenchang Wang, Email: cjr.wzhch@vip.163.com.

References

  • 1.Estimation of the global prevalence of dementia. in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health Feb. 2022;7(2):e105–25. 10.1016/s2468-2667(21)00249-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet Aug. 2020;8(10248):413–46. 10.1016/s0140-6736(20)30367-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Borshchev YY, Uspensky YP, Galagudza MM. Pathogenetic pathways of cognitive dysfunction and dementia in metabolic syndrome. Life Sci Nov. 2019;15:237:116932. 10.1016/j.lfs.2019.116932. [DOI] [PubMed] [Google Scholar]
  • 4.Lee SH, Han K, Cho H, et al. Variability in metabolic parameters and risk of dementia: a nationwide population-based study. Alzheimers Res Ther Oct. 2018;27(1):110. 10.1186/s13195-018-0442-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lim S, Chung SH, Kim JH, Kim YH, Kim EJ, Joo HJ. Effects of metabolic parameters’ variability on cardiovascular outcomes in diabetic patients. Cardiovasc Diabetol May. 2023;15(1):114. 10.1186/s12933-023-01848-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kim MK, Han K, Park YM, et al. Associations of Variability in Blood Pressure, Glucose and Cholesterol Concentrations, and Body Mass Index With Mortality and Cardiovascular Outcomes in the General Population. Circulation Dec. 2018;4(23):2627–37. 10.1161/circulationaha.118.034978. [DOI] [PubMed] [Google Scholar]
  • 7.Stevens SL, Wood S, Koshiaris C, et al. Blood pressure variability and cardiovascular disease: systematic review and meta-analysis. Bmj Aug. 2016;9:354:i4098. 10.1136/bmj.i4098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Cologne J, Takahashi I, French B, et al. Association of Weight Fluctuation With Mortality in Japanese Adults. JAMA Netw Open Mar. 2019;1(3):e190731. 10.1001/jamanetworkopen.2019.0731. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Livingston JM, McDonald MW, Gagnon T, et al. Influence of metabolic syndrome on cerebral perfusion and cognition. Neurobiol Dis Apr. 2020;137:104756. 10.1016/j.nbd.2020.104756. [DOI] [PubMed] [Google Scholar]
  • 10.Claassen J, Thijssen DHJ, Panerai RB, Faraci FM. Regulation of cerebral blood flow in humans: physiology and clinical implications of autoregulation. Physiol Rev Oct. 2021;1(4):1487–559. 10.1152/physrev.00022.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Sun J, Zeng N, Hui Y, et al. Association of variability in body size with neuroimaging metrics of brain health: a population-based cohort study. Lancet Reg Health West Pac Mar. 2024;44:101015. 10.1016/j.lanwpc.2024.101015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lane CA, Barnes J, Nicholas JM, et al. Associations between blood pressure across adulthood and late-life brain structure and pathology in the neuroscience substudy of the 1946 British birth cohort (Insight 46): an epidemiological study. Lancet Neurol Oct. 2019;18(10):942–52. 10.1016/s1474-4422(19)30228-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Chung CP, Chou KH, Peng LN, et al. Associations between low circulatory low-density lipoprotein cholesterol level and brain health in non-stroke non-demented subjects. Neuroimage Nov 1. 2018;181:627–34. 10.1016/j.neuroimage.2018.07.049. [DOI] [PubMed] [Google Scholar]
  • 14.Shimoda M, Kaneko K, Nakagawa T, et al. Relationship Between Fasting Blood Glucose Levels in Middle Age and Cognitive Function in Later Life: The Aichi Workers’ Cohort Study. J Epidemiol Feb. 2023;5(2):76–81. 10.2188/jea.JE20210128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Hamer M, Batty GD. Association of body mass index and waist-to-hip ratio with brain structure: UK Biobank study. Neurology Feb. 2019;5(6):e594–600. 10.1212/wnl.0000000000006879. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Li X, Hui Y, Shi H, et al. Gray matter volume mediates the association of long-term blood pressure variability with cognitive function in an adult population. Alzheimers Dement Jul. 2024;20(7):4476–85. 10.1002/alz.13865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Zeki Al Hazzouri A, Caunca MR, Jawadekar N, et al. Associations Between 20-Year Lipid Variability Throughout Young Adulthood and Midlife Cognitive Function and Brain Integrity. J Gerontol Biol Sci Med Sci Jan. 2022;7(1):114–21. 10.1093/gerona/glab108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Smit RA, Trompet S, Sabayan B, et al. Higher Visit-to-Visit Low-Density Lipoprotein Cholesterol Variability Is Associated With Lower Cognitive Performance, Lower Cerebral Blood Flow, and Greater White Matter Hyperintensity Load in Older Subjects. Circulation Jul. 2016;19(3):212–21. 10.1161/circulationaha.115.020627. [DOI] [PubMed] [Google Scholar]
  • 19.Sible IJ, Nation DA. Blood Pressure Variability and Cerebral Perfusion Decline: A Post Hoc Analysis of the SPRINT MIND Trial. J Am Heart Assoc Jun. 2023;20(12):e029797. 10.1161/jaha.123.029797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lane CA, Barnes J, Nicholas JM, et al. Associations Between Vascular Risk Across Adulthood and Brain Pathology in Late Life: Evidence From a British Birth Cohort. JAMA Neurol Feb. 2020;1(2):175–83. 10.1001/jamaneurol.2019.3774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Song R, Xu H, Dintica CS, et al. Associations Between Cardiovascular Risk, Structural Brain Changes, and Cognitive Decline. J Am Coll Cardiol. May 2020;26(20):2525–34. 10.1016/j.jacc.2020.03.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Wu S, An S, Li W, et al. Association of Trajectory of Cardiovascular Health Score and Incident Cardiovascular Disease. JAMA Netw Open May. 2019;3(5):e194758. 10.1001/jamanetworkopen.2019.4758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Li X, Zhao P, Qiu X, et al. Altered cerebral blood flow in patients with unilateral venous pulsatile tinnitus: an arterial spin labeling study. Br J Radiol Apr. 2021;1(1120):20200990. 10.1259/bjr.20200990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Li Y, Ying Y, Yao T, et al. Decreased water exchange rate across blood-brain barrier in hereditary cerebral small vessel disease. Brain Jul. 2023;3(7):3079–87. 10.1093/brain/awac500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Li X, Hui Y, Shi H, et al. Association of blood pressure with brain perfusion and structure: A population-based prospective study. Eur J Radiol Aug. 2023;165:110889. 10.1016/j.ejrad.2023.110889. [DOI] [PubMed] [Google Scholar]
  • 26.Bang J, Spina S, Miller BL. Frontotemporal dementia. Lancet. Oct 24. 2015;386(10004):1672-82. 10.1016/s0140-6736(15)00461-4 [DOI] [PMC free article] [PubMed]
  • 27.Yan L, Liu CY, Wong KP, et al. Regional association of pCASL-MRI with FDG-PET and PiB-PET in people at risk for autosomal dominant Alzheimer’s disease. Neuroimage Clin. 2018;17:751–60. 10.1016/j.nicl.2017.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Mahinrad S, Kurian S, Garner CR et al. Cumulative Blood Pressure Exposure During Young Adulthood and Mobility and Cognitive Function in Midlife. Circulation. Mar 3. 2020;141(9):712–724. 10.1161/circulationaha.119.042502 [DOI] [PMC free article] [PubMed]
  • 29.Cui X, Abduljalil A, Manor BD, Peng CK, Novak V. Multi-scale glycemic variability: a link to gray matter atrophy and cognitive decline in type 2 diabetes. PLoS ONE. 2014;9(1):e86284. 10.1371/journal.pone.0086284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Gu T, Fu C, Shen Z, et al. Age-Related Whole-Brain Structural Changes in Relation to Cardiovascular Risks Across the Adult Age Spectrum. Front Aging Neurosci. 2019;11:85. 10.3389/fnagi.2019.00085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Ungvari Z, Toth P, Tarantini S, et al. Hypertension-induced cognitive impairment: from pathophysiology to public health. Nat Rev Nephrol Oct. 2021;17(10):639–54. 10.1038/s41581-021-00430-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Ferreira D, Verhagen C, Hernández-Cabrera JA, et al. Distinct subtypes of Alzheimer’s disease based on patterns of brain atrophy: longitudinal trajectories and clinical applications. Sci Rep Apr. 2017;18:7:46263. 10.1038/srep46263. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Mrdjen D, Fox EJ, Bukhari SA, Montine KS, Bendall SC, Montine TJ. The basis of cellular and regional vulnerability in Alzheimer’s disease. Acta Neuropathol Nov. 2019;138(5):729–49. 10.1007/s00401-019-02054-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Byun MS, Kim SE, Park J, et al. Heterogeneity of Regional Brain Atrophy Patterns Associated with Distinct Progression Rates in Alzheimer’s Disease. PLoS ONE. 2015;10(11):e0142756. 10.1371/journal.pone.0142756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Sweeney MD, Kisler K, Montagne A, Toga AW, Zlokovic BV. The role of brain vasculature in neurodegenerative disorders. Nat Neurosci Oct. 2018;21(10):1318–31. 10.1038/s41593-018-0234-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Iadecola C. The Neurovascular Unit Coming of Age: A Journey through Neurovascular Coupling in Health and Disease. Neuron Sep. 2017;27(1):17–42. 10.1016/j.neuron.2017.07.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Ferri F, Deschênes SS, Power N, Schmitz N. Association between depressive symptoms, metabolic risk factors, and cognitive function: cross-sectional results from a community study in Quebec, Canada. Aging Ment Health Nov. 2021;25(11):2003–10. 10.1080/13607863.2020.1792412. [DOI] [PubMed] [Google Scholar]
  • 38.Sherlock L, Lee SF, Cukierman-Yaffe T, et al. Visit-to-visit variability in multiple biological measurements and cognitive performance and risk of cardiovascular disease: A cohort study. Cereb Circ Cogn Behav. 2024;6:100223. 10.1016/j.cccb.2024.100223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Zhou Z, Moran C, Murray AM, et al. Association of Year-to-Year Lipid Variability With Risk of Cognitive Decline and Dementia in Community-Dwelling Older Adults. Neurology Feb. 2025;25(4):e210247. 10.1212/wnl.0000000000210247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Schaich CL, Bancks MP, Hayden KM, et al. Visit-to-Visit Glucose Variability, Cognition, and Global Cognitive Decline: The Multi-Ethnic Study of Atherosclerosis. J Clin Endocrinol Metab Dec. 2023;21(1):e243–52. 10.1210/clinem/dgad444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ye BS, Jang EY, Kim SY, et al. Unstable Body Mass Index and Progression to Probable Alzheimer’s Disease Dementia in Patients with Amnestic Mild Cognitive Impairment. J Alzheimers Dis. 2016;49(2):483–91. 10.3233/jad-150556. [DOI] [PubMed] [Google Scholar]
  • 42.Yaffe K, Bahorik AL, Hoang TD, et al. Cardiovascular risk factors and accelerated cognitive decline in midlife: The CARDIA Study. Neurology Aug. 2020;18(7):e839–46. 10.1212/wnl.0000000000010078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Steven S, Frenis K, Oelze M, et al. Vascular Inflammation and Oxidative Stress: Major Triggers for Cardiovascular Disease. Oxid Med Cell Longev. 2019;2019:7092151. 10.1155/2019/7092151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Zhong S, Li L, Shen X, et al. An update on lipid oxidation and inflammation in cardiovascular diseases. Free Radic Biol Med Nov. 2019;20:144:266–78. 10.1016/j.freeradbiomed.2019.03.036. [DOI] [PubMed] [Google Scholar]
  • 45.Ransohoff RM. How neuroinflammation contributes to neurodegeneration. Science Aug 19. 2016;353(6301):777–83. 10.1126/science.aag2590. [DOI] [PubMed] [Google Scholar]
  • 46.Takahashi S. Metabolic Contribution and Cerebral Blood Flow Regulation by Astrocytes in the Neurovascular Unit. Cells Feb. 2022;25(5). 10.3390/cells11050813. [DOI] [PMC free article] [PubMed]
  • 47.Stebbins GT, Nyenhuis DL, Wang C, et al. Gray matter atrophy in patients with ischemic stroke with cognitive impairment. Stroke Mar. 2008;39(3):785–93. 10.1161/strokeaha.107.507392. [DOI] [PubMed] [Google Scholar]
  • 48.Sible IJ, Yew B, Dutt S, et al. Selective vulnerability of medial temporal regions to short-term blood pressure variability and cerebral hypoperfusion in older adults. Neuroimage Rep Mar. 2022;2(1). 10.1016/j.ynirp.2022.100080. [DOI] [PMC free article] [PubMed]
  • 49.Yu JH, Kim REY, Park SY, et al. Night blood pressure variability, brain atrophy, and cognitive decline. Front Neurol. 2022;13:963648. 10.3389/fneur.2022.963648. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Morrison C, Oliver MD, Kamal F, Dadar M. Beyond Hypertension: Examining Variable Blood Pressure’s Role in Cognition and Brain Structure. J Gerontol B Psychol Sci Soc Sci Sep. 2024;1(9). 10.1093/geronb/gbae121. [DOI] [PMC free article] [PubMed]
  • 51.Sible IJ, Nation DA. Blood pressure variability and medial temporal atrophy in apolipoprotein ϵ4 carriers. Brain Imaging Behav Apr. 2022;16(2):792–801. 10.1007/s11682-021-00553-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Soreca I, Rosano C, Jennings JR, et al. Gain in adiposity across 15 years is associated with reduced gray matter volume in healthy women. Psychosom Med Jun. 2009;71(5):485–90. 10.1097/PSY.0b013e3181a5429d. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Kai H, Kudo H, Takayama N, Yasuoka S, Kajimoto H, Imaizumi T. Large blood pressure variability and hypertensive cardiac remodeling–role of cardiac inflammation. Circ J Dec. 2009;73(12):2198–203. 10.1253/circj.cj-09-0741. [DOI] [PubMed] [Google Scholar]
  • 54.Miao CY, Xie HH, Zhan LS, Su DF. Blood pressure variability is more important than blood pressure level in determination of end-organ damage in rats. J Hypertens Jun. 2006;24(6):1125–35. 10.1097/01.hjh.0000226203.57818.88. [DOI] [PubMed] [Google Scholar]
  • 55.Aoki Y, Kai H, Kajimoto H, et al. Large blood pressure variability aggravates arteriolosclerosis and cortical sclerotic changes in the kidney in hypertensive rats. Circ J. 2014;78(9):2284–91. 10.1253/circj.cj-14-0027. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (1.2MB, docx)

Data Availability Statement

Access to the deidentified participant data and code for the statistical analysis will be granted based on reasonable requests to the designated authors (18810833234@163.com; [cjr.wzhch@vip.163.com](mailto: cjr.wzhch@vip.163.com) ).

Articles from Alzheimer's Research & Therapy are provided here courtesy of BMC

RESOURCES