Association between dietary magnesium intake, inflammation, and neurodegeneration

Background Consistent evidence shows that magnesium (Mg) intake is associated with lower blood pressure (BP), and that lower BP is associated with improved cerebral health. However, recent findings indicate that the positive effect of dietary Mg intake on cerebral health is not mediated by a decrease in BP. As Mg’s anti-inflammatory action is a plausible alternative mechanism, the objective of this study was to investigate the associations between Mg intake and inflammation to determine whether it mediates any neuroprotective effect. Methods Participants from the UK Biobank (n = 5775, aged 40–73 years, 54.7% female) were assessed for dietary magnesium using an online food questionnaire, brain and white matter lesion (WML) volumes were segmented with FreeSurfer software, and inflammation markers including high-sensitivity C-reactive protein (hs-CRP), leukocyte, erythrocyte count, and Glycoprotein acetylation (GlycA) were measured using specific laboratory techniques such as immunoturbidimetry, automated cell counting, and nuclear magnetic resonance. Hierarchical linear regression models were performed to investigate the association between dietary Mg, and inflammatory markers and between dietary Mg, brain and WMLs volumes. Mediation analysis was performed to test a possible mediation role of inflammation on the association between dietary Mg and brain and WMLs volumes. Results Higher dietary Mg intake was associated with lower inflammation: hs-CRP level (− 0.0497%; 95% confidence interval [CI] − 0.0497%, − 0.0199%) leukocytes count (− 0.0015%; 95%CI − 0.00151%, − 0.0011%), and GlycA (− 0.0519%; 95%CI − 0.1298%, − 0.0129%). Moreover, higher dietary Mg intake was associated with larger grey matter volume (0.010%; 95%CI 0.004%, 0.017%), white matter volume (0.012%; 95%CI 0.003, 0.022) and right hippocampal volume (0.002%; 95%CI 0.0007, –0.0025%). Lower hs-CRP levels mediated the positive association between higher dietary Mg intake and larger grey matter volume. Conclusions The anti-inflammatory effects of dietary Mg intake in the general population, appears to mediate its neuroprotective effect. Supplementary Information The online version contains supplementary material available at 10.1007/s00394-024-03383-1.

Alateeq 4 Figure S2 shows the results of hierarchical regression analysis exploring the association between dietary magnesium (Mg) intake and inflammatory markers (leukocytes, erythrocyte, C-reactive protein (hs-CRP), and Glycoprotein acetylation (GlycA)) in the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for main covariates, including age, sex, and education.Model 2 was further adjusted for BP medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and body mass index (BMI).Model 3 tested the twoway interactions between Mg and smoking, Mg and BP medication, Mg and hypertension, and Mg and cholesterol level.Note that error bars represent standard errors from the same model.Significance levels were indicated as *p < 0.05, **p < 0.01, and ***p < 0.001.where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Note that the hierarchical analysis results of the association between dietary Mg intake and inflammatory markers, including leukocytes, erythrocyte count, and GlycA were obtained from three models.Model 1 was adjusted for main covariates, including age, sex, and education.Model 2 was further adjusted for other covariates, such as antihypertensive medication, HDL, cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and BMI.Model 3 also examined the two-way interactions between Mg and cholesterol level.Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Note that the hierarchical analysis results of the association between dietary Mg intake and leukocytes were obtained from three models.Model 1 was adjusted for main covariates, including age, sex, and education.Model 2 was further adjusted for additional covariates, including antihypertensive medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and BMI.Model 3 tested the two-way interaction between Mg and smoking status.
Figure S4.The figure presents the results of hierarchical regression analysis examining the association between dietary magnesium intake and erythrocyte count in the UK Biobank study The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates including age, sex, and education.Model 2 was additionally adjusted for antihypertensive medication, HDL, cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and body mass index (BMI).
Model 3 was additionally tested for the two-way interactions between magnesium intake and the covariates.shows unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.It should be noted that the hierarchical analysis results of the association between dietary Mg intake and erythrocytes were obtained from three models.Model 1 was adjusted for the main covariates including age, sex, and education.Model 2 was additionally adjusted for other covariates, such as antihypertensive medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, and alcohol intake.Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.
It should be noted that the hierarchical analysis results of the association between dietary Mg intake and erythrocytes were obtained from three models.Model 1 was adjusted for the main covariates including age, sex, and education.Model 2 was additionally adjusted for other covariates, such as antihypertensive medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, and alcohol intake.represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.It should be noted that the hierarchical analysis results of the association between dietary Mg intake and erythrocytes were obtained from three models.Model 1 was adjusted for the main covariates including age, sex, and education.
Model 2 was additionally adjusted for other covariates, such as antihypertensive medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, and alcohol intake.Finally, Model 3 was additionally tested for the two-way interactions between Mg x Cholesterol level.
Figure S7.The interaction effects between a) dietary magnesium intake and smoking status in predicting the leukocytes, b) between dietary magnesium intake and Hypertension, BP medication, in predicting the C-reactive protein (hs-CRP) and c) between dietary magnesium intake and cholesterol level in predicting the GlycA.Note.
the shadow represents the 95% confidence intervals (CI).

Mg, inflammation & Brain
Alateeq 16   Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.
Note that the hierarchical analysis results of the association between Mg intake and GM volume.Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.
Note that the hierarchical analysis results of the association between Mg intake and WM volume.SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Note that the hierarchical analysis results of the association between Mg intake and LHC volume.Model 1 was adjusted for the main covariates including age, sex, education, and ICV.Model 2 was additionally adjusted for other covariates: antihypertensive medication, HDL, cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and BMI.Model 3 was additionally tested the two-way interactions between Mg x covariates.SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Note that the hierarchical analysis results of the association between Mg intake and WMLs volume.
Model 1 was adjusted for the main covariates including age, sex, education, and ICV.Model 2 was additionally adjusted for other covariates: antihypertensive medication, HDL, cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and BMI.Model 3 was additionally tested the two-way interactions between Mg x covariates.

FigureS1.
FigureS1.Flow chart of the participants selection process.

Mg
FigureS3.The figure presents the results of hierarchical regression analysis examining the association between dietary magnesium intake and leukocyte count in the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for main covariates, including age, sex, and education.Model 2 was further adjusted for antihypertensive medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and body mass index (BMI).Model 3 tested the two-way interaction between magnesium intake and smoking status.Note: Error bars represent standard errors from the same model.Significance levels are indicated as * p < 0.05; ** p < 0.01; *** p < 0.001.

"
Figure S5 The figure presents the results of hierarchical regression analysis examining the association between dietary magnesium (Mg) intake and high-sensitivity C-reactive protein (hs-CRP) levels in the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for key covariates including age, sex, and education.Model 2 additionally controlled for antihypertensive medication, HDL cholesterol, diabetes mellitus, smoking status, higher education, physical activity, and body mass index (BMI).Model 3 tested two-way interactions between Mg x Hypertension and Mg x BP medication.Error bars represent the standard error from the same model.Statistical significance was indicated by *p < 0.05, **p < 0.01, and ***p < 0.001 Figure S6 The figure presents the results of hierarchical regression analysis examining the association between dietary magnesium (Mg) intake and glycoprotein acetylation (GlycA) in the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 adjusted for age, sex, and education.Model 2 additionally adjusted for antihypertensive medication, high-density lipoprotein (HDL) cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and body mass index (BMI).Model 3 tested two-way interactions between Mg x cholesterol level.Statistical significance was denoted by *p<0.05,**p<0.01,and ***p<0.001.

Alateeq 18 Figure S8 .
Figure S8.The figure presents the results of hierarchical regression analysis examining the association between dietary magnesium (Mg) intake and gray matter volume (GM) in the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates including age, sex, education, and intracranial volume (ICV).Model 2 was additionally adjusted for other covariates: antihypertensive medication, high-density lipoprotein (HDL) cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and body mass index (BMI).Model 3 additionally tested the two-way interactions between Mg x Age.Note: Error bars represent standard error from the same model.Statistical significance was denoted by *p<0.05,**p<0.01,and ***p<0.001.
Figure S9.The figure presents the results of hierarchical regression analysis examining the association between dietary magnesium (Mg) intake and white matter volume (WM) in the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates including age, sex, education, and intracranial volume (ICV).Model 2 was additionally adjusted for other covariates: antihypertensive medication, high-density lipoprotein (HDL) cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and BMI.Model 3 additionally tested the twoway interactions between Mg x Age.Note that error bars represent the standard error from the same model.Statistical significance was denoted by *p<0.05,**p<0.01,and ***p<0.001.
Figure S10.The figure presents the results of hierarchical regression analysis examining the association between magnesium (Mg) intake and left hippocampus (LHC) volume at the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates including age, sex, education, and intracranial volume (ICV).Model 2 was additionally adjusted for other covariates: antihypertensive medication, high-density lipoprotein (HDL) cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and body mass index (BMI).Model 3 was additionally tested for two-way interactions between covariates.Note that error bars represent standard errors from the same model.Significance is denoted by * p<0.05, ** p<0.01, and *** p<0.001.

Alateeq 26 Figure S11 .
Figure S11.The figure presents the results of hierarchical regression analysis examining the association between Magnesium intake and right hippocampal (RHC) volume in the UK Biobank study.The data are presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates, including age, sex, education, and ICV.Model 2 was additionally adjusted for other covariates: antihypertensive medication, high-density lipoprotein (HDL) cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and BMI.Model 3 additionally tested the two-way interactions between Mg x covariates.Note.Error bars represent standard error from the same model.Significance is at * p<0.05; ** p<0.01; *** p<0.001.

Alateeq 28 Figure S12 .
Figure S12.Hierarchical regression analysis results of the association between Magnesium intake and white matter lesions (WMLs) at the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates including age, sex, education, and ICV.Model 2 was additionally adjusted for other covariates: antihypertensive medication, high-density lipoprotein (HDL), cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and BMI.Model 3 was additionally tested for the two-way interactions between Mg and covariates.Note.Error bars represent standard error from the same model.Significance is denoted by *p<0.05,**p<0.01,and ***p<0.001.

Alateeq 30 Figure S13 .
Figure S13.The figure presents the results of hierarchical regression analysis examining the association between magnesium intake and brain volumes, including gray matter (GM), white matter (WM), left hippocampal (LHC), right hippocampal (RHC), and white matter lesions (WMLs), at the UK Biobank study.The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Model 1 was adjusted for the main covariates, including age, sex, education, and intercranial volume (ICV).Model 2 was additionally adjusted for other covariates, including antihypertensive medication, high-density lipoprotein (HDL), cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake, and body mass index (BMI).Model 3 was additionally tested for the two-way interactions between Mg and Age.Note.Error bars represent standard error from the same model.Significance is denoted by *p<0.05,**p<0.01,and ***p<0.001.

Figure S14 .
Figure S14.The interaction effects between dietary magnesium intake and age in predicting a) gray matter volume and b) white matter volume are shown.Note.The shadow represents the 95% confidence intervals (CI).

Table S .
Demographic and health characteristics of included and excluded participants.

Table S1 . Association between dietary Mg intake and inflammation markers.
Mg refers to magnesium, and GlycA is a measure of glycoprotein acetylation HDL, high-density lipoprotein; BMI, body mass index.The data represents unstandardized Beta correlation with the standard error (+/-SE),

Table S2 .
The Association Between Dietary Magnesium Intake and Leukocytes

Table S3 .
The association between dietary magnesium intake and erythrocytes in the UK Biobank Study.

Table S6 .
The association between dietary Mg intake and brain volumes in the UK Biobank Study.WMLs, white matter lesions; HDL, high-density lipoprotein; ; BMI, body mass index; ICV, intracranial volume..The data is presented as unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.It should be noted that the hierarchical analysis results of the association between Mg intake and brain volumes, including GM, WM, LHC, RHC, and WMLs.Model 1 was adjusted for the main covariates, including age, sex, education, and intracranial volume (ICV).Model 2 was additionally adjusted for other covariates, such as antihypertensive medication, high-density lipoprotein (HDL), cholesterol, diabetes mellitus, smoking status, higher education, physical activity, alcohol intake and BMI.Finally, Model 3 tested two-way interactions between Mg x Age, Mg x BP medication, and Mg x physical activity.

Table S7 .
The association between dietary Mg intake and gray matter (GM) volume in the UK Biobank Study.

Table S8 .
The association between dietary Magnesium intake and white matter (WM) volume in the UK Biobank volume.The data presented here shows unstandardized Beta correlation with the standard error (+/-SE), where

Table S9 .
The association between dietary Mg intake and left hippocampus (LHC) volume in the UK Biobank Mg, magnesium; LHC, left hippocampus volume; HDL, high-density lipoprotein; ; BMI, body mass index; ICV, intracranial volume.The data presented here shows unstandardized Beta correlation with the standard error (+/-

Table S10 .
The association between dietary Mg intake and right hippocampal (RHC) volume in the UK Biobank The data presented here shows unstandardized Beta correlation with the standard error (+/-SE), where Beta represents the effect size of a 1 mg increase in Mg intake, expressed in SD units of the dependent variable.Note that the hierarchical analysis results of the association between Mg intake and RHC volume.Model 1 was adjusted for the main covariates including age, sex, education, and ICV.Model 2 was additionally adjusted

Table S11 .
The association between dietary Mg intake and white matter lesions (WMLs) in the UK Biobank Mg, magnesium; WMLs, white matter lesions; HDL, high-density lipoprotein; BMI, body mass index; ICV, intracranial volume.The data presented here shows unstandardized Beta correlation with the standard error (+/-

Table S12 .
Association between dietary Mg intake and inflammatory markers controlled for calcium (Ca) and Energy intake.Significance.* p<0.05; ** p<0.01; *** p<0.001.Abbreviations: CI-confidence interval -standard error; Mg -magnesium; hs-CRP -high-sensitivity C-reactive protein; GlycA -Glycoprotein acetylation.Note: Sensitivity analysis controlled for the calcium levels and energy intake in the association between dietary Mg intake and inflammatory markers, including leukocytes, erythrocytes, hs-CRP, and GlycA, using data from the UK Biobank study.The data represents unstandardized beta coefficients with 95% Confidence Interval.Beta values correspond to a 1mg unit increment in Mg intake variables.

Table 13
Association between dietary Mg intake and inflammatory markers in men and women.Significance.* p<0.05; ** p<0.01; *** p<0.001.Abbreviations: CI-confidence interval -standard error; Mg -magnesium; hs-CRP -high-sensitivity C-reactive protein; GlycA -Glycoprotein acetylation.Note: The sensitivity analysis stratified the sample by gender to investigate the association between dietary Mg intake and inflammatory markers, including leukocytes, erythrocytes, hs-CRP, and GlycA, using data from the UK Biobank study.The data represents unstandardized beta coefficients with 95% Confidence Interval.Beta values correspond to a 1mg unit increment in Mg intake variables.

Table 14 .
Association between dietary Mg intake and brain volumes and WMLs in men and women.