WIXLIB

Osteoporosis and cerebral white matter hyperintensitymonk, NY, USA), R language version 3.01 (R Foundation forStatistical Computing, Vienna, Austria), or SAS version 9.1 (SASInstitute Inc., Cary, NC, USA), and P-values 0.05 were regarded as significant.cantly correlated with both PVMH (odd ratio [OR], –0.416; 95%confidence interval [CI], –0.654 to –0.178; P 0.001) and DVMH(OR, 0.306; 95% CI, –0.546 to –0.067; P 0.012) (Table 2). Chang-es of PWMH and DWMH grades exert synergistic effect on BMD(Fig. 2). In addition, several factors showed significant correla-RESULTStions; old age and high serum homocysteine for high PWMHFinally, 408 patients were included in the study; 75.37 6.17years old and 213 (52.2%) were male. Mean BMD T-score wasgrade; old age, AF history, and high serum total calcium for highDWMH grade.–2.439 1.15 and 202 (49.5%) had osteoporosis, T-score low-er than –2.5. Distribution of PWMH subgroup was as follows;124 for grade 1, 155 for grade 2, and 129 for grade 3. Distribu-–2.0for grade 2, and 67 for grade 3. In comparison of clinical char-–2.2tion of DWMH subgroup was as follows; 203 for grade 1, 138acteristics among PWMH subgroups, age, height, BMD, hypermocysteine, ALP, and GFR were significantly different. Incomparison of clinical characteristics among DWMH sub-–2.4BMDtension history, atrial fibrillation (AF) history, triglyceride, ho-–2.6groups, age, sex, height, and BMD were significantly different–2.8(Table 1).In univariable regression analysis of PWMH as a response–3.0variable, age, sex, weight, height, BMD, hypertension history,AF history, GFR, and homocysteine showed significant association with P-value less than 0.1 (Supplementary Table 1). Inunivariable regression analysis of DWMH as a response variable, age, sex, weight, height, BMD, hypertension history, AFhistory, total calcium, and GFR showed P-value less than 0.1(Table 1). In multivariable regression analysis, BMD was signifi-Grade 1Grade ona),b)Atrial sphorusb)a),b)Total –0.0040.1230.001–0.146PWMH95% CIP-value–0.387 to 0.831–0.022 to 0.045–0.654 to –0.199–0.199 to 0.761–0.909 to 0.037–0.020 to 0.0110.063 to 0.182–0.008 to 0.010-0.4750.508 0.001c)0.2510.0710.592 0.001c)0.820-0.002 to 0.073–0.022 to 0.0451radee2GGrad3eHGradDWMFig. 2. Mean bone mineral density (BMD) of subgroups, classified withthe grade of periventricular white matter hyperintensity (PWMH) anddeep white matter hyperintensity (DWMH). According to increase ofgrade, BMD tends to decrease. There was no subjects with PWMHgrade 1 and DWMH grade 3.Table 2. Relationships between clinical factors and subtypes of cerebral white matter hyperintensityVariableGrade 4–0.0090.002–0.147–0.474DWMH95% CIP-value–0.108 to 0.910–0.006 to 0.041–0.563 to –0.088–0.243 to 0.756–1.124 to –0.145–0.022 to 0.004–0.006 to 0.010–0.543 to .021 to 0.092–0.945 to –0.0020.002c)0.049Multivariable ordinal regression analysis was performed.PWMH, periventricular white matter hyperintensity; DWMH, deep white matter hyperintensity; OR, odd ratio; CI, confidence interval; BMD, bonemineral density; GFR, glomerular filtration rate; ALP, alkaline phosphatase.a)Variables for PWMH; b)Variables for DWMH; C)P 0.05 was presented as bold characters.120http://pfmjournal.org

DISCUSSIONJi-Sun Kim, et al.In the present study, severity of cerebral white matter hyperin-risk of cardiovascular disease such as myocardial infarction orstroke [21,22]. Bisphosphonate has been suggested to reducecardiovascular disease by the following mechanisms; inhibi-tensity on MRI showed significant associations with BMD, com-tion of intimal hyperplasia, macrophage-related inflamma-significant association between presence of osteoporosis andta-analysis showed different results that bisphosphonate dopatible results with the previous study. Previous study showedcerebral SVD on CT [7]. Although other previous study statedno significant correlation between SVD and low BMD, the studyused phalynx rather than femur neck as measuring site of BMD[13]. Because phalynx is not a recommended site for BMD measuring to assess osteoporosis in clinical practice, BMD mea-sured in phalynx maybe insufficiently reflect the presence ofosteoporosis [14].It is suggested that sharing of risk factor might explain theassociation between osteoporosis and cerebral SVD [7]. Tra-tion, and vascular calcification [22,23]. However, recent me-not act as beneficial or harmful factor on atherosclerosis [24].Meanwhile, calcium supplements may be harmful to cardio-vascular disease [25,26]. Contrast to daily calcium intake dose,artificial calcium supplements lead to significant increase ofserum calcium level. The increased calcium level is associatedwith increased cardiovascular risk [25,27,28]. Whether activetreatment of osteoporosis reduces the risk of cerebral SVD isnot clear and careful approach is needed.In addition to BMD, several factors were associated with ce-ditionally, cerebral SVD that refers to atherosclerosis hasrebral SVD. Age was independent risk factor for both PWMHsion, diabetes mellitus, dyslipidemia, smoking, or homocys-ly correlated with DWMH (P 0.011; OR, –0.634). Although sta-been known as having risk factors such as aging, hypertenteine [15]. Osteoporosis has several risk factors similar to cerebral SVD in part, like smoking and old age. However, in thisand DWMH, as reported in previous study [29]. AF was inversetistically not significant, PWMH also showed tendency of nega-tive association with AF (P 0.071; OR, –0.436). The results canstudy, the association between cerebral white matter hyper-be explained by the following. The study was performed withthe association could not be explained by risk factors shar-oembolic stroke, and patients with cardioembolic stroke haveintensity and BMD was independent of such risk factors, thusing. Although the underlying pathophysiologic mechanismsbetween osteoporosis and cerebral SVD are still on debate,one of the possible explanations is that osteoporosis-relatedabnormal calcium metabolism could increase vascular calcification and arterial wall stiffness [5,16]. Vascular calcifica-tion has been suggested as a marker of cerebral SVD in recent observations [17,18]. The association may also be explained by common pathophysiology of the two conditions.Atherosclerosis, a main pathology of cerebral SVD, and oste-acute symptomatic stroke patients. AF is a major cause of cardi-smaller cerebral white matter hyperintensity volume thanthose with large artery atherosclerosis or small vessel occlusion[30,31]. Homocysteine showed significant association withPVMH but not with DVMH. In the contrast, serum calcium levelis associated with DVMH only. Discrepancy of significant association factors between PWMH and DWMH may indicate that thetwo categories have different pathomechanism in part, whichis not clearly identified, and reserved for future investigation.The present study has several limitations. First, the studyoporosis are known to have common pathogenesis like bonepopulation consisted of a single-race, was selected from hos-factor-kappaB ligand (RANKL)/RANK/osteoprotegerin (OPG)were not included. Thus there are some limitations to gener-morphogenetic proteins, the receptor activator of nuclearpathway, matrix Gla protein (MGP) and vitamin K [19]. Lowvitamin D level also is known to be associated with increasedvascular risk [20].If osteoporosis is an independent risk factor of cerebral SVD,we can assume that treatment of osteoporosis reduces the occurrence or progression of cerebral SVD. However, such con-cept should be considered with caution. Two well-establishedagents for treating osteoporosis, bisphosphonate, and calci-um, showed a dissimilar influence to atherosclerosis. Therehas been no consensus about the effect of bisphosphonate onatherosclerosis. Several observational studies demonstratedthat bisphosphonate treatment had benefit on reducing d symptomatic stroke patients and healthy subjectsalize the results. Second, subjects taking osteoporosis medications were excluded for the study. The strict inclusion crite-ria was unavoidable due to more homogeneous study populations; however, subjects with severe or symptomatic osteoporosis may be eliminated from the study and it can attenuate the results. Finally, MRI equipment used for the study wasnot identical to each other, one was 1.5 Tesla and the otherwas 3.0 Tesla. Because measurement of cerebral white matter intensities is a precision task and requires high resolutionimages, heterogeneity of MRI equipment may influence tomeasured size of cerebral white matter intensities. However,possible missing or mis-sized lesions on low tesla MRI are of121

Osteoporosis and cerebral white matter hyperintensityten small-sized; hence, they do not significantly influence tograding of cerebral white matter hyperintensity, because thegrade in the study was decided by size of biggest lesion [32].In conclusion, low BMD was associated with a high severityof cerebral white matter hyperintensity in elderly Asian strokepatients. This association was independent of a shared etiol-ogy of osteoporosis and cerebral SVD. Osteoporosis could be10. Alagiakrishnan K, Hsueh J, Zhang E, Khan K, Senthilsel-van A. Small vessel disease/white matter disease of thebrain and its association with osteoporosis. J Clin MedRes 2015;7:297-302.11. Kim DH, Rogers JR, Fulchino LA, Kim CA, Solomon DH,Kim SC. Bisphosphonates and risk of cardiovascularevents: a meta-analysis. PLoS One 2015;10:e0122646.linked to dementia or stroke through cerebral SVD, it would12. Foley RN, Collins AJ, Ishani A, Kalra PA. Calcium-phos-fect of osteoporosis treatment on cerebral SVD is still on de-ty-dwelling adults: the Atherosclerosis Risk in Communi-be important to screen osteoporosis in advance. But the efbate and further longitudinal studies are requested.CONFLICTS OF INTERESTNo potential conflict of interest relevant to this article was reported.REFERENCESphate levels and cardiovascular disease in communities (ARIC) Study. Am Heart J 2008;156:556-63.13. Kang JH, Keller JJ, Lin HC. A population-based 2-year follow-up study on the relationship between bisphospho-nates and the risk of stroke. Osteoporos Int 2012;23:25517.14. Noh Y, Lee Y, Seo SW, Jeong JH, Choi SH, Back JH, et al. Anew classification system for ischemia using a combination of deep and periventricular white matter hyperintensities. J Stroke Cerebrovasc Dis 2014;23:636-42.1. Prevention and management of osteoporosis. World15. Rost NS, Rahman RM, Biffi A, Smith EE, Kanakis A, Fitz-2. Minn YK, Suk SH, Do SY. Osteoporosis as an independentcreased in small vessel stroke subtypes. Neurology 2010;Health Organ Tech Rep Ser 2003;921:1-164.risk factor for silent brain infarction and white matterchanges in men and women: the PRESENT project. Osteoporos Int 2014;25:2465-9.3. Lampropoulos CE, Papaioannou I, D’Cruz DP. Osteoporo-patrick K, et al. White matter hyperintensity volume is in75:1670-7.16. Choi JY, Seo WK, Kang SH, Jung JM, Cho KH, Yu S, et al.Statins improve survival in patients with cardioembolicstroke. Stroke 2014;45:1849-52.sis: a risk factor for cardiovascular disease? Nat Rev17. Seo WK, Lee JM, Park MH, Park KW, Lee DH. Cerebral mi-4. Bolland MJ, Avenell A, Baron JA, Grey A, MacLennan GS,stiffness in stroke patients. Cerebrovasc Dis 2008;26:618-Rheumatol 2012;8:587-98.Gamble GD, et al. Effect of calcium supplements on riskcrobleeds are independently associated with arterial23.of myocardial infarction and cardiovascular events: me-18. Mussolino ME, Madans JH, Gillum RF. Bone mineral den-5. Santos LL, Cavalcanti TB, Bandeira FA. Vascular effects of19. Hong NR, Seo HS, Lee YH, Kim JH, Seol HY, Lee NJ, et al.ta-analysis. BMJ 2010;341:c3691.bisphosphonates-a systematic review. Clin Med InsightsEndocrinol Diabetes 2012;5:47-54.6. Kang JH, Keller JJ, Lin HC. Bisphosphonates reduced therisk of acute myocardial infarction: a 2-year follow-upstudy. Osteoporos Int 2013;24:271-7.7. Carda S, Cisari C, Invernizzi M, Bevilacqua M. Osteoporo-sis after stroke: a review of the causes and potentialtreatments. Cerebrovasc Dis 2009;28:191-200.sity and stroke. Stroke 2003;34:e20-2.The correlation between carotid siphon calcification andlacunar infarction. Neuroradiology 2011;53:643-9.20. Poole KE, Reeve J, Warburton EA. Falls, fractures, and osteoporosis after stroke: time to think about protection?Stroke 2002;33:1432-6.21. Wardlaw JM, Smith C, Dichgans M. Mechanisms of sporadic cerebral small vessel disease: insights from neuroimaging. Lancet Neurol 2013;12:483-97.8. Meyer JS, Kawamura J, Terayama Y. White matter lesions22. Frost ML, Grella R, Millasseau SC, Jiang BY, Hampson G, Fo-9. Wardlaw JM, Brindle W, Casado AM, Shuler K, Hendersonrotic plaque and arterial stiffness to bone mineral densityin the elderly. J Neurol Sci 1992;110:1-7.M, Thomas B, et al. A systematic review of the utility of 1.5versus 3 Tesla magnetic resonance brain imaging in clinical practice and research. Eur Radiol 2012;22:2295-303.122gelman I, et al. Relationship of calcification of atheroscleand osteoprotegerin in postmenopausal women referredfor osteoporosis screening. Calcif Tissue Int 2008;83:11220.http://pfmjournal.org

Ji-Sun Kim, et al.23. Jorde R, Sundsfjord J, Fitzgerald P, Bonaa KH. Serum cal-and neurodegeneration. Lancet Neurol 2013;12:822-38.cium and cardiovascular risk factors and diseases: the28. Golob AL, Laya MB. Osteoporosis: screening, prevention,24. Iijima K, Akishita M, Ouchi Y. Coronary artery calcification29. Beveridge LA, Witham MD. Vitamin D and the cardiovas-Tromso study. Hypertension 1999;34:484-90.and cerebral small vessel disease. Association of systemic atherosclerosis. Circ J 2011;75:272-3.25. Lee SB, Cho AH, Butcher KS, Kim TW, Ryu SY, Kim YI. Lowbone mineral density is associated with poor clinical outcome in acute ischemic stroke. Int J Stroke 2013;8:68-72.and management. Med Clin North Am 2015;99:587-606.cular system. Osteoporos Int 2013;24:2167-80.30. Wahlund LO, Barkhof F, Fazekas F, Bronge L, Augustin M,Sjogren M, et al. A new rating scale for age-related whitematter changes applicable to MRI and CT. Stroke 2001;32:1318-22.26. Myint PK, Clark AB, Kwok CS, Loke YK, Yeong JK, Luben RN,31. Bolland MJ, Barber PA, Doughty RN, Mason B, Horne A,pean prospective investigation into cancer-norfolk popula-ceiving calcium supplementation: randomised controlledet al. Bone mineral density and incidence of stroke: Euro-tion-based study, systematic review, and meta-analysis.Stroke 2014;45:373-82.27. Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, FazekasF, Frayne R, et al. Neuroimaging standards for researchinto small vessel disease and its contribution to ageinghttps://doi.org/10.23838/pfm.2018.00114Ames R, et al. Vascular events in healthy older women retrial. BMJ 2008;336:262-6.32. Chutinet A, Biffi A, Kanakis A, Fitzpatrick KM, Furie KL,Rost NS. Severity of leukoaraiosis in large vessel atherosclerotic disease. AJNR Am J Neuroradiol 2012;33:1591-5.123

Supplementary Table 1. Univariable regression analysis between clinical factors and cerebral white matter hyperintensity subtypesVariableAgeORPWMH95% CIP-value0.070 to 0.7900.019a)0.0790.048 to 0.109Weight–0.022–0.040 to –0.004BMI–0.031–0.093 to 0.0320.4830.056 to 0.911SexHeightBMDClinical historyHypertensionDiabetes mellitus0.430–0.026–0.438–0.047 to –0.006–0.602 to –0.275–0.016–0.034 to 0.002–0.057 to –0.0130.002a)a) 0.001–0.414–0.583 to –0.245 0.001a)0.027a)0.4610.010 to 0.9120.045a)0.332ALP0.004–0.003 to 0.0110.269Total calcium0.026–0.393 to 0.4450.902Laboratory eLipoprotein-α0.095–0.355 to 0.787–0.237 to 0.4270.088a)–0.0350.5730.2130.706 0.001a)0.013a)–0.193 to 0.349Previous stroke 0.001a)0.015a)0.078Smoking0.332 to 1.0800.048 to 0.1100.615–0.919 to –0.073P-value0.079–0.467 to 0.276–0.49695% CI 0.001a)–0.095Atrial fibrillationORDWMH0.021–0.264–0.043 to 0.085–0.650 to 0.1220.5210.1800.022–0.443–0.887 to 0.0020.051a)0.4590.421–0.157 to 0.9980.153a)–0.173–0.457 to 0.1100.2300.003–0.005 to 0.0100.494–0.451–0.887 to –0.0140.043a)0.573–0.029–0.371 to 0.3120.867–0.022–0.033 to –0.011 0.001–0.011–0.022 to 0.0000.042a)0.000–0.016 to 0.0160.964–0.001–0.018 to 0.0150.881 0.001a)0.0240.0020.0020.1130.017–0.004 to 0.008–0.002 to 0.0050.066 to 0.160–0.108 to 0.143a)0.4230.3660.7880.0040.0000.038–0.002 to 0.011–0.003 to 0.004–0.021 to 0.068–0.092 to 0.1670.1580.8300.3030.568Univariable ordinal regression analysis was performed.PWMH, periventricular white matter hyperintensity; DWMH, deep white matter hyperintensity; OR, odd ratio; CI, confidence interval; BMI, bodymass index; BMD, bone mineral density; ALP, alkaline phosphatase; GFR, glomerular filtration rate; LDL-C, low density lipoprotein cholesterol;HDL-C, high density lipoprotein cholesterol.P 0.1.a)

Results: Four hundred eight patients were included in the study. High severity grade of both MWMH and DWMH was independently correlated with low BMD. . School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Korea Tel: 82-2-3410-0799 . Belford, MA, USA) for lumbar spine and pre-sented as T-score: T-score (individual BMD-reference BMD)/ .