Article ID: CJ-16-0692
Background: Some disagreements surround the effects of calcium-channel blockers (CCBs) on the risk of dementia. The purpose of this study was to investigate the protective effects of CCBs on dementia among elderly hypertensive Koreans.
Methods and Results: We conducted a large population-based cohort study using the senior cohort database of the Korean National Health Insurance Service (2002–2013). Subjects were elderly hypertensive Koreans older than 60 years of age. A total of 18,423 patients (CCB user group: 13,692 patients; non-CCB antihypertensive user group: 4,731 patients) were statistically analyzed using the Cox proportional hazard regression model to estimate the adjusted hazard ratio (aHR) and confidence intervals (CIs) of dementia associated with CCB use. There were 2,881 cases (21.0%) of dementia in the CCB user group and 1,124 cases (23.8%) in the non-user group. CCB use significantly reduced the risk of total dementia (aHR 0.81, 95% CI 0.75–0.87, P<0.0001), Alzheimer’s dementia (aHR 0.80, 95% CI 0.72–0.88, P<0.0001), and vascular dementia (aHR 0.81, 95% CI 0.70–0.94, P=0.0067).
Conclusions: CCB use had a protective effect on the risk of dementia among elderly hypertensive Koreans.
Hypertension is the most powerful correctable factor for cardiovascular and cerebrovascular diseases1 and many regard it as the most important pathological factor for Alzheimer’s dementia (AD) and vascular dementia. It is also thought that long-lasting hypertension causes severe atherosclerosis and dysregulation of the cerebrovascular system.2,3 Despite these expectations, however, controversy surrounds the effects of antihypertensive medication use on the risk of dementia.2,4
Editorial p ????
In the substudy of the PROGRESS (Preventing Strokes by Lowering Blood Pressure in Patients With Cerebral Ischemia) trial, after treatment with angiotensin-converting-enzyme inhibitor (ACEI) and/or diuretics (Ds), it was shown that the risks of dementia and cognitive decline among recurrent stroke patients were reduced to 34% and 45%, respectively, over an average of 3.9 years, but those of non-recurrent stroke patients were not significantly affected.5 In SCOPE (The Study on Cognition and Prognosis in the Elderly), comparison of cognitive function measured by mini-mental state examination with angiotensin II receptor blocker (ARB) treatment groups and placebo groups revealed no significant changes.6 Additionally, in the HYVET-COG (Hypertension in the Very Elderly Trial assessing Cognitive decline and dementia incidence) study, it was found that the prevalence rate of dementia associated with use of diuretics compared with non-use was not significantly different after a mean of 2.2 years since the start of treatment.7
There are 4 classes of voltage-gated calcium channels (VGCCs): T, L, P/Q, and N types.8 Most L-type VGCCs in neurons are targeted by calcium-channel blockers (CCBs).9 Because of calcium’s importance as a key intracellular “messenger” in aging and “neurodegeneration”,10,11 it was hypothesized that L-type CBBs, especially dihydropyridine (DHP), might have protective effects for AD, or at least delay the course of the disease by blocking influx pathway of intracellular calcium in neurons. Therefore, many preclinical12,13 and clinical14,15 studies have been conducted to demonstrate these effects. According to the “calcium hypothesis,” calcium abnormalities could accelerate the formation of β-amyloid and Tau proteins, which contribute to neurodegeneration in AD.16 In the Syst-Eur trial, there was a 50% reduction in the prevalence rate of dementia after CCB treatment for a median of 2 years,14 but there was no significance associated with the effects of CCB on cognitive function in another study.15
In this study, to investigate the effect of antihypertensive medication on the risk of dementia, mainly focusing on the CCB, we conducted a senior cohort study to analyze the risk of dementia associated with use of CCBs, using the large claims database of the Korean National Health Insurance Service (NHIS).
The dataset was provided by the Korean NHIS, which was founded in 2000 as a single-insurer system. It integrated more than 366 of the medical insurance organizations in the Republic of Korea and stored and converted medical records of individuals into a national database called ‘the National Health Information Database (NHID)’, which contains personal information, demographics and medical treatment data for Korean citizens who were categorized as insured employees, insured self-employed individuals or medical aid beneficiaries. Of them, NHIS-Senior Cohort (NHIS-SC) is a population-based cohort. Because of limitations on time and money, we randomly extracted large samples that represented a Korean senior population consisting of 558,147 subjects who accounted for approximately 10% of the target 5,500,000 patients aged ≥60 years in 2002–2013. Insurance coverage was continued until death, emigration or immigration was recorded.17 A total of 672 strata were constructed by sex, location, and income level. The NHIS-SC database contains information on insurance membership and income, history of medical use, medical check-ups, and long-term care and provides medical health examinations every 2 years for all elderly beneficiaries, including blood pressure (BP) levels, health behaviors (smoking, alcohol), past medical history, etc.17
During January 1, 2002 to December 31, 2005, a total of 208,763 seniors had undergone medical check-ups. Of these, 88,139 had hypertension code (KCD-I10, I11, I12, I13, and I15 [KCD is the Korean Standard Classification of Diseases]) records for at least 1 admission or ≥3 outpatient visits18 during January 1, 2003 to December 31, 2004. Patients were excluded if they had (1) diagnostic codes for dementia or cerebrovascular disease during 2002–2005 (28,767 patients), (2) diagnostic codes of malignant cancers (13,008 patients), or (3) no history of use of antihypertensive medications (504 patients). The final group consisted of 50,755 patients (Figure 1).
Flowchart of calcium-channel blockers (CCB) cohort study. MPR, medication possession ratio; NHIS, National Health Insurance Service.
The 5 classes of antihypertensive medications were categorized as ACEIs, ARBs, β-blockers (BBs), CCBs and Ds. Classes of antihypertensive medications taken by a patient as part of a combination therapy were calculated separately.
The medication possession ratio (MPR) was calculated by dividing the sum of prescription days by total prescription periods (between the first date of prescription and the final follow-up day or the diagnostic day of dementia first recorded) using medication history extracted from the individual’s medical database during 2003–2013. In terms of CCB exposure, we defined 20% MPR as a cutoff point; patients with CCB MPR >20% were assigned to the CCB exposure group and <20% were assigned to the CCB non-exposure group.19 Within the CCB exposure group, we assigned patients who had >80% MPR to the CCB user group (CCB+) during the study period. Similar methods were applied to the non-CCB antihypertensive exposure group to assign patients to the non-CCB antihypertensive user group (non-CCB+) if they had >80% MPR. We hypothesized that each class of antihypertensive medications acts independently, which would have long-term effects on the risk of dementia.20 Person-years were calculated by adding up all the follow-up times from recruitment until the time that dementia occurred or until final examination for each individual.21
Covariate Ascertainment and Potential ConfoundersWe defined dementia subjects as those with at least 1 dementia code (F00, F01, F02, F03, F04, F05, F06, F07, G20, G21, G22, G23, G24, G25, G26, G30, G31, G32) record as an admission or 3 records as an outpatient. The index date was the day of dementia diagnosis first recorded in the study. Risk factors for dementia were considered as potential confounders, including underlying diseases (coronary artery disease (CAD), hypercholesterolemia, and diabetes), number of antihypertensive drugs, sociodemographic characteristics (age, sex, income, residential area), body mass index (BMI), BP level, and lifestyle (smoking, alcohol drinking). Income level was categorized into 2 groups, lower (1–3 deciles) and upper (4–10 deciles), based on the Insurance Premium Rating System of the Korean National Insurance Corporation. We also divided residential locality into 7 metropolitan and 9 township areas based on a 2014 report of a survey of Korean older persons.
Ethics StatementThis study was approved by the Kyung Hee University Hospital Research Ethics Committee (Approval ID: KMC IRB 1601-09).
Statistical AnalysisThe primary endpoint of this study was event-free survival, which was defined as the time from the evaluation of the induction to the date of incidence of any dementia. We compared the distribution of demographic and clinical factors and also examined the relationships between dementia events and variables. We used the Wilcoxon 2-sample test for continuous variables and Fisher’s exact test for categorical variables. Survival analysis, including Kaplan-Meier and Cox proportional hazards regression methods, was performed. The log-rank test and Cox proportional hazard regression methods were performed to examine the differences in the CCB+ and non-CCB+ cohort populations. Finally, we evaluated hazard ratio (HR) adjusted for age, sex, BMI, BP, income, urbanization, smoking, alcohol drinking, and underlying diseases (CAD, hypercholesterolemia, and diabetes).
All statistical analyses were conducted using R version 3.3.0 (R Foundation for Statistical Computing, Vienna, Austria), and SAS version 9.3 (SAS Institute Inc, Cary, NC, USA) was used to explore and modify big data.
Among elderly hypertensive Koreans older than 60 years of age, a total of 50,755 had at least 1 history of antihypertensive medication use. From 2006 to 2013, there were 12,440 (24.51%) patients who had been diagnosed with dementia (all causes), 6,360 (12.5%) patients with AD, and 2,782 (5.5%) patients with vascular dementia during 470,738.2 person-years, 488,390.3 person-years, and 496,266.7 person-years, respectively. The overall incidence rates were 26.43 cases (95% confidence interval (CI) 25.96–26.89) per 1,000 person-years for dementia (all causes), 13.02 cases (95% CI 12.70–13.34) per 1,000 person-years for AD, and 5.61 cases (95% CI 5.40–5.81) per 1,000 person-years for vascular dementia.
Incidence rates of the 5 classes of antihypertensive medication were 30.20, 30.36, 29.78, 29.04, and 31.52 per 1,000 person-years, respectively. By performing a multivariable-adjusted Cox proportional hazards regression model, we revealed that CCB had the lowest adjusted HR (aHR 0.88, 95% CI 0.85–0.90), suggesting a protective effect of CCB use on the risk of dementia. These tendencies were similarly shown for AD and vascular dementia (Table 1).
| Class of antihypertensive drug*** |
No. of users |
Dementia (all cause) | Alzheimer’s dementia | Vascular dementia | |||||
|---|---|---|---|---|---|---|---|---|---|
| No. of cases |
Person- years |
Incidence rate** |
aHR* (95% CI) |
Incidence rate** |
aHR* (95% CI) |
Incidence rate** |
aHR* (95% CI) |
||
| ARB | 21,431 | 6,160 | 204,002.4 | 30.20 (29.44–30.95) |
0.90 (0.87–0.93) |
14.39 (13.38–15.40) |
0.90 (0.86–0.94) |
6.61 (5.93–7.30) |
0.91 (0.85–0.96) |
| ACEI | 5,594 | 1,554 | 51,187.45 | 30.36 (28.85–31.87) |
1.00 (0.95–1.06) |
14.24 (13.73–14.75) |
0.98 (0.91–1.05) |
6.61 (6.27–6.95) |
0.97 (0.87–1.08) |
| BB | 13,388 | 3,736 | 125,439.7 | 29.78 (28.83–30.74) |
1 | 14.24 (13.60–14.88) |
1 | 6.53 (6.10–6.96) |
1 |
| CCB | 25,926 | 7,105 | 244,621.3 | 29.04 (28.37–29.72) |
0.88 (0.85–0.90) |
13.62 (13.17–14.07) |
0.87 (0.83–0.90) |
6.31 (6.00–6.61) |
0.88 (0.83–0.93) |
| Diuretics | 19,506 | 5,691 | 180,525.3 | 31.52 (30.71–32.34) |
1.00 (0.97–1.03) |
14.79 (14.24–15.34) |
0.96 (0.92–1.00) |
6.64 (6.28–7.00) |
0.96 (0.90–1.02) |
*Adjusted for age, sex, blood pressure, body mass index, income, urbanization, smoking, alcohol, coronary artery disease, hypercholesterolemia, diabetes, number of hypertensive drugs. **Incidence rate: per 1,000 person-years. ***Subjects with at least 1 hypertensive medication history in the NHIS-Senior cohort during the study period. If patients took multiple drugs in combination therapy, each class of antihypertensive medication was accepted individually. ACEI, angiotensin-converting-enzyme inhibitor; aHR, adjusted hazard ratio; ARB, angiotensin-receptor blocker; BB, β-blockers; CCB, calcium-channel blocker; CI, confidence interval.
A total of 18,423 patients were included as final subjects, with 13,692 patients in the CCB user group (CCB+) and 4,731 patients in the non-CCB antihypertensive user group (non-CCB+; ARB, ACEI, BB, D) (Figure 1). We found that the frequencies of underlying diseases were, in order, CAD (19.0%), diabetes (18.8%), hypercholesterolemia (12.6%), and chronic kidney disease (0.5%; Table 2). Among them, 2,881 patients (21.0%) in the CCB+ and 1,124 patients (23.8%) in the non-CCB+ groups were diagnosed with dementia (all causes). The average age of subjects was 67.25 years, and 57.2% were female. We observed that the incidence rate of dementia was 21.70 (95% CI 20.91–22.49) per 1,000 person-years in the CCB+ group and 25.83 (95% CI 24.32–27.34) per 1,000 person-years in the non-CCB+ group. The aHR of dementia associated with CCB+ compared with non-CCB+ was 0.81 (95% CI 0.75–0.87), with approximately 19% reduction of dementia risk, significantly (Table 3, Figure 2). In the subgroups of dementia, the incidence rate of AD was 10.39 (95% CI 9.85–10.93) per 1,000 person-years in the CCB+ and 12.52 (95% CI 11.49–13.55) per 1,000 person-years in the non-CCB+ groups, and the incidence rate of vascular dementia was 4.58 (95% CI 4.22–4.93) per 1,000 person-years in the CCB+ and 5.56 (95% CI 4.88–6.24) per 1,000 person-years in the non-CCB+ groups. The aHR was 0.80 (95% CI 0.72–0.88) for AD and 0.81 (95% CI 0.70–0.94) for vascular dementia, with significant risk reduction of approximately 20% and 19%, respectively (Table 3).
| Characteristic | Total (n=18,423) No. (%) |
Non-CCB antihypertensive users* (non-CCB+) (n=4,731) No. (%) |
CCB users* (CCB+) (n=13,692) No. (%) |
P value |
|---|---|---|---|---|
| Age (years) | 67.25 | 67.34 | 67.22 | 0.5582 |
| Sex | <0.0001 | |||
| Female | 10,531 (57.2) | 2,837 (60.0) | 7,694 (56.2) | |
| Male | 7,892 (42.8) | 1,894 (40.0) | 5,998 (43.8) | |
| Blood pressure | ||||
| sBP mean (SD) (mmHg) | 140.07 (±18.79) | 136.52 (±18.85) | 141.30 (±18.62) | <0.0001 |
| DBP mean (SD) (mmHg) | 83.82 (±11.57) | 82.37 (±11.68) | 84.33 (±11.49) | <0.0001 |
| PP (ΔBP=SBP–DBP) | 56.25 (±13.87) | 54.16 (±13.63) | 56.97 (±13.88) | <0.0001 |
| Body mass index | <0.0001 | |||
| Mean (SD) (kg/m2) | 25.00 (±3.08) | 24.82 (±3.13) | 25.06 (±3.06) | |
| Income (decile) | 0.0687 | |||
| Lower | 8,001 (43.4) | 2,001 (42.3) | 6,000 (43.8) | |
| Higher | 10,422 (56.6) | 2,730 (57.7) | 7,692 (56.1) | |
| Residential area | 0.5305 | |||
| Metropolitan | 8,288 (45.0) | 2,147 (45.4) | 6,141 (44.9) | |
| Town | 10,135 (55.0) | 2,584 (54.6) | 7,551 (55.2) | |
| Smoking | 0.0121 | |||
| No (never) | 14,824 (80.5) | 3,866 (81.7) | 10,958 (80.0) | |
| Yes (ever smoked) | 3,599 (19.5) | 865 (18.3) | 2,734 (20.0) | |
| Alcohol | <0.0001 | |||
| No (never) | 13,291 (72.1) | 3,608 (76.3) | 9,683 (70.7) | |
| Yes (ever drank) | 5,132 (27.9) | 1,123 (23.7) | 4,009 (29.3) | |
| Medical disease | ||||
| CAD | 3,508 (19.0) | 612 (12.9) | 2,896 (21.2) | 0.4744 |
| Hypercholesterolemia | 2,324 (12.6) | 661 (14.0) | 1,663 (12.2) | 0.0013 |
| Diabetes | 3,459 (18.8) | 905 (19.1) | 2,554 (18.7) | 0.4761 |
| CKD | 85 (0.5) | 24 (0.5) | 61 (0.5) | 0.6186 |
| Antihypertensive drugs | <0.0001 | |||
| Mean no. (mean±SD) | 1.66 (±0.66) | 1.48 (±0.62) | 1.73 (±0.66) | |
*CCB users and non-CCB antihypertensive users were those who had >80% MPR of CCB antihypertensive medication (CCB+) or other classes of medication except CCB (non-CCB+) in the study period. In the case of CCB MPR <20%, the subject was defined as a non-CCB antihypertensive user and assigned to the non-CCB antihypertensive user group. CAD, coronary artery disease; dBP, diastolic blood pressure; PP, pulse pressure; sBP, systolic blood pressure.
| CCB group |
No. of users |
Dementia (all cause) | Alzheimer’s dementia | Vascular dementia | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Person- years |
No. of cases |
Incidence rate** |
aHR* (95% CI) |
No. of cases |
Incidence rate** |
aHR* (95% CI) |
No. of cases |
Incidence rate** |
aHR* (95% CI) |
||
| CCB users (CCB+) |
13,692 | 132,765.46 | 2,881 | 21.70 (20.91– 22.49) |
0.81 (0.75– 0.87) |
1,421 | 10.39 (9.85– 10.93) |
0.80 (0.72– 0.88) |
634 | 4.58 (4.22– 4.93) |
0.81 (0.70– 0.94) |
| Non-CCB users (non-CCB +) |
4,731 | 43,509.95 | 1,124 | 25.83 (24.32– 27.34) |
1 | 566 | 12.52 (11.49– 13.55) |
1 | 256 | 5.56 (4.88– 6.24) |
1 |
*Adjusted for age, sex, blood pressure, body mass index, income, urbanization, smoking, alcohol, CAD, hypercholesterolemia, diabetes, and number of hypertensive drugs. **Incidence rate: per 1,000 person-years. CCB+, CCB user group; CI, confidence intervals; non-CCB+, non-CCB user group. Other abbreviations as in Tables 1,2.
Kaplan-Meier curves for survival free of all-cause dementia among eligible subjects assigned to CCB+ (n=13,692) and non-CCB+ (n=4,731) treatment without adjustment. CCB, calcium-channel blockers.
We conducted a subanalysis of the CCB+ group comparing DHP and non-DHP users. Of the 13,692 patients in the CCB+ group, 11,816 (86.3%) were DHP users, 269 (2.0%) were non-DHP users, and 1,607 used both DHP and non-DHP CCBs (11.7%). Among the CCB+ patients, 2,444 (20.7%) in the DHP user group and 54 (20.1%) in the non-DHP user group were diagnosed with dementia (all causes). The Cox proportional hazard regression model was used to identify the risk of dementia according to the differences between DHP and non-DHP users, and the HR was 1.08 (95% CI 0.83–1.42) in dementia (all causes), 1.13 (95% CI 0.78–1.65) in AD, and 1.00 (95% CI 0.55–1.81) in vascular dementia. These differences were not statistically significant (Table 4).
| Type of CCB | HR* | 95% CI | P value | No. of cases |
|---|---|---|---|---|
| Dementia (all cause) | ||||
| DHP | 1.00 | Reference | – | 2,444 |
| Non-DHP | 1.08 | 0.83–1.42 | 0.56 | 54 |
| Alzheimer’s dementia | ||||
| DHP | 1.00 | Reference | – | 1,206 |
| Non-DHP | 1.13 | 0.78–1.65 | 0.51 | 28 |
| Vascular dementia | ||||
| DHP | 1.00 | Reference | – | 536 |
| Non-DHP | 1.00 | 0.55–1.81 | 0.99 | 11 |
DHP, dihydropyridine. Other abbreviations as in Tables 1,3.
In this study, CCBs were shown to have greater protective effects on various types of dementia than other classes of antihypertensive medications. We observed approximately 19% risk reduction in general dementia and approximately 20% and 19% reduction for AD and vascular dementia, respectively.
From these results, we interpret CCBs to be effective not only in reducing BP as antihypertensive drugs but also in providing potential “neuroprotection”. How might CCBs protect against dementia in humans? The mechanism is still not fully understood, but it is thought that CCBs might have a neuroprotective action in addition to decreasing BP directly. Calcium is an intracellular “messenger”, and its homeostasis is likely to be crucial for maintaining “the function and viability of neurons”.11 In the process of aging, it has been shown that imbalance of intracellular calcium plays a key role in “neurodegeneration”, and thus changes in calcium influx might lead to cell pathology and apoptosis.10,11 Calcium abnormalities could accelerate AD by affecting the formation of β-amyloid and Tau proteins.16 In vitro studies show that blockade of calcium channels in neurons has a protective effect on “neurodegeneration” caused by β-amyloid.10,22
Nevertheless, it is not yet clear how to choose different types of CCBs for maximizing the neuroprotective benefits because of controversy surrounding the effects of CCB on cognitive function and reports that DHP CCBs would be more effective than non-DHP CCBs. Additionally, previous research has found that specific drugs, such as nitrendipine and nivaldipine, which can penetrate the blood-brain barrier, are more likely to affect cognitive function and improve “β-amyloid clearance”.23 Results from the present study did not support previous results on DHPs; however, we had few non-DHP users in our study because we specifically included subjects being treated for hypertension. Although other clinical studies related to antihypertensive medications and dementia did not report significant effects on dementia,7 this likely stems from insufficient statistical power and unexpected confounding variables caused by combination treatments.14
Study LimitationsFirst, the definition of CCB exposure might not be a perfect classification, which could possibly affect the results of this study because we assigned <20% MPR as a non-CCB exposure group in the study period. In a previous study, it was found that at least 3–6 months of lercanidipine use might have an influence on cognitive function,24 but it was uncertain if this resulted directly from the reduction in BP over a short-term period or originated from a “neuroprotective effect” of the drug. Therefore, it is necessary to execute further study associated with exposure periods. Second, this result was obtained using a set of data at 1 particular point within the recruiting period based on the subject’s medical check-ups undergone biennially. Therefore, it is unclear when hypertension was diagnosed and for how long the patient had been affected, how many drugs were used, whether BP was maintained or changed during follow-ups. Nevertheless, there might be some obvious points to increase the accuracy of the results. At baseline, the CCB user group has higher BP compared with non-users, but CCB users rather have reduced dementia events. By adjustment for the mean number of antihypertensive drugs and BP, we set the same conditions at baseline. Through these findings, we think that the protective effect of CCBs is actually reasonable as expected, not exaggerated. Definitely, it is important to confirm the trends in differences in BP during the study period but unfortunately we could not collect that data because of limitations of the dataset. Therefore, we defined CCB users and non-CCB users with strict MPR ≥80% in each drug category and this supports our definition that BP was well controlled in both groups during the entire study period. Also, we executed adjustment analysis for the differences in the number of hypertensive drugs compared between CCB and non-CCB users and the final result was similar when compared with that not adjusted for the number of hypertensive drugs. Third, we analyzed medication history on the basis of the prescription data system, so we could not accurately reflect the true compliance status of treatment. However, by reviewing the payment data for drugs, we supposed a high possibility that patients actually took the recorded medicine because patients had already paid for the drugs. Fourth, with regard to the diagnosis of dementia, incidence rates may have been overestimated or underestimated because we calculated these based on dementia codes extracted from databases not certificated by specific neurologists or psychiatrists. According to another study related to the accuracy of Medicare claims records in regard to “clinically-diagnosed dementia”, it has been found that there was 0.85 and 0.89 (0.64 and 0.95 for AD) sensitivity and specificity, respectively, of Medicare claims in the diagnosis of dementia.25 Fifth, because of the limits regarding fixed lists of health check-ups in the data, there may be more potential confounding factors for dementia not addressed in this study, such as circadian BP variation,26,27 genetic factors, education and physical activity. Also, arterial stiffness is another confounding factor we have to consider. It has been noted that pulse wave velocity (PWV) is a significant predictor for adverse cardiovascular events and cognitive decline. In a recent study of relatively healthy young adults, higher PWV values showed a steeper decline in specific domains such as executive cognitive score, working memory score, and memory score.28 In another study of older adults, high brachial-ankle PWV had an inverse relationship and may have a key role as an “independent” predictor of cognitive decline in older adults.29 Moreover, CCB should not be used in specific conditions, such as primary aldosteronism, chronic kidney disease or CCB-related peripheral edema developed from the use of CCB. Primary aldosteronism and CCB-related peripheral edema are not risk factors for dementia; on the other hand, chronic kidney disease is the most important risk factor for dementia. Also, we could think of the possibility that ACEi or ARB are more frequently used in most patients with chronic kidney disease. However, the incidence of diabetes mellitus and chronic kidney disease was not higher among the non-CCB users in this study. Finally, we did not perform dose-response or duration-response analyses, so we could not identify the effects of dose and duration.
However, there are several notable strengths of this study; it was a large population-based cohort study over a long-term period involving a general population using big data from the NHIS. We also analyzed the outcome of dementia by categorizing its subtype, and we performed selection of subjects using a strict criterion of ≥80% MPR during the study period.
CCBs have protective effects on the risk of dementia, including AD and vascular dementia, compared with other classes of antihypertensive medication among elderly Korean hypertensive patients. Further studies are needed to increase our understanding of the potential mechanisms of CCBs in dementia prevention.
This study was supported by the National Health Insurance Service-Senior cohort database and was not supported by any foundations.
This study used NHIS-senior (2002–2013) data (NHIS-2016-2-031) offered by the Korean NHIS. The authors declare no conflicts of interest with the NHIS.
No declared.
