Abstract
Aim: This study aimed to investigate the associations of leukocyte count with the risks of stroke and coronary heart disease among the general Japanese population.
Methods: A total of 5,242 residents aged 40–69 years living in two Japanese communities underwent leukocyte count measurements between 1991 and 2000, and the data were updated using 5- or 10-year follow-ups or both. Participants who had histories of stroke, coronary heart disease, or high values of leukocyte count (＞130×10 2 cells/mm 3 ) were excluded. Hazard ratios with 95% confidence intervals (CIs) were calculated according to quartiles of cumulative average leukocyte count.
Results: During follow-up of 21 years, 327 stroke and 130 coronary heart disease cases were determined. After adjustments for age, sex, community, and updated cardiovascular risk factors, the multivariable hazard ratio (95% CI) for the highest versus lowest quartile of leukocyte count was 1.50 (1.08–2.08) for ischemic stroke, 1.59 (1.00–2.51) for lacunar infarction, 1.42 (0.90–2.26) for non-lacunar infarction, 2.17 (1.33–3.55) for coronary heart disease, and 1.40 (1.11–1.76) for total cardiovascular disease. In smoking status-stratified analyses, the corresponding multivariable hazard ratio (95% CI) was 2.45 (1.11–5.38) for ischemic stroke, 2.73 (1.37–5.44) for coronary heart disease in current smokers, 2.42 (1.07–5.46), 1.55 (0.58–4.15) in former smokers, and 1.17 (0.75–1.82), 1.78 (0.83–3.82) in never smokers.
Conclusion: Leukocyte count was positively associated with the risks of ischemic stroke and coronary heart disease among the general Japanese population, especially in current smokers.
Keywords: Leukocyte count, Coronary heart disease, Ischemic stroke, Lacunar infarction, Cohort study
Introduction
Leukocytes are major immune system cells responsible for protecting the body from infection and foreign materials. Increased leukocyte count reflects the presence of inflammation, which promotes the pathogenesis of atherosclerosis 1) . Previous prospective cohort studies have reported the positive associations of leukocyte count with the risks of ischemic stroke and coronary heart disease 2 - 7) . However, two studies reported no association between leukocyte count and the risk of ischemic stroke 8 , 9) . Leukocyte count was positively correlated with several cardiovascular risk factors, especially for cigarette smoking 10) . The positive association of leukocyte count with the risk of ischemic stroke was reported in current smokers 2 , 3) , former smokers 3) , but not in never smokers, whereas the positive association with the risk of coronary heart disease was reported in both current, former, and never smokers 2 , 8 , 9) .
In this study, we aimed to investigate the associations of leukocyte count with the risks of stroke and coronary heart disease among the general Japanese population. We hypothesized that leukocyte count is positively associated with the risks of ischemic stroke and coronary heart disease, and the associations could be observed in both current, former, and never smokers.
Methods
Study Population
This study is a part of the Circulatory Risk in Communities Study (CIRCS), an ongoing dynamic community cohort study of cardiovascular disease among the general Japanese population since 1963 11 , 12) . Residents from Ikawa town, a rural community in Akita Prefecture in northwestern Japan, and Kyowa district of Chikusei city, a rural community in Ibaraki Prefecture in mid-eastern Japan were invited to participate in the annual health checkups. A total of 5,360 residents aged 40–69 years were tested for leukocyte count at the baseline from 1991 to 2000. After the exclusion of 104 participants who had histories of stroke or coronary heart disease, and 14 who had high values of leukocyte count (＞130×10 2 cells/mm 3 ), 5,242 participants were available in the analyses. Among them, 1,861 (36%) participants continued to participate in the annual health checkups at the 5-year follow-up, and 2,763 (53%) at the 10-year follow-up, so that the baseline examination data could be updated using the 5-year follow-up or 10-year follow-up or both. The informed consent was obtained from the representatives of the community according to the guidelines of the Council for International Organizations of Medicine of Sciences. This study obtained ethical approval from the Ethics Committees of the Osaka Center for Cancer and Cardiovascular Disease Prevention and Osaka University.
Baseline Examination
Blood was collected into a serum separating tube from participants who are in a seated position and separated within 30 min to obtain the serum. The serum sample of leukocyte count was collected in a tube containing EDTA-2K and measured using automated hematology analyzer (model S-Plus VI, STKR, or Gen-S; Beckman Coulter Inc., Brea, CA). Serum total cholesterol and triglycerides were measured using the enzymatic method during 1991–1993, and the enzymatic method for free glycerol during 1994–2000. Serum high-density lipoprotein levels after heparin-manganese precipitation were measured using the direct Liebermann–Burchard method during 1991–1993, and the dextran sulfate–phosphotungstate–MgCl 2 precipitation method during 1994–2000. Serum glucose was measured using the hexokinase method during 1991–1993, and the glucokinase method during 1994–2000. Blood measurements were performed at the Osaka Medical Central for Cancer and Cardiovascular Disease, an international member of the US National Cholesterol Reference Method Laboratory Network 13 , 14) .
In the annual health checkups, height (in stockinged feet) and weight (in light clothing) were measured, and body mass index was calculated. Information on smoking status, number of cigarettes per day, usual weekly intake of alcohol evaluated by units of “go” (a traditional Japanese unit of volume corresponding to 23-g ethanol), and medication uses were obtained via interviews. Blood pressure was measured on the right arm of participants who are in a seated position using standard mercury sphygmomanometers and unified epidemiological methods 15) . Hypertension was defined as systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, and use of antihypertensive medication. Diabetes mellitus was defined as a fasting glucose level of ≥ 7.0 mmol/L, non-fasting glucose level of ≥ 11.1 mmol/L, and use of medication for diabetes mellitus.
Follow-Up and Ascertainment of Cases
The participants were followed up to the end of 2017 for Ikawa and 2014 for Kyowa to determine incident stroke and coronary heart disease. The median follow-up period was 21 years. As described previously 11 , 12) , the CIRCS researchers ascertained the information of candidate cases of stroke and coronary heart disease from death certificates, national insurance claims, annual household questionnaires, annual cardiovascular risk surveys, and reports by either local physicians, public health nurses, or healthy volunteers. All living suspected cases were telephoned, visited, or invited to take part in a questionnaire survey, and medical records were obtained from the local clinics and hospitals to confirm the diagnosis. For deaths, medical histories were obtained from families or attending physicians, and related medical records were surveyed.
Stroke was defined as a focal neurological disorder with rapid onset and persisted at least 24 h or until death. Stroke subtypes, including hemorrhagic stroke and ischemic stroke (lacunar infarction or non-lacunar infarction), were mainly diagnosed using computed tomography (CT) or magnetic resonance imaging (MRI) 16) , which were available for 94% of stroke cases. When CT or MRI data were unavailable, stroke subtypes were classified according to clinical criteria as hemorrhagic stroke, ischemic stroke, and unclassified stroke. The modified coronary heart disease criteria proposed by the World Health Organization Expert Committee 17) were used to confirm the coronary heart disease cases in the medical records. Definite myocardial infarction was diagnosed by typical chest pain, lasting ≥ 30 min without definite nonischemic causes, and the appearance of abnormal and persistent Q or QS waves on the electrocardiogram and consistent changes in cardiac enzyme activity or both. These patients who had typical chest pain but whose electrocardiographic and enzyme levels were non-diagnostic or unavailable were diagnosed as possible myocardial infarction. Angina pectoris was diagnosed as repeated episodes of chest pain when exerting effort, especially when walking, usually disappearing rapidly after the cessation of effort or with the use of sublingual nitroglycerin. Sudden cardiac death was diagnosed as death ≤ 1 h after onset, a witnessed cardiac arrest, or abrupt collapse not preceded by ＞1 h of symptoms. The initial case of definite or probable myocardial infarction, angina pectoris, or sudden cardiac death was indicated as coronary heart disease. The final diagnoses were based on a panel of experienced physician–epidemiologists using the same diagnostic criteria and blinded to the data from the risk factors survey.
Statistical Analyses
Participants were classified according to quartiles of leukocyte count. To reduce misclassification, cumulative averages of leukocyte count at the baseline, 5- or 10-year follow-up surveys or both were used to rank the quartiles instead of the single measurements at baseline. Continuous confounding variables (body mass index, ethanol intake, blood pressure, and lipids) were also updated using the same method. For categorical variables (smoking status, antihypertensive medication use, and diabetes mellitus), the last observations were used. Among participants, 64% had no data on the 5-year follow-up survey, and 47% had no data on the 10-year follow-up survey. For participants who had no data on the 5- and 10-year follow-up surveys, only the baseline data were used. For those who had data on the 5-year follow-up survey but no data on the 10-year follow-up survey, averages of the baseline and 5-year follow-up survey for continuous variables or the observations at the 5-year follow-up survey for categorical variables were used. Moreover, for those who had data on the 10-year follow-up survey but no data on the 5-year follow-up survey, averages of the baseline and 10-year follow-up survey or the observations at the 10-year follow-up survey were used. When participants died, moved out from communities, or had a stroke/coronary heart disease before the 5- or 10-year follow-up surveys, the baseline or updated data before those events were used for analyses.
The analyses of covariance were used to calculate age-, sex-, and community-adjusted mean values or prevalence of updated cardiovascular risk factors. Cox proportional hazard models with updated leukocyte count and confounding variables at the baseline and 5- and 10-year follow-up surveys were used to calculate hazard ratios with 95% confidence intervals (CIs). The hazard ratios were performed according to quartiles of cumulative average leukocyte count referenced to the lowest quartiles, and 1 standard deviation (SD) increment of cumulative average leukocyte count (16.4×10 2 cells/mm 3 ). The trends in the associations were tested using the median values of the pertinent leukocyte count quartiles.
The initial hazard ratio models were adjusted for age, sex, and community, and these multivariable models were further adjusted for updated body mass index (kg/m 2 ), cigarette smoking status (never, former, and current 1–19 or 20 cigarettes per day), ethanol intake (g/day), systolic blood pressure (mmHg), antihypertensive medication use (no or yes), tertiles of serum non-HDL cholesterol levels and HDL cholesterol levels (mmol/L), tertiles of serum triglyceride levels (mmol/L), and diabetes mellitus (no or yes). The smoking status-stratified analyses were performed according to never, former, and current smokers. All analyses were conducted using the SAS System for Windows (version 9.4; SAS Inc., Cary, NC).
Results
The cardiovascular risk characteristics of the baseline population according to quartiles of cumulative average leukocyte count are listed in Table 1 . The range of each leukocyte count quartile was 18–52, 53–62, 63–74, and 75–130×10 2 cells/mm 3 , and the median value was 47, 58, 68, and 83×10 2 cells/mm 3 , respectively. Participants who had higher leukocyte count quartiles were younger and had a higher proportion of men than those who had the lowest quartiles. After adjustments for age, sex, and community, leukocyte count was positively associated with body mass index, systolic and diastolic blood pressure, antihypertensive medication use, serum non-HDL cholesterol levels, serum triglycerides, ethanol intake, the prevalence of current smokers and diabetes mellitus, and inversely associated with serum HDL cholesterol levels.
Table 1. Cardiovascular risk characteristics of participants according to quartiles of cumulative average leukocyte count.
| Leukocyte count quartiles (×10 2 cells/mm 3 ) |
P for difference |
||||
|---|---|---|---|---|---|
| Q1 (low) | Q2 | Q3 | Q4 (high) | ||
| Range of leukocyte count,×10 2 cells/mm 3 | 18–52 | 53–62 | 63–74 | 75–130 | |
| Median leukocyte count,×10 2 cells/mm 3 | 47 | 58 | 68 | 83 | |
| No. at risk | 1,310 | 1,311 | 1,308 | 1,313 | |
| Age at baseline, year | 56 (0.2) | 55 (0.2) | 55 (0.2) | 54 (0.2) | ＜0.001 |
| Men at baseline, % | 32 | 36 | 41 | 50 | ＜0.001 |
| Body mass index, kg/m 2 | 23.3 (0.1) | 23.9 (0.1) | 24.1 (0.1) | 24.3 (0.1) | ＜0.001 |
| Systolic blood pressure, mmHg | 128 (0.4) | 130 (0.5) | 132 (0.5) | 134 (0.5) | ＜0.001 |
| Diastolic blood pressure, mmHg | 76 (0.3) | 78 (0.3) | 79 (0.3) | 81 (0.3) | ＜0.001 |
| Antihypertensive medication use, % | 21 | 22 | 27 | 24 | ＜0.001 |
| Serum non-HDL cholesterol, mmol/L | 3.27 (0.02) | 3.41 (0.02) | 3.46 (0.02) | 3.55 (0.02) | ＜0.001 |
| Serum HDL cholesterol, mmol/L | 1.40 (0.01) | 1.37 (0.01) | 1.33 (0.01) | 1.32 (0.01) | ＜0.001 |
| Serum triglycerides, mmol/L | 1.03 (0.02) | 1.19 (0.02) | 1.32 (0.02) | 1.49 (0.02) | ＜0.001 |
| Diabetes mellitus, % | 4 | 6 | 7 | 7 | ＜0.001 |
| Current smokers, % | 15 | 24 | 32 | 56 | ＜0.001 |
| Ethanol intake, g/day | 8.5 (0.4) | 10.2 (0.4) | 11.8 (0.4) | 14.6 (0.4) | ＜0.001 |
Values were presented as means (standard errors) or proportions, adjusted for age, sex, and community.
During a median follow-up of 21 years totaling 99,206 person-years, there were 327 incident strokes, including 105 hemorrhagic strokes, 219 ischemic strokes (113 lacunar infarctions, and 106 non-lacunar infarctions), and 3 unclassified strokes, and 130 incident coronary heart diseases.
Table 2 shows the associations of cumulative average leukocyte count with the risks of stroke subtypes, coronary heart disease, and total cardiovascular disease. After adjustments for age, sex, community, and updated conventional cardiovascular risk factors, the multivariable hazard ratio (95% CI) for the highest versus lowest quartile of leukocyte count was 1.50 (1.08–2.08); P for trend=0.01 for ischemic stroke, 2.17 (1.33–3.55); P for trend ＜0.001 for coronary heart disease, and 1.40 (1.11–1.76); P for trend ＜0.001 for total cardiovascular disease. The multivariable hazard ratio (95% CI) for 1 SD increment of leukocyte count was 1.14 (1.03–1.27) for ischemic stroke, 1.30 (1.13–1.49) for coronary heart disease, and 1.13 (1.05–1.22) for total cardiovascular disease. There were no significant differences pertaining to sex between leukocyte count and the risks of ischemic stroke ( P for interaction=0.74) and coronary heart disease ( P for interaction=0.45) after multivariable adjustments. The positive association with the risk of ischemic stroke was similarly observed for both lacunar infarction and non-lacunar infarction, whereas the trend in non-lacunar infarction was of borderline statistical significance. The multivariable hazard ratio (95% CI) for the highest versus lowest quartile of leukocyte count was 1.59 (1.00–2.51); P for trend=0.04 for lacunar infarction, 1.42 (0.90–2.26); P for trend=0.09 for non-lacunar infarction. The multivariable hazard ratio (95% CI) for 1 SD increment of leukocyte count was 1.17 (1.01–1.36) for lacunar infarction 1.12 (0.96–1.31) for non-lacunar infarction. A similar association was not observed between leukocyte count and the risk of hemorrhagic stroke.
Table 2. Hazard ratios (HRs, 95% CIs) of cardiovascular diseases according to quartiles of cumulative average leukocyte count.
| Leukocyte count quartiles (× 10 2 cells/mm 3 ) | P for trend | 1 SD increment ＊ | ||||
|---|---|---|---|---|---|---|
| Q1 (low) | Q2 | Q3 | Q4 (high) | |||
| No. at risk | 1,310 | 1,311 | 1,308 | 1,313 | ||
| Person-years | 24,760 | 25,191 | 24,580 | 24,675 | ||
| Total stroke | ||||||
| No. of events | 69 | 66 | 98 | 94 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.12 (0.87, 1.43) | 1.41 (1.11, 1.80) | 1.45 (1.13, 1.87) | ＜0.001 | 1.15 (1.05, 1.25) |
| Multivariable HR (95% CI) | 1.00 | 1.02 (0.79, 1.31) | 1.17 (0.91, 1.50) | 1.20 (0.93, 1.56) | 0.08 | 1.06 (0.97, 1.16) |
| Hemorrhagic stroke | ||||||
| No. of events | 29 | 21 | 30 | 25 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 0.86 (0.57, 1.29) | 1.14 (0.77, 1.69) | 0.87 (0.55, 1.36) | 0.91 | 0.94 (0.80, 1.10) |
| Multivariable HR (95% CI) | 1.00 | 0.76 (0.50, 1.15) | 0.95 (0.63, 1.43) | 0.68 (0.43, 1.08) | 0.21 | 0.86 (0.73, 1.01) |
| Ischemic stroke | ||||||
| No. of events | 40 | 45 | 67 | 67 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.31 (0.95, 1.79) | 1.60 (1.17, 2.18) | 1.78 (1.30, 2.44) | ＜0.001 | 1.22 (1.10, 1.35) |
| Multivariable HR (95% CI) | 1.00 | 1.20 (0.87, 1.65) | 1.32 (0.96, 1.81) | 1.50 (1.08, 2.08) | 0.01 | 1.14 (1.03, 1.27) |
| Lacunar infarction | ||||||
| No. of events | 19 | 27 | 31 | 36 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.41 (0.91, 2.20) | 1.55 (0.99, 2.43) | 1.95 (1.25, 3.05) | 0.002 | 1.25 (1.09, 1.44) |
| Multivariable HR (95% CI) | 1.00 | 1.26 (0.81, 1.98) | 1.25 (0.79, 1.99) | 1.59 (1.00, 2.51) | 0.04 | 1.17 (1.01, 1.36) |
| Non-lacunar infarction | ||||||
| No. of events | 21 | 18 | 36 | 31 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.20 (0.76, 1.88) | 1.63 (1.06, 2.50) | 1.61 (1.03, 2.52) | 0.02 | 1.19 (1.03, 1.37) |
| Multivariable HR (95% CI) | 1.00 | 1.14 (0.72, 1.79) | 1.38 (0.90, 2.15) | 1.42 (0.90, 2.26) | 0.09 | 1.12 (0.96, 1.31) |
| Coronary heart disease | ||||||
| No. of events | 13 | 22 | 44 | 51 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.56 (0.93, 2.61) | 2.51 (1.56, 4.05) | 2.96 (1.84, 4.76) | ＜0.001 | 1.43 (1.26, 1.62) |
| Multivariable HR (95% CI) | 1.00 | 1.41 (0.84, 2.37) | 1.95 (1.19, 3.18) | 2.17 (1.33, 3.55) | ＜0.001 | 1.30 (1.13, 1.49) |
| Total cardiovascular disease | ||||||
| No. of events | 79 | 85 | 138 | 139 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.17 (0.94, 1.47) | 1.61 (1.30, 2.00) | 1.74 (1.40, 2.18) | ＜0.001 | 1.23 (1.14, 1.32) |
| Multivariable HR (95% CI) | 1.00 | 1.07 (0.85, 1.35) | 1.32 (1.06, 1.65) | 1.40 (1.11, 1.76) | ＜0.001 | 1.13 (1.05, 1.22) |
＊ 1 SD increment of leukocyte count was 16.4×10 2 cells/mm 3 .
Q1: 18–52×10 2 cells/mm 3 , Q2: 53–62×10 2 cells/mm 3 , Q3: 63–74×10 2 cells/mm 3 , and Q4: 75–130×10 2 cells/mm 3 .
Multivariable hazard ratios were further adjusted for updated body mass index, systolic blood pressure, cigarette smoking status, ethanol intake, serum non-HDL and HDL cholesterol levels, serum triglycerides, diabetes mellitus, and antihypertensive medication use.
Table 3 shows the smoking status-stratified associations of cumulative average leukocyte count with the risks of ischemic stroke and coronary heart disease according to updated smoking status. In never smokers, the multivariable hazard ratio (95% CI) for the highest versus lowest quartile of leukocyte count was 1.17 (0.75–1.82); P for trend=0.53 for ischemic stroke, 1.78 (0.83–3.82); P for trend=0.14 for coronary heart disease. In former smokers, the corresponding multivariable hazard ratio (95% CI) was 2.42 (1.07–5.46); P for trend=0.03 for ischemic stroke, 1.55 (0.58–4.15); P for trend=0.35 for coronary heart disease. In current smokers, the corresponding multivariable hazard ratio (95% CI) was 2.45 (1.11–5.38); P for trend=0.05 for ischemic stroke, 2.73 (1.37–5.44); P for trend=0.009 for coronary heart disease. The multivariable hazard ratio (95% CI) for 1 SD increment of leukocyte count was 1.03 (0.88–1.21) for ischemic stroke, 1.23 (0.95–1.59) for coronary heart disease in never smokers, 1.20 (0.97–1.49), 1.13 (0.87–1.47) in former smokers, and 1.26 (1.03–1.54), 1.36 (1.09–1.71) in current smokers.
Table 3. Hazard ratios (HRs, 95% CIs) of ischemic stroke and coronary heart disease according to quartiles of cumulative average leukocyte count, stratified by updated smoking status.
| Leukocyte count quartiles (×10 2 cells/mm 3 ) | P for trend | 1 SD increment ＊ | ||||
|---|---|---|---|---|---|---|
| Q1 (low) | Q2 | Q3 | Q4 (high) | |||
| Never smokers | ||||||
| No. at risk | 953 | 899 | 812 | 666 | ||
| Person-years | 18,466 | 17,785 | 16,052 | 13,177 | ||
| Ischemic stroke | ||||||
| No. of events | 28 | 28 | 29 | 25 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.05 (0.71, 1.56) | 1.12 (0.74, 1.69) | 1.53 (1.00, 2.35) | 0.05 | 1.14 (0.99, 1.33) |
| Multivariable HR (95% CI) | 1.00 | 0.94 (0.63, 1.40) | 0.92 (0.60, 1.41) | 1.17 (0.75, 1.82) | 0.53 | 1.03 (0.88, 1.21) |
| Coronary heart disease | ||||||
| No. of events | 9 | 13 | 12 | 11 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.53 (0.74, 3.13) | 1.76 (0.86, 3.61) | 2.48 (1.19, 5.17) | 0.01 | 1.37 (1.08, 1.73) |
| Multivariable HR (95% CI) | 1.00 | 1.49 (0.71, 3.10) | 1.42 (0.68, 3.00) | 1.78 (0.83, 3.82) | 0.14 | 1.23 (0.95, 1.59) |
| Former smokers | ||||||
| No. at risk | 172 | 158 | 158 | 130 | ||
| Person-years | 3,134 | 2,892 | 2,797 | 2,264 | ||
| Ischemic stroke | ||||||
| No. of events | 6 | 8 | 13 | 9 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.72 (0.77, 3.81) | 2.33 (1.08, 5.00) | 2.56 (1.16, 5.63) | 0.01 | 1.28 (1.05, 1.57) |
| Multivariable HR (95% CI) | 1.00 | 1.71 (0.75, 3.90) | 1.99 (0.90, 4.43) | 2.42 (1.07, 5.46) | 0.03 | 1.20 (0.97, 1.49) |
| Coronary heart disease | ||||||
| No. of events | 3 | 5 | 8 | 9 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 1.21 (0.48, 3.08) | 1.79 (0.73, 4.36) | 1.93 (0.77, 4.83) | 0.11 | 1.22 (0.96, 1.55) |
| Multivariable HR (95% CI) | 1.00 | 1.19 (0.44, 3.24) | 1.61 (0.61, 4.19) | 1.55 (0.58, 4.15) | 0.35 | 1.13 (0.87, 1.47) |
| Current smokers | ||||||
| No. at risk | 185 | 254 | 338 | 517 | ||
| Person-years | 3,160 | 4,514 | 5,731 | 9,234 | ||
| Ischemic stroke | ||||||
| No. of events | 6 | 9 | 25 | 33 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 2.14 (0.95, 4.82) | 2.87 (1.32, 6.22) | 2.58 (1.20, 5.57) | 0.04 | 1.28 (1.06-1.55) |
| Multivariable HR (95% CI) | 1.00 | 1.95 (0.86, 4.43) | 2.60 (1.18, 5.74) | 2.45 (1.11, 5.38) | 0.05 | 1.26 (1.03-1.54) |
| Coronary heart disease | ||||||
| No. of events | 1 | 4 | 24 | 31 | ||
| Age-, sex-, and community-adjusted HR (95% CI) | 1.00 | 3.18 (1.58, 6.37) | 3.20 (1.63, 6.26) | 0.001 | 1.43 (1.15, 1.76) | |
| Multivariable HR (95% CI) | 1.00 | 2.75 (1.35, 5.62) | 2.73 (1.37, 5.44) | 0.009 | 1.36 (1.09, 1.71) | |
＊ 1 SD increment of leukocyte count was 16.4×10 2 cells/mm 3 .
Q1: 18–52×10 2 cells/mm 3 , Q2: 53–62×10 2 cells/mm 3 , Q3: 63–74×10 2 cells/mm 3 , and Q4: 75–130×10 2 cells/mm 3 .
Multivariable hazard ratios were further adjusted for updated body mass index, systolic blood pressure, ethanol intake, serum non-HDL and HDL cholesterol levels, serum triglyceride, diabetes mellitus, and antihypertensive medication use.
Discussion
Leukocyte count was positively associated with the risks of ischemic stroke and coronary heart disease among the general Japanese population. The positive association with the risk of ischemic stroke was similarly observed in both lacunar and non-lacunar infarctions, although the trend in non-lacunar infarction was of borderline statistical significance. In smoking status-stratified analyses, positive associations with the risks of ischemic stroke and coronary heart disease were more pronounced in current smokers, and similar associations of borderline statistical significance were observed in former and never smokers.
The positive associations of leukocyte count with the risks of ischemic stroke and coronary heart disease in this study are consistent with previous findings. The Atherosclerosis Risk in Communities (ARIC) Study of 13,555 African–Americans and Caucasian aged 45–64 years with a follow-up of 8 years reported that the multivariable hazard ratio (95% CI) of ≥ 70×10 2 cells/mm 3 versus ＜48×10 2 cells/mm 3 leukocytes was 2.00 (1.29–2.99); P for trend=0.002 for ischemic stroke, 1.80 (1.32–2.43); P for trend ＜0.001 for coronary heart disease 2) . Another ARIC study of 14,448 Americans aged 45–64 years with a follow-up of 13.4 years further investigated the associations between 1 SD increment (20×10 2 cells/mm 3 ) of leukocyte count and the risks of ischemic stroke subtypes; the multivariable hazard ratio (95% CI) was 1.14 (1.02–1.28) for lacunar infarction and 1.07 (0.99–1.15) for non-lacunar infarction 4) . The Honolulu Heart Program of 3,342 Japanese–American men aged 71–93 years with a follow-up of 8 years reported the multivariable hazard ratio (95% CI) of 72–640×10 2 cells/mm 3 versus 9–50×10 2 cells/mm 3 leukocytes was 1.69 (0.99–2.89); P for trend=0.04 for ischemic stroke 5) . The Women’s Health Initiative of 74,375 American postmenopausal women aged 50–79 years with a follow-up of 16 years reported that the multivariable hazard ratio (95% CI) of the updated leukocyte count at the baseline and 3-years follow-up (≥ 80×10 2 cells/mm 3 versus 53–56.99×10 2 cells/mm 3 [the fifth decile]) was 2.06 (1.60–2.65); P for trend ＜0.001 for coronary heart disease mortality 6) . However, a Japanese prospective cohort study of 4,492 male employees aged 40–59 years with a follow-up of 9 years and the Malmö Preventive Project of 20,063 Sweden men and women aged 27–61 years with a follow-up of 24 years reported no significant association of leukocyte count with the risk of ischemic stroke 8 , 9) .
Cigarette smoking has a long-term negative impact on leukocyte count 10) . The ARIC study reported that the multivariable hazard ratio (95% CI) of ≥ 70×10 2 cells/mm 3 versus ＜48×10 2 cells/mm 3 leukocytes for ischemic stroke was 1.59 (0.85–2.30); P for trend=0.10 in never smokers, 2.13 (0.97–4.69); P for trend=0.05 in former smokers, and 3.15 (1.07–9.23); P for trend=0.07 in current smokers. The corresponding multivariable hazard ratio (95% CI) for coronary heart disease was 1.73 (1.08–2.79); P for trend=0.03, 2.01 (1.16–3.47); P for trend=0.02, and 1.44 (0.76–2.73); P for trend=0.06, respectively 2) . The Malmö Diet and Cancer Study of 26,972 Sweden men and women aged 45–73 years with a follow-up of 13.6 years reported that the multivariable hazard ratio (95% CI) of ≥ 73×10 2 cells/mm 3 versus ＜52×10 2 cells/mm 3 leukocytes for ischemic stroke was 1.1 (0.8–1.4); P for trend=0.94 in never smokers, 1.5 (1.1–1.9); P for trend=0.01 in former smokers, 1.7 (1.2–2.5); P for trend ＜0.001 in current smokers 3) . The Japanese prospective cohort study reported that the multivariable hazard ratio (95% CI) of ≥ 60×10 2 cells/mm 3 versus ＜60×10 2 cells/mm 3 leukocytes for myocardial infarction was 3.5 (1.0–1.2) in non-current smokers, 2.6 (1.0–1.2) in current smokers 8) . The Malmö Preventive Project reported that the multivariable hazard ratio (95% CI) of ≥ 70×10 2 cells/mm 3 versus ＜47×10 2 cells/mm 3 leukocytes for coronary heart disease was 1.38 (1.04–1.82); P for trend=0.22 in never smokers, 1.68 (1.24–2.28); P for trend=0.004 in former smokers, 1.30 (1.09–1.54); P for trend ＜0.001 in current smokers 9) .
The potential mechanisms underlying the associations of leukocyte count with the risks of ischemic stroke and coronary heart disease have been explored. Leukocyte count is a marker of systemic inflammation, which is considered to play an important role in the risk of ischemic cardiovascular disease. Leukocytes as a source of thrombogenic stimuli could enhance atherogenesis and lead to atherosclerotic plaque destabilization 18 , 19) . Pro-inflammatory leukocytes adhere by adhesion molecules to endothelial cells and enter the vascular intima. Monocytes differentiate into macrophages and concentrate the per-oxidized lipids to become foam cells, which are the prototypical cells in the atherosclerotic plaque 18) . Neutrophils attract molecules to promote foam cell formation and aggravate atherosclerotic lesions 20) . Cigarette smoking is strongly associated with increased leukocyte count and could intermediate the association with the risk of ischemic cardiovascular disease 10 , 21) . An early CIRCS report of 557 men aged 60–75 years found that the association of leukocyte count with mean carotid arteries intima-media thickness (IMT) had no significant difference between current and non-current smokers ( P for interaction=0.65), and parts of lymphocyte subpopulations (CD4＋CD45RO＋T cells and CD19＋CD80＋B cells) were positively associated with mean IMT in non-current smokers 22) .
The strength of our study lied in the large population-based study design with a 21-year follow-up. Also, we used updated leukocyte count and confounding variables to examine the associations, which could help reduce misclassification of study exposure. As for study limitations, the CIRCS did not measure C-reactive protein at the 1991–2000 baseline, which was another important inflammatory marker and is positively associated with the risks of ischemic stroke, coronary heart disease, and recurrent stroke among the general Japanese population 23 - 25) . However, subgroup analyses from the Women’s Health Initiative Study found that the association of leukocyte count with cardiovascular disease mortality did not change materially after further adjustment for C-reactive protein 6) .
In conclusion, leukocyte count was positively associated with the risks of ischemic stroke and coronary heart disease among the general Japanese population, especially in current smokers.
Appendix The CIRCS Investigators
The CIRCS Investigators are Takeo Okada, Yuji Shimizu, Yasuhiko Kubota, Shinichi Sato, Mina Hayama-Terada and Masahiko Kiyama, Osaka Center for Cancer and Cardiovascular Disease Prevention; Hironori Imano, Renzhe Cui, Isao Muraki, Akihiko Kitamura, Hiroshige Jinnouchi, Mizuki Sata, and Hiroyasu Iso, Osaka University; Kazumasa Yamagishi, Mitsumasa Umesawa and Tomoko Sankai, University of Tsukuba; Koutatsu Maruyama, Ehime University; Ai Ikeda and Takeshi Tanigawa, Juntendo University, and Masanori Nagao and Tetsuya Ohira, Fukushima Medical University.
Acknowledgements
We thank our colleagues from Osaka University Center of Medical Data Science, Advanced Clinical Epidemiology Investigator’s Research Project for providing their insight and expertise for our research.
Sources of Funding
This study was supported by Grant-in-Aid for Co-operative Research (A) [grant number 04304036], Scientific Research B [grant numbers 08457125, 10470103 and 19H03901], and Challenging Research (Exploratory) [grant number 17K19810] from the Japan Society for the Promotion of Science.
Competing Interest
None declared.
References
- 1).Libby P. Inflammation in atherosclerosis. Nature, 2002; 420: 868-874 [DOI] [PubMed] [Google Scholar]
- 2).Lee CD, Folsom AR, Nieto FJ, Chambless LE, Shahar E, Wolfe DA. White blood cell count and incidence of coronary heart disease and ischemic stroke and mortality from cardiovascular disease in African-American and White men and women: atherosclerosis risk in communities study. Am J Epidemiol, 2001; 154: 758-764 [DOI] [PubMed] [Google Scholar]
- 3).Zia E, Melander O, Björkbacka H, Hedblad B, Engström G. Total and differential leucocyte counts in relation to incidence of stroke subtypes and mortality: a prospective cohort study. J Intern Med, 2012; 272: 298-304 [DOI] [PubMed] [Google Scholar]
- 4).Ohira T, Shahar E, Chambless LE, Rosamond WD, Mosley TH Jr, Folsom AR. Risk factors for ischemic stroke subtypes: the Atherosclerosis Risk in Communities study. Stroke, 2006; 37: 2493-2498 [DOI] [PubMed] [Google Scholar]
- 5).Huh JY, Ross GW, Chen R, Abbott RD, Bell C, Willcox B, Launer L, Petrovitch H, Kaya B, Masaki K. Total and differential white blood cell counts in late life predict 8-year incident stroke: the Honolulu Heart Program. J Am Geriatr Soc, 2015; 63: 439-446 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6).Kabat GC, Kim MY, Manson JE, Lessin L, Lin J, Wassertheil-Smoller S, Rohan TE. White Blood Cell Count and Total and Cause-Specific Mortality in the Women’s Health Initiative. Am J Epidemiol, 2017; 186: 63-72 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7).Welsh C, Welsh P, Mark PB, Celis-Morales CA, Lewsey J, Gray SR, Lyall DM, Iliodromiti S, Gill JMR, Pell J, Jhund PS, Sattar N. Association of Total and Differential Leukocyte Counts With Cardiovascular Disease and Mortality in the UK Biobank. Arterioscler Thromb Vasc Biol, 2018; 38: 1415-1423 [DOI] [PubMed] [Google Scholar]
- 8).Imano H, Sato S, Kitamura A, Kiyama M, Ohira T, Shimamoto T, Iso H. Leukocyte count is an independent predictor for risk of acute myocardial infarction in middle-aged Japanese men. Atherosclerosis, 2007; 195: 147-152 [DOI] [PubMed] [Google Scholar]
- 9).Li C, Engström G, Hedblad B. Leukocyte count is associated with incidence of coronary events, but not with stroke: a prospective cohort study. Atherosclerosis, 2010; 209: 545-550 [DOI] [PubMed] [Google Scholar]
- 10).Nieto FJ, Szklo M, Folsom AR, Rock R, Mercuri M. Leukocyte count correlates in middle-aged adults: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Epidemiol, 1992; 136: 525-537 [DOI] [PubMed] [Google Scholar]
- 11).Imano H, Kitamura A, Sato S, Kiyama M, Ohira T, Yamagishi K, Noda H, Tanigawa T, Iso H, Shimamoto T. Trends for blood pressure and its contribution to stroke incidence in the middle-aged Japanese population: The Circulatory Risk in Communities Study (CIRCS). Stroke, 2009; 40: 1571-1577 [DOI] [PubMed] [Google Scholar]
- 12).Yamagishi K, Muraki I, Kubota Y, Hayama-Terada M, Imano H, Cui R, Umesawa M, Shimizu Y, Sankai T, Okada T, Sato S, Kitamura A, Kiyama M, Iso H. The Circulatory Risk in Communities Study (CIRCS): A Long-Term Epidemiological Study for Lifestyle-Related Disease Among Japanese Men and Women Living in Communities. J Epidemiol, 2019; 29: 83-91 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13).Nakamura M, Iso H, Kitamura A, Imano H, Kiyama M, Yokoyama S, Kayamori Y, Koyama I, Nishimura K, Nakai M, Dasti M, Vesper HW, Teramoto T, Miyamoto Y. Total cholesterol performance of Abell-Levy-Brodie-Kendall reference measurement procedure: Certification of Japanese in-vitro diagnostic assay manufacturers through CDC’s Cholesterol Reference Method Laboratory Network. Clin Chim Acta, 2015; 445: 127-132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14).Nakamura M, Iso H, Kitamura A, Imano H, Noda H, Kiyama M, Sato S, Yamagishi K, Nishimura K, Nakai M, Vesper HW, Teramoto T, Miyamoto Y. Comparison between the triglycerides standardization of routine methods used in Japan and the chromotropic acid reference measurement procedure used by the CDC Lipid Standardization Programme. Ann Clin Biochem, 2016; 53: 632-639 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15).Kirkendall WM, Feinlieb M, Freis ED, Mark AL. Recommendations for human blood pressure determination by sphygmomanometers. Subcommittee of the AHA Postgraduate Education Committee. Circulation, 1980; 62: 1146A-1155A [PubMed] [Google Scholar]
- 16).Iso H, Rexrode K, Hennekens CH, Manson JE. Application of computer tomography-oriented criteria for stroke subsubtype classification in a prospective study. Ann Epidemiol, 2000; 10: 81-87 [DOI] [PubMed] [Google Scholar]
- 17).WHO Expert Committee. Arterial hypertension and ischemic heart disease, preventive aspect: WHO technical report series no. 231. Geneva: World Health Organization, 1962 [Google Scholar]
- 18).Libby P, Nahrendorf M, Swirski FK. Leukocytes Link Local and Systemic Inflammation in Ischemic Cardiovascular Disease: An Expanded “Cardiovascular Continuum.” J Am Coll Cardiol, 2016; 67: 1091-1103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19).Lafont A. Basic aspects of plaque vulnerability. Heart, 2003; 89: 1262-1267 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20).Döring Y, Soehnlein O, Weber C. Neutrophil Extracellular Traps in Atherosclerosis and Atherothrombosis. Circ Res, 2017; 120: 736-743 [DOI] [PubMed] [Google Scholar]
- 21).SahBandar IN, Ndhlovu LC, Saiki K, Kohorn LB, Peterson MM, D’Antoni ML, Shiramizu B, Shikuma CM, Chow DC. Relationship between Circulating Inflammatory Monocytes and Cardiovascular Disease Measures of Carotid Intimal Thickness. J Atheroscler Thromb, 2020; 27: 441-448 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22).Tanigawa T, Kitamura A, Yamagishi K, et al. Relationships of differential leukocyte and lymphocyte subpopulations with carotid atherosclerosis in elderly men. J Clin Immunol, 2003; 23: 469-476 [DOI] [PubMed] [Google Scholar]
- 23).Iso H, Noda H, Ikeda A, Yamagishi K, Inoue M, Iwasaki M, Tsugane S. The impact of C-reactive protein on risk of stroke, stroke subtypes, and ischemic heart disease in middle-aged Japanese: the Japan public health center-based study. J Atheroscler Thromb, 2012; 19: 756-766 [PubMed] [Google Scholar]
- 24).Chei CL, Yamagishi K, Kitamura A, Kiyama M, Imano H, Ohira T, Cui R, Tanigawa T, Sankai T, Ishikawa Y, Sato S, Iso H. C-reactive protein levels and risk of stroke and its subtype in Japanese: The Circulatory Risk in Communities Study (CIRCS). Atherosclerosis, 2011; 217: 187-193 [DOI] [PubMed] [Google Scholar]
- 25).Kitagawa K, Hosomi N, Nagai Y, Kagimura T, Ohtsuki T, Maruyama H, Origasa H, Minematsu K, Uchiyama S, Nakamura M, Matsumoto M; J-STARS collaborators. Cumulative Effects of LDL Cholesterol and CRP Levels on Recurrent Stroke and TIA. J Atheroscler Thromb, 2019; 26: 432-441 [DOI] [PMC free article] [PubMed] [Google Scholar]