Carotid Intima-Media Thickness and Plaque Characteristics as a Risk Factor for Stroke in Japanese Elderly Men
Background and Purpose— Few cohort studies have examined the association of carotid intima-media thickness (IMT) and plaque characteristics with the risk of stroke in apparently healthy persons. We examined the relationship of carotid IMT and the surface, morphology, and calcification of carotid plaques with the incidence of stroke among Japanese men.
Methods— Carotid IMT and plaque were evaluated bilaterally with ultrasonography in 1289 men aged 60 to 74 years without a previous stroke or coronary heart disease. In this cohort, the subsequent incidence of stroke was investigated.
Results— During the 4.5-year follow-up, 34 strokes occurred. The multivariate-adjusted relative risk (95% CI) for the highest versus lowest quartiles of maximum IMT of the common carotid artery (CCA; ≥1.07 versus ≤0.77 mm) was 3.0 (1.1 to 8.3) for stroke. The combination of CCA and internal carotid artery (ICA) wall thickness was a better predictor of the risk of stroke than was CCA wall thickness alone. Men with a plaque, defined as a focal wall thickness of ≥1.5 mm, in the ICA had a 3-fold higher risk of stroke than those without a plaque, and the plaque surface irregularity further increased the stroke risk. A significant excess risk of stroke was confined to men with an uncalcified plaque.
Conclusions— Increased IMT of the CCA and an uncalcified plaque in the ICA, as assessed by ultrasonography, are risk factors for stroke in elderly Japanese men.
Prospective population-based studies in Europe and the United States have documented that carotid intima-media thickness (IMT) and plaque are positively associated with the subsequent incidence of stroke.1–4 Furthermore, 2 prospective studies have investigated the association between echogenicity of plaques or carotid artery lesions and stroke risk in American populations; 1 study indicated that hypoechoic plaque was associated with increased risk of ischemic stroke,5 and the other study implied that acoustic shadowing on carotid artery lesions is predictive of ischemic stroke.6
In Asian countries, a previous follow-up study of Japanese patients with cardiovascular diseases or high-risk profiles showed a positive association between the severity of carotid plaque and the risk of ischemic stroke,7 but no prospective study has been conducted in a general population. The aim of this prospective study was to systematically examine the incidence of stroke in relation to carotid IMT and the surface, morphology, and calcification of carotid plaque in community-dwelling Japanese elderly men.
Subjects and Methods
Subjects comprised 1358 men aged 60 to 74 years who were all participants of the target ages in a cardiovascular risk survey between 1996 and 2000. Subjects lived in 1 urban and 2 rural communities in Japan (Minamitakayasu district in Yao City, an urban community in Osaka Prefecture, 278 miles west of Tokyo, with a total census of 23 552 in 2000; Ikawa town, a rural community in Akita Prefecture, 280 miles northeast of Tokyo, with a total census of 6116; and Noichi town, a rural community in Kochi Prefecture, 393 miles southwest of Tokyo, with a total census of 16 595). Participants were recruited from all the residents >40 years old in each community for the purpose of primary prevention of cardiovascular diseases, and the number of participants were 420 in the urban community, 435 in the northeast rural community, and 503 in the southwest rural community. The respective participation rate for the census population within the target ages was 31%, 79%, and 45%. Individuals with a history of stroke (n=46) or coronary heart disease (n=23) were excluded, leaving 1289 men for the analysis.
Carotid arteries were evaluated with high-resolution B-mode ultrasonography. We adopted the same ultrasonography protocol used in one of the largest population-based studies of carotid atherosclerosis among elderly Americans [the Cardiovascular Health Study (CHS)].2,5 A single trained physician (A.K.) conducted the ultrasonographic scanning and interpreted the results. The imaging protocol involved obtaining a single longitudinal lateral view of the distal 10 mm of the right and left common carotid arteries (CCAs) and 3 longitudinal views (anterior-oblique, lateral, and posterior-oblique) of each internal carotid artery (ICA). The ICA was defined as including the carotid bulb, identified by loss of the parallel wall present in the CCA, and the 10-mm segment of the ICA distal to the tip of the flow divider that separates the internal and external carotid arteries. To quantify the degree of thickening of the carotid artery walls, we assessed the maximum IMT; maximum IMT of the CCA was defined as the single thickest wall of the near and far right or left walls of the CCA. Maximum IMT of the ICA was defined in the same way from the 3 scan views on the right and left ICA. Plaque in the ICA wall was defined as a focal wall thickness of ≥1.5 mm and was categorized according to several surface characteristics: smoothness (smooth, mildly irregular, markedly irregular, or ulcerated), morphology (homogeneous or heterogeneous), and density (hypodense, isodense, hyperdense, or calcified). In cases with multiple plaques, only the thickest plaque was evaluated. These definitions were similar to those used in the CHS.5
Cardiovascular risk factors were also measured at baseline examination, using the same protocol among the 3 communities. Systolic blood pressure (SBP) and fifth-phase diastolic blood pressure (DBP) in the right arm were measured by trained nurses using standard mercury sphygmomanometers in quietly seated participants after a ≥5- minute rest. When the first SBP was ≥140 mm Hg or the first DBP was ≥90 mm Hg, the measurement was repeated. The second reading was used in the analyses; otherwise, the first reading was used. Physical and biochemical examinations, ECG, assessment of smoking and drinking habits by interviews, and quality control procedures have been reported previously.8 Hypertension was defined as an SBP ≥140 mm Hg, DBP ≥90 mm Hg, or use of antihypertensive medication. ST-T abnormalities and left ventricular hypertrophy (LVH) on ECG were defined as Minnesota Codes 4-1 to 4-3 or 5-1 to 5-3 and 3-1 plus 5-1 to 5-3, respectively. Diabetes mellitus was defined as a fasting glucose level of ≥126 mg/dL, or a nonfasting glucose level of ≥200 mg/dL, or use of medication for diabetes.
Ascertainment of Incident Cardiovascular Diseases
Subjects were followed to determine the incidence of stroke occurring by the end of 2002. Clinical end points were assessed blinded to the ultrasound measurements. Follow-up was terminated when subjects withdrew from the cohort because of nonstroke death (n=78) or moved out from the communities (n=15). Subjects were censored from the follow-up analysis at the date of the withdrawal.
Stroke was defined as a constellation of neurological deficits, sudden or rapid in onset, lasting ≥24 hours, or until death. A diagnosis of embolic infarction was made when evidence of an embolic source was reported in the medical records and if imaging studies and a neurology consult supported the diagnosis. Classification of other stroke subtypes (large-artery occlusive infarction, lacunar infarction, intraparenchymal hemorrhage, and subarachnoid hemorrhage) was based on imaging studies.9 Nonembolic stroke cases without imaging studies available were categorized according to clinical signs and symptoms.9 The study protocol was approved by the human ethics review committee of Osaka Medical Center for Health Science and Promotion.
Differences in mean values and prevalence of baseline characteristics between men who developed stroke and men who remained free of stroke were tested by the t test and χ2 test, respectively. Incidence rates per 1000 person years for stroke were calculated by quartiles of maximum IMT of the CCA, maximum IMT of the ICA, and the combination of maximum IMT of the CCA and the ICA, each split at the fourth quartile. Kaplan–Meier method was used to examine cumulative stroke-free rates according to quartiles of carotid artery IMT, and differences in cumulative stroke-free rates by quartiles were tested using log-rank test. The Cox proportional-hazards regression model was used to estimate the relative risk of stroke relative to the lowest quartile of maximum IMT or relative to men having no plaque in the ICA, adjusting for age and other potential confounding factors. The potential confounding factors included age (years), SBP (mm Hg), antihypertensive medication use (yes or no), ST-T abnormalities (yes or no), body mass index (kg/m2), and community (dummy variables). All analyses were performed using a standard statistical package (SPSS/PC, version 11.0; SPSS Inc.).
During the mean follow-up period of 4.5 years, we identified 34 strokes. Twenty-five stroke cases were confirmed by imaging studies comprising 9 intraparenchymal hemorrhages, 10 lacunar infarctions, 4 large-artery occlusive infarctions, and 2 embolic infarctions. Nine stroke cases without imaging studies consisted of 6 probable lacunar infarctions with minor motor and sensory symptoms, 1 probable large-artery occlusive infarction with cortical signs, and 2 unclassified strokes.
Table 1 shows the baseline risk characteristics of incident stroke and those who remained free of stroke. The mean age, SBP and DBP levels, prevalence of hypertension, the mean maximum IMT of the ICA, and prevalence of plaque were significantly higher in men who developed stroke than in men who remained free of stroke. Prevalence of antihypertensive medication use, ST-T abnormalities and LVH on ECG, and the mean body mass index tended to be higher in men with incident stroke than in men without stroke, although the differences were not statistically significant. Adjustment for age did not alter these findings (data not shown).
Figure 1 shows the incidence rates per 1000 person years of stroke according to quartiles of maximum CCA or ICA IMT. Compared with men with the lowest quartile of CCA wall thickness, the incidence of stroke tended to be higher in men in the second or third quartile and was >4-fold higher in men in the highest quartile. The gradient of increased risk of stroke was less pronounced for ICA than for CCA wall thickness. The combined categories of wall thickness of the CCA and the ICA showed a more prominent risk gradient than each measure alone. The cumulative stroke-free rate declined more for the fourth quartile of CCA IMT than for the other 3 quartiles of CCA (Figure 2). A similar but weaker trend was observed for ICA IMT. The divergence of the stroke-free rate according to the combined categories was most evident.
Table 2 provides the relative risks of stroke by quartiles of maximum CCA or ICA wall thickness. The age-adjusted relative risk among men in the highest versus lowest quartiles of maximum CCA wall thickness was 3.5 for total stroke and 3.8 for ischemic stroke. Further adjustment for potential cardiovascular factors attenuated these associations slightly, but relative risk remained statistically significant for total stroke. In contrast, there was no significant association between the degree of maximum ICA IMT and the incidence of total or ischemic stroke. The combined category of maximum CCA and ICA wall thickness was a better predictor than was maximum CCA wall thickness alone.
The age-adjusted relative risks (95% CI) in men with a plaque (n=535) in the ICA compared with men without plaque (n=754) was 3.3 (1.5 to 7.1) for total stroke and 4.4 (1.6 to 12.2) for ischemic stroke. The respective multivariate relative risk was 3.2 (1.4 to 7.1) and 4.2 (1.5 to 11.8; not shown in the table).
The relative risks of stroke according to plaque characteristics in the ICA are shown in Table 3. Age-adjusted and multivariate-adjusted relative risks of total and ischemic stroke were higher with increasing surface irregularity. These excess risks were similar among men with homogeneous plaque and those with heterogeneous plaque. The excess risks were higher among men with an uncalcified plaque than among men with a calcified plaque.
The present study showed a positive association between carotid IMT, a measure of carotid and generalized atherosclerosis, and the incidence of stroke in Japanese male residents aged 60 to 74 years. Maximum IMT of the CCA and plaque in the ICA are strong predictors for risk of stroke. The combination of CCA and ICA wall thickness provided a better prediction for stroke than did either IMT variable alone.
Three population-based prospective studies in the United States and Netherlands reported a positive association between intima-media thickening in the CCA and the incidence of stroke.2,3,10 The CHS reported that the multivariate relative risk of stroke for the highest versus lowest quintiles of maximum IMT in the CCA (≥1.18 mm versus <0.87 mm) was 2.13 (95% CI, 1.38 to 3.28).2 The Atherosclerosis Risk in Communities (ARIC) Study showed that the multivariate relative risk of ischemic stroke for the highest versus lowest tertiles of mean IMT of the far wall of the CCA (≥0.70 mm versus <0.59 mm) was 2.69 (95% CI, 1.49 to 4.87) in men and 1.65 (95%CI, 0.85 to 3.19) in women.3 In the Rotterdam Study, the multivariate relative risk of total stroke for the highest tertile versus lowest tertile of mean IMT in the CCA (>0.84 mm versus <0.72 mm) was 2.42 (95% CI, 1.51 to 3.89).10
Because we used the same protocols for carotid ultrasonography as those used in the CHS, our findings could be compared with data from the CHS. The excess risk of stroke with CCA intima-media thickening tended to be greater in elderly Japanese than in elderly American counterparts; however, the differences in the range of age, gender, cut-off values of IMT, and probably distributions of stroke subtypes must be taken into consideration. Although the CHS did not describe the distribution of stroke subtypes, on the basis of results from previous population-based studies in the United States,11,12 we inferred that their strokes comprised a higher proportion of large-artery occlusive infarctions than lacunar infarctions and intraparenchymal hemorrhage. In contrast, our stroke cases comprised a higher proportion of definite or probable lacunar infarctions (47%) and intraparenchymal hemorrhage (26%) compared with large-artery occlusive infarctions (15%). Increased carotid IMT may be a strong marker for arteriosclerotic lesions in intracranial small-vessel arteries. Our findings are supported by the results of the Rotterdam Study, which showed that carotid plaques were associated with the risk of lacunar infarction.4
To our knowledge, this is the first prospective study for general populations to demonstrate significant associations of plaque surface irregularity with the subsequent risk of stroke, although the Medical Research Council (MRC) European Carotid Surgery Trial showed that patients with irregular plaque had a higher risk of stroke than did those with smooth plaque.13 We speculate that plaque irregularity in a carotid artery may be a marker of the severity of arteriosclerosis in intracranial small vessels as well as a source of carotid artery-to-brain artery embolism. This speculation is supported by the results of the CHS, which showed a positive association between carotid plaque surface irregularity and the prevalent lacunar infarction detected by MRI.14
Evidence on calcified plaque of the carotid artery and risk of stroke is sparse. A cross-sectional study of Finnish elderly men reported a lack of the association between carotid mineralizations and clinical cardiovascular disease.15 We did not have a significant excess risk of total or ischemic stroke associated with calcified plaque. The CHS also reported a nonsignificant relative risk of 6.93 (95% CI, 0.8 to 60.2) for incident stroke with calcified plaque.5 Plaque accompanied by calcification may be stable and therefore not increase the stroke risk so much. However, it is also possible that the statistical insignificance of the association was attributable to only low stroke incidence. On the other hand, there was a significant excess risk of total and ischemic strokes among men with a plaque free of calcification compared with men with no plaque. The ARIC study showed that carotid artery lesions without acoustic shadowing, after adjustment for several risk factors, were predictive of ischemic stroke in men,6 which supported our findings.
We did not evaluate the risk of stroke for echolucent uncalcified plaques, which have been reported to increase risk for stroke in clinical and epidemiological studies;5,16 the proportion of hypoechoic plaques in our sample was too low (1.2%). Another limitation of the present study is we did not examine the associations between carotid IMT and risks of stroke subtypes because of the small numbers of cases. A large follow-up study is necessary to clarify these relationships.
We conclude that wall thickening of the CCA and uncalcified plaque formation in the ICA, as assessed by ultrasonography, are positively associated with an increased risk of stroke in elderly Japanese men.
This study was supported by grants-in-aid research B (11877069 in 1998 to 1999 and 12470092 in 2000 to 2001) from the Ministry of Health, Welfare and Labor, Japan. We thank Professor Aaron R. Folsom, University of Minnesota, Minneapolis, for the valuable comments. We also thank Professor Daniel H. O’Leary from Tufts University School of Medicine, Boston, Mass, for providing training sessions for scanning and reading the carotid ultrasound.
- Received July 6, 2004.
- Revision received September 14, 2004.
- Accepted September 29, 2004.
Bots ML, Hoes AW, Koudstaal PJ, Hofman A, Grobbee DE. Common carotid intima-media thickness and risk of stroke and myocardial infarction: the Rotterdam Study. Circulation. 1997; 96: 1432–1437.
Chambless LE, Folsom AR, Clegg LX, Sharrett AR, Shahar E, Nieto FJ, Rosamond WD, Evans G. Carotid wall thickness is predictive of incident clinical stroke: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Epidemiol. 2000; 151: 478–487.
Hollander M, Bots ML, Del Sol AI, Koudstaal PJ, Witteman JCM, Grobbee DE, Hofman A, Breteler MMB. Carotid plaques increase the risk of stroke and subtypes of cerebral infarction in asymptomatic elderly: the Rotterdam study. Circulation. 2002; 105: 2872–2877.
Hunt KJ, Evans GW, Folsom AR, Sharrett AR, Chambless LE, Tegeler CH, Heiss G. Acoustic shadowing on B-mode ultrasound of the carotid artery predicts ischemic stroke: the Atherosclerosis Risk in Communities (ARIC) Study. Stroke. 2001; 32: 1120–1126.
Handa N, Matsumoto M, Maeda H, Hougaku H, Kamada T. Ischemic stroke events and carotid atherosclerosis. Results of the Osaka Follow-up Study for Ultrasonographic Assessment of Carotid Atherosclerosis (the OSACA Study). Stroke. 1995; 26: 1781–1786.
Hollander M, Hak AE, Koudstaal PJ, Bots ML, Grobbee DE, Hofman A, Witteman JC, Breteler MM. Comparison between measures of atherosclerosis and risk of stroke: the Rotterdam Study. Stroke. 2003; 34: 2367–2372.
Sudlow CL, Warlow CP. Comparable studies of the incidence of stroke and its pathological types: results from an international collaboration. International Stroke Incidence Collaboration. Stroke. 1997; 28: 491–499.
Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers DO. Ischemic stroke subtypes: a population-based study of incidence and risk factors. Stroke. 1999; 30: 2513–2516.
Manolio TA, Burke GL, O’Leary DH, Evans G, Beauchamp N, Knepper L, Ward B. Relationships of cerebral MRI findings to ultrasonographic carotid atherosclerosis in older adults: the Cardiovascular Health Study. CHS Collaborative Research Group. Arterioscler Thromb Vasc Biol. 1999; 19: 356–365.
Salonen R, Tervahauta M, Salonen JT, Pekkanen J, Nissinen A, Karvonen MJ. Ultrasonographic manifestations of common carotid atherosclerosis in elderly eastern Finnish men. Prevalence and associations with cardiovascular diseases and risk factors. Arterioscler Thromb. 1994; 14: 1631–1640.
Mathiesen EB, Bonaa KH, Joakimsen O. Echolucent plaques are associated with high risk of ischemic cerebrovascular events in carotid stenosis: the Tromso Study. Circulation. 2001; 103: 2171–2175.