Cigarette smoking is a causative factor for premature atherosclerosis [1, 2]. However, the exact mechanism of smoking-induced damage to the arterial wall and its relation to the atherosclerotic process is still largely obscure. The adverse effects of smoking on vascular wall structure has been historically evaluated by autopsy studies [3]. Clinically applicable diagnostic measurements such as ultrasound imaging are widely utilized in recent years [4, 5].

The B-mode ultrasonographic image for a normal carotid artery is characterized by two parallel echogenic lines separated by an echolucent (hypoechoic) space. The parallel echoes correspond to the adventitial and intimal layers of the arterial wall, and the intervening echolucent region represents the media [6], although there may be systematic discrepancies between sonographic and histological measurements of intimal and medial thickness [7, 8].

Prior reports have been mixed regarding the relations between cigarette smoking and ultrasound-measured carotid intima-medial thickness (IMT). Whereas some researchers reported no significant differences in IMT of the common carotid artery (CCA) or differences in progression of IMT between smokers and nonsmokers [9, 10], others reported greater CCA-IMT in smokers [11–13], and that smoking is associated with echogenic plaques [14]. In addition, it is suggested that some harmful effects of smoking on the vessel wall are sex-related [14–16]. However, no study has been performed to determine specifically the association between cigarette smoking and carotid morphological changes in terms of arterial wall components detected by ultrasonography and whether there are any sex differences on those effects.

Changes in arterial wall thickness indicate structural arterial changes resulting from arterial remodeling, most often due to the atherosclerotic process. CCA-IMT is conventionally used as a marker of subclinical atherosclerosis [17], although important limitations of this marker have been recognized [18]. The aim of this study was to evaluate the chronic association between cigarette smoking on carotid IMT and IMT components. The following hypotheses were tested among middle-age adults: (1) smoking-related carotid artery wall changes involve both intimal and medial thickness; (2) the smoking-related alteration in carotid arterial wall is partly reversible upon smoking cessation in former smokers; and (3) there are sex differences among these relations.

Prior study has demonstrated that smoking has a specific fibrogenic effect which causes intimal thickening [24]. Autopsy studies have provided strong evidence for the effects of smoking on aortic [25] and coronary atherosclerosis [26]. Furthermore, the degree of atherosclerosis correlates with the intensity of smoking (both with the average number of cigarettes smoked per day and the duration of smoking) [14]. These findings are also consistent with the evidence that cigarette smoking and duration of smoking are positively associated with markers of inflammation [27], which may be one of many mechanisms of intimal injury from smokers.

Pre-clinical atherosclerosis is thought to be mainly an intimal atherosclerotic process [28]; However, the manner in which the media changes during the progression of atherosclerosis is still unclear. Some researchers have suggested that media hypertrophy accompanies atherosclerosis progression [29–31], while others have suggested that there is a moderate attenuation of the media [32], or that the media is unaffected in type I to III lesions, but may be modified in very advanced stages of atherosclerosis [28]. Adding more complexity, there is an argument that IMT is not an ideal indicator for early atherosclerosis because it consists mainly of media. The current study provides evidence that the medial layer (most likely corresponding to echolucent layer) may not necessarily get thickened during the smoking-related atherosclerotic process, which may partially explain poor correlation of IMT with clinical endpoints such as coronary disease in some studies [18, 33].

The current study reveals that current cigarette smoking correlated to a thinner echolucent layer in women. This effect (p = .03) does not reach statistical significance if Bonferroni correction is considered. However, the opposite effects of smoking effects on two IMT components (thickening of echogenic layer and thinning of echolucent layer) resulted in null association between current smoking and overall IMT in women (p = 0.54). Why thinning echolucent layer occurs particularly among female smokers is not clear, but there may be three putative mechanisms underlying this pathological phenomenon. First, because the matrix macromolecules of the media are produced by smooth muscle cells (SMC), the thinned echolucent layer could be caused by media stretching without a compensatory increase in matrix protein synthesis. Second, the thinned echolucent layer may be caused by smoking-induced SMC apoptosis [34] and consequent medial atrophy. There was indirect supporting evidence that smoking is related to decreased arterial smooth muscle mass. For example, smoking is an important risk factor for the development of abdominal aortic aneurysms (AAA) [35, 36]. The histopathology of aortic aneurysms is dominated by a transmural degenerative process characterized by medial atrophy and a gradual thinning and weakening of the vessel wall [37]. Medial SMC density is significantly decreased in AAA compared to normal aorta, and SMC apoptosis contributes to medial degeneration [38]. Therefore, if the smoking-related association in the common carotid is similar to that in the abdominal aorta (both are elastic-type arteries), the thinning carotid echolucent layer found in our study may be attributed to the medial atrophy. Third, the specific fibrogenic effect of smoking may not only cause intimal thickening, but also cause inflammation and density change of the media, rendering it isoechoic with the intima and adventitia [39, 40], and thus an apparent thinner echolucent layer in ultrasonic image. This pathological change could help explain why cigarette smoking was associated with greater distensibility in spite of more severe and echogenic lesions in another study [14].

The use of IMT progression over time as the sole outcome (without including baseline IMT) in an epidemiological study may result in spurious conclusions if not interpreted properly. In the current study, after controlling for smoking pack-years, the progression rate of IMT for current and former smokers did not differ among both men and women (see Table 4), which is consistent with other studies [41]. Nevertheless, it is inappropriate to infer that the atherogenic effect of smoking is irreversible [41]. Rather, our data suggest that the adverse effects of smoking are partly reversible among former smokers. In comparison with current smokers with similar pack-year exposure, former smokers manifested lesser extent of intimal thickening among both men and women. Particularly, the echolucent-layer-thinning in association with smoking was lessened among the subgroup of former female smokers.

However, former smokers still manifest elevated atherosclerotic risk in comparison with never smokers. Prior reports from population-based epidemiologic studies examining relations between smoking cessation and carotid atherosclerosis support our observation. Time-since-quitting was related to the atherosclerotic score [42]. In the Atherosclerosis Risk in Community (ARIC) study, former smokers display a large risk reduction in the first two to three years after quitting but maintain a modest elevated CAD risk for many years compared with that among never smokers [41]. This biphasic decrease in cardiovascular risk following smoking cessation indicates that some adverse effects of smoking may be cumulative and irreversible [41]. In the current study, the average time since cessation among former smokers was 14 years (range 0–40 yr). The relatively small samples of former smoker (n = 117) limited the power to test the change of IMT and its components as a function of time since cessation.

There are at least two limitations in this study. First, only participants with IMT measurement 0.7 mm or greater were included; the participants who fulfilled this criterion may potentially possess higher levels of other risk factors relative to those not included (as shown in Table 1). Thus the results may not be generalized to the general population. Second, the relatively short follow-up period (3 years) and the small sample size (n = 413) may have limited the power for detecting correlations between smoking and carotid morphological changes. However, these two limitations were at least partially compensated by the precision of the IMT measurement in this study [19].

This study examined carotid arterial morphological associations (thickness of arterial wall and components) detected by ultrasonography with cigarette smoking and smoking cessation. The evidence provided may help motivate early modification of the smoking habit and thus delay or reverse the disease process [43]. Furthermore, smoking-related carotid arterial morphological associations appeared to differ by sex; women may be more susceptible to arterial damage caused by cigarette smoking. This is evidenced by simultaneous thickening of echogenic layer and thinning of echolucent layer among female current smokers. Overall IMT alone may not be a valid measure for detecting smoking-induced atherosclerosis, especially among women. However, the results need to be verified by larger longitudinal epidemiologic studies using ultrasonography.