2023 Volume 30 Issue 8 Pages 934-942
Aim: Epidemiological evidence regarding the relationship between fish and fatty acid intake and carotid intima–media thickness (CIMT) has been limited and inconsistent. The current cross-sectional study investigated this issue using baseline data from the Aidai Cohort Study.
Methods: Study subjects were 2024 Japanese men and women aged 34–88 years. Right and left CIMT were measured at the common carotid artery using an automated carotid ultrasonography device. Maximum CIMT was defined as the largest CIMT value in either the left or right common carotid artery. Carotid wall thickening was defined as a maximum CIMT value ＞1.0 mm.
Results: The prevalence of carotid wall thickening was 13.0%. In men, intake of n-3 polyunsaturated fatty acids (PUFA) was independently positively related to the prevalence of carotid wall thickening, while no associations were found between intake of fish and the other fatty acids and carotid wall thickening or maximum CIMT. In women, intake levels of fish, n-3 PUFA, eicosapentaenoic acid, docosahexaenoic acid, and arachidonic acid were independently inversely associated with carotid wall thickening and intake levels of fish, n-3 PUFA, α-linolenic acid, n-6 PUFA, and linoleic acid were independently inversely associated with the maximum CIMT. No significant relationships were found between intake of total fat, saturated fatty acids, or monounsaturated fatty acids and carotid wall thickening or maximum CIMT regardless of sex.
Conclusions: In women, higher intake of fish and n-3 and n-6 PUFA may be associated with a lower prevalence of carotid wall thickening and a decrease in maximum CIMT.
The traditional Mediterranean diet and traditional Japanese diet, which are low in saturated fatty acids and high in n-3 and n-6 polyunsaturated fatty acids (PUFA), meet the criteria for a protective effect against coronary heart disease1). A meta-analysis in 2020 demonstrated that higher fish consumption is associated with a lower coronary heart disease incidence and mortality2). Long-term prospective cohort studies consistently demonstrate an association between a higher intake of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), marine n-3 PUFA, or higher levels of EPA and DHA in the body and a lower risk of coronary heart disease, myocardial infarction, and cardiovascular mortality3). However, the role of the mainly plant-derived n-3 PUFA α-linolenic acid in atherosclerotic coronary vascular disease is less clear, and conflicting findings have been shown4). n-6 PUFA and saturated fatty acids do not increase the coronary vascular risk, as was previously suggested, and there is no proof that monounsaturated fatty acids have a preventative effect on coronary heart disease1).
Carotid intima–media thickness (CIMT) measurements may not be the best candidate predictor for the risk of coronary vascular disease in asymptomatic and low-to-intermediate risk individuals5). Epidemiological evidence regarding the relationship between fatty acid intake and CIMT has been limited and inconsistent6-10), and no epidemiological study has examined the association between fish intake and CIMT. Here, we conducted a cross-sectional study examining the association between both intake of fish and specific fatty acids and the carotid wall thickening and maximum CIMT in Japanese adults using baseline data from the Aidai Cohort Study (AICOS) conducted in Yawatahama, Uchiko, Seiyo, and Ainan.
AICOS is a prospective study11-16) that included a baseline survey in 2015, and it is still ongoing. The present cross-sectional study used data collected in Yawatahama City in 2015, Uchiko Town in 2016, and Seiyo City and Ainan Town in 2017, with total populations of nearly 36,000, 17,000, 38,000, and 22,000, respectively. These municipalities are four of the 20 municipalities in Ehime Prefecture on Shikoku Island, which is located south of Japan’s Main Island. In Yawatahama City in 2015, Uchiko Town in 2016, Seiyo City in 2017, and Ainan Town in 2017, 798, 347, 524, and 755 study subjects, respectively, participated in AICOS, and they were recruited from people who had undergone health checkups that were conducted by their municipality of residence or through one of several alternative recruiting procedures. These 2424 study subjects were 33–89 years old (894 men, 35–89 years old; 1530 women, 33–85 years old) provided written informed consent and completed the baseline questionnaire. We excluded 258 study subjects who had missing or incomplete data on the study factors and 142 study subjects with a self-reported history of cardiovascular disease, leaving 2024 study subjects for the present analyses (727 men aged 35–88 years and 1297 women aged 34–85 years). The ethics committee at the Ehime University Graduate School of Medicine approved the AICOS.
MeasurementsTrained laboratory technicians or medical doctors measured the right and left CIMT at the common carotid artery using an automated onscreen carotid ultrasonography device (GM-72P00A [The CardioHealth Station]; Panasonic Healthcare Co., Ltd., Ehime, Japan) within a limited timeframe. The excellent reproducibility and face validity of this automated device were shown by the UK Biobank17). The maximum CIMT was defined as the largest CIMT value in either the left or the right common carotid artery. Carotid wall thickening was defined as a maximum CIMT ＞1.0 mm according to the Japan Academy of Neurosonology and the Japan Society of Ultrasonics in Medicine guidelines18, 19).
Study subjects completed a self-administered questionnaire, and research technicians filled in missing or illogical data using telephone interviews. In the first part of the questionnaire, the following information was obtained: age; sex; smoking habits; alcohol drinking habits; physical activity; current use of antihypertensive, cholesterol-lowering, and diabetic medications; employment; and education. Leisure time physical activity was defined as present if the study subjects had engaged in at least 30 minutes of any type of moderate-to-vigorous physical activity such as brisk walking, playing golf, gardening, jogging, or playing tennis at least once a week. Two blood pressure measurements were taken using an automated sphygmomanometer, with the study subject in the sitting position after at least 5 minutes of rest; the second measurement was used for the present study. Hypertension was defined as systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or current use of antihypertensive medication. Blood samples were collected from an antecubital vein after an overnight fast. Serum concentrations of low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, plasma glucose concentrations, and hemoglobin A1c levels were measured at an external laboratory (Shikoku Chuken, Ehime, Japan). Dyslipidemia was defined as a serum low-density lipoprotein cholesterol concentration ≥ 140 mg/dL, high-density lipoprotein cholesterol concentration ＜40 mg/dL, triglyceride concentration ≥ 150 mg/dL, or current use of cholesterol-lowering medication. Diabetes mellitus was defined as a fasting plasma glucose level ≥ 126 mg/dL, hemoglobin A1c level ≥ 6.5%, or current use of diabetic medication. Body height and weight were measured in light clothing without shoes, and body mass index (BMI) was calculated as body weight in kilograms divided by height in meters squared. Waist circumference was measured at the umbilical level with the study subject in a standing position.
The second part of the questionnaire comprised a 169-item semiquantitative food-frequency questionnaire (FFQ) that included 520 food items. In the FFQ, study subjects were asked how often on average they consumed each of the food items listed and what was the usual serving size of each item during the previous year. Estimates of daily intake of food, energy, and selected nutrients were calculated based on the Standard Tables of Food Composition in Japan20). Fatty acid composition was evaluated using data published by Sasaki et al.21). Detailed information on the FFQ, including its validity and reproducibility, has been published previously22). Spearman correlation coefficients between the questionnaire and 12 daily diet records kept over a 1-year period for total fat, saturated fatty acids, monounsaturated fatty acids, and PUFA intake were 0.24, 0.47, 0.34, and 0.27, respectively, in men and 0.52, 0.54, 0.56, and 0.51, respectively, in women. Energy-adjusted intake calculated according to the residual method was used for the analyses23).
Statistical AnalysisAge, smoking status, alcohol consumption, leisure time physical activity, hypertension, dyslipidemia, diabetes mellitus, BMI, waist circumference, employment, and education were selected a priori as potential confounding factors. Age, BMI, and waist circumference were used as continuous variables. Multiple logistic regression analysis was performed to estimate the adjusted odds ratio (OR) and 95% confidence interval (CI) of carotid wall thickening according to the quartile of dietary factors under study, with the lowest quartile was used as the reference. Analysis of covariance was used to calculate adjusted mean of the maximum CIMT results according to the quartile of dietary factors under study allowing for confounding factors. A trend of an association was assessed using a multiple logistic regression model or multiple linear regression analysis assigning consecutive integers to the quartiles of dietary factors under study. A two-sided p value less than 0.05 was considered to be statistically significant. Because the distribution of the maximum CIMT was skewed to the right, natural logarithms of the values were used; therefore, the mean maximum CIMT values and their 95% CIs were geometric. All statistical analyses were performed using the SAS software version 9.4 (SAS Institute, Inc., Cary, NC, USA).
The maximum CIMT range was 0.423 to 1.925 mm among the 2024 study subjects. The median and 95th percentile values were 0.770 and 1.116 mm, respectively, and the geometric mean of the maximum CIMT was 0.783 (95% CI: 0.775–0.790) mm. The prevalence of carotid wall thickening was 13.0%. Fish intake was positively associated with age, diabetes mellitus, and waist circumference and inversely related to employment and education regardless of sex (Table 1). In men, fish intake was positively associated with BMI. In women, fish intake was positively associated with leisure time physical activity, hypertension, and dyslipidemia and inversely related to alcohol consumption.
| Menb | |||||
|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | Q4 | p for trendd | |
| n | 181 | 182 | 182 | 182 | |
| Age, years | 62.0 (51.0–68.0) | 64.0 (55.0–69.0) | 66.0 (58.0–71.0) | 68.0 (61.0–71.0) | ＜0.0001 |
| Smoking status | 0.19 | ||||
| Never | 55 (24.8) | 56 (30.8) | 59 (32.4) | 52 (28.6) | |
| Former | 89 (49.2) | 97 (53.3) | 101 (55.5) | 110 (60.4) | |
| Current | 37 (20.4) | 29 (15.9) | 22 (12.1) | 20 (11.0) | |
| Alcohol consumption | 0.06 | ||||
| Never | 23 (12.7) | 32 (17.6) | 30 (16.5) | 32 (17.6) | |
| Former | 6 (3.3) | 16 (8.8) | 15 (8.2) | 18 (9.9) | |
| Current | 152 (84.0) | 134 (73.6) | 137 (75.3) | 132 (72.5) | |
| Leisure time physical activity | 71 (39.2) | 72 (38.6) | 87 (47.8) | 78 (42.9) | 0.24 |
| Hypertension | 88 (48.6) | 96 (52.8) | 105 (57.7) | 101 (55.5) | 0.12 |
| Dyslipidemia | 81 (44.8) | 88 (48.4) | 98 (53.9) | 88 (48.4) | 0.33 |
| Diabetes mellitus | 13 (7.2) | 12 (6.6) | 28 (15.4) | 30 (16.5) | 0.0005 |
| Body mass index, kg/m2 | 23.8 (21.7–25.7) | 23.5 (22.0–26.0) | 23.8 (22.1–25.7) | 23.9 (22.2–26.4) | 0.02 |
| Waist circumference, cm | 83.7 (78.0–89.8) | 84.1 (78.3–90.0) | 84.6 (81.0–90.3) | 85.2 (80.0–91.0) | 0.0008 |
| Employment | 131 (72.4) | 130 (71.4) | 113 (62.1) | 112 (61.5) | 0.007 |
| Educatione | 0.008 | ||||
| Low | 19 (10.5) | 26 (14.3) | 20 (11.0) | 37 (20.3) | |
| Middle | 82 (45.3) | 87 (47.8) | 90 (49.5) | 88 (47.3) | |
| High | 80 (44.2) | 69 (37.9) | 72 (39.6) | 59 (32.4) | |
| Carotid wall thickening | 25 (13.8) | 37 (20.3) | 30 (16.5) | 47 (25.8) | 0.01 |
| Maximum carotid intima–media thickness, mm | 0.808 (0.731–0.885) | 0.808 (0.693–0.924) | 0.770 (0.693–0.924) | 0.828 (0.731–1.001) | 0.03 |
| Womenc | |||||
|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | Q4 | p for trendd | |
| n | 324 | 324 | 324 | 325 | |
| Age, years | 58.0 (48.0–67.0) | 62.0 (52.0–68.0) | 64.0 (57.0–69.0) | 66.0 (60.0–71.0) | ＜0.0001 |
| Smoking status | 0.23 | ||||
| Never | 286 (88.3) | 295 (91.1) | 288 (88.9) | 300 (92.3) | |
| Former | 23 (7.1) | 26 (8.0) | 31 (9.6) | 13 (4.0) | |
| Current | 15 (4.6) | 3 (0.9) | 5 (1.5) | 12 (3.7) | |
| Alcohol consumption | 0.01 | ||||
| Never | 172 (53.1) | 196 (60.5) | 172 (53.1) | 212 (65.2) | |
| Former | 15 (4.6) | 21 (6.5) | 21 (6.5) | 17 (5.2) | |
| Current | 137 (42.3) | 107 (33.0) | 131 (40.4) | 96 (29.5) | |
| Leisure time physical activity | 130 (40.1) | 147 (45.4) | 149 (46.0) | 156 (48.0) | 0.0497 |
| Hypertension | 94 (29.0) | 116 (35.8) | 125 (38.6) | 133 (40.9) | 0.001 |
| Dyslipidemia | 153 (47.2) | 172 (53.1) | 179 (55.3) | 198 (60.9) | 0.0005 |
| Diabetes mellitus | 16 (4.9) | 13 (4.0) | 18 (5.6) | 29 (8.9) | 0.02 |
| Body mass index, kg/m2 | 22.3 (20.2–25.0) | 22.3 (20.1–24.3) | 22.6 (20.5–24.4) | 23.1 (21.0–25.0) | 0.07 |
| Waist circumference, cm | 81.0 (73.6–88.6) | 80.0 (74.1–86.0) | 81.5 (75.2–87.7) | 82.0 (77.1–89.0) | 0.02 |
| Employment | 202 (62.4) | 173 (53.4) | 159 (49.1) | 150 (46.2) | ＜ 0.0001 |
| Educatione | ＜ 0.0001 | ||||
| Low | 37 (11.4) | 41 (12.7) | 38 (11.7) | 72 (22.2) | |
| Middle | 139 (42.9) | 148 (45.7) | 158 (48.8) | 162 (49.9) | |
| High | 148 (45.7) | 135 (41.7) | 128 (39.5) | 91 (28.0) | |
| Carotid wall thickening | 31 (9.6) | 31 (9.6) | 30 (9.3) | 33 (10.2) | |
| Maximum carotid intima–media thickness, mm | 0.731 (0.616–0.866) | 0.731 (0.654–0.885) | 0.770 (0.654–0.885) | 0.770 (0.693–0.885) | 0.02 |
a Values are medians (interquartile ranges) for continuous variables and numbers (percentages) of subjects for categorical variables.
b Fish intake range of each quartile was as follows: Q1 (＜71.27 g), Q2 (71.27–＜101.58 g), Q3 (101.58–＜132.64 g), Q4 (≥ 132.64 g).
c Fish intake range of each quartile was as follows: Q1 (＜62.53 g), Q2 (62.53–＜85.15 g), Q3 (85.15–＜113.34 g), Q4 (≥ 113.34 g).
d For continuous variables, a linear trend test was used; for categorical variables, a Mantel–Haenszel χ2 -test was used.
e Low, junior high school; medium, high school; high, junior college, vocational technical school or university.
Table 2 provides adjusted ORs and 95% CIs for carotid wall thickening and adjusted geometric means of the maximum CIMT according to quartiles of dietary intake of fish and specific types of fatty acids. In men, fish intake was not related to carotid wall thickening or maximum CIMT in the multivariate model. However, in women, an independent inverse exposure–response association was observed between fish intake and the prevalence of carotid wall thickening although the inverse relationship with the highest quartile of fish intake was not statistically significant (adjusted OR [95% CI] between extreme quartiles=0.58 [0.33–1.01], p for trend=0.04). Moreover, there was a significant inverse association between fish intake and the maximum CIMT: among those with lowest and highest quartiles, the multivariate-adjusted geometric means of the maximum CIMT were 0.775 and 0.753 mm, respectively (p for trend=0.03). In men, n-3 PUFA intake was independently associated with a higher prevalence of carotid wall thickening: the adjusted OR (95% CI, p for trend) between extreme quartiles was 1.97 (1.11–3.56, 0.02); however, no significant relationship was found between fish intake and maximum CIMT. Intake levels of α-linolenic acid, EPA, DHA, n-6 PUFA, linoleic acid, and arachidonic acid were not related to carotid wall thickening or maximum CIMT. In women, inverse exposure–response relationships were shown between intake levels of n-3 PUFA, EPA, DHA, and arachidonic acid and carotid wall thickening. However, only the adjusted OR between extreme quartiles of DHA, but not n-3 PUFA, EPA, or arachidonic acid, was statistically significant; the adjusted OR (95% CI, p for trend) between extreme quartiles was 0.57 (0.31–1.02, 0.04) for n-3 PUFA, 0.57 (0.32–1.00, 0.02) for EPA, 0.56 (0.32–0.96, 0.03) for DHA, and 0.61 (0.34–1.06, 0.045) for arachidonic acid. Intake levels of α-linolenic acid, n-6 PUFA, and linoleic acid were not related to carotid wall thickening. Intake levels of n-3 PUFA, α-linolenic acid, n-6 PUFA, and linoleic acid, but not EPA, DHA, or arachidonic acid, were significantly inversely associated with maximum CIMT.
| Men | Women | |||||||
|---|---|---|---|---|---|---|---|---|
| Quartile (Q) of intake | Median intake, g/day | Prevalence (%) | Adjusted OR (95% CI)a | Adjusted mean of maximum carotid intima–media thickness, mm (95% CI) a | Median intake, g/day | Prevalence (%) | Adjusted OR (95% CI) a | Adjusted mean of maximum carotid intima–media thickness, mm (95% CI) a |
| Fish | ||||||||
| Q1 | 49.4 | 25/181 (13.8) | 1.00 | 0.830 (0.806–0.854) | 45.9 | 31/324 (9.6) | 1.00 | 0.775 (0.760–0.790) |
| Q2 | 87.9 | 37/182 (20.3) | 1.35 (0.75–2.46) | 0.827 (0.803–0.851) | 74.7 | 31/324 (9.6) | 0.82 (0.46–1.46) | 0.769 (0.754–0.784) |
| Q3 | 115.1 | 30/182 (16.5) | 0.79 (0.42–1.48) | 0.785 (0.762–0.808) | 96.6 | 30/324 (9.3) | 0.68 (0.38–1.20) | 0.759 (0.745–0.774) |
| Q4 | 162.7 | 47/182 (25.8) | 1.41 (0.79–2.56) | 0.834 (0.810–0.858) | 142.4 | 33/325 (10.2) | 0.58 (0.33–1.01) | 0.753 (0.738–0.768) |
| p for trend | 0.53 | 0.61 | 0.04 | 0.03 | ||||
| n-3 Polyunsaturated fatty acids | ||||||||
| Q1 | 2.0 | 25/181 (13.8) | 1.00 | 0.819 (0.795–0.843) | 2.1 | 29/324 (9.0) | 1.00 | 0.775 (0.760–0.791) |
| Q2 | 2.7 | 29/182 (15.9) | 1.07 (0.58–2.00) | 0.802 (0.779–0.826) | 2.6 | 36/324 (11.1) | 1.09 (0.63–1.91) | 0.769 (0.754–0.784) |
| Q3 | 3.1 | 34/182 (18.7) | 1.11 (0.61–2.05) | 0.815 (0.792–0.839) | 3.0 | 32/324 (9.9) | 0.90 (0.51–1.59) | 0.759 (0.744–0.774) |
| Q4 | 3.8 | 51/182 (28.0) | 1.97 (1.11–3.56) | 0.838 (0.814–0.863) | 3.6 | 28/325 (8.6) | 0.57 (0.31–1.02) | 0.753 (0.738–0.768) |
| p for trend | 0.02 | 0.21 | 0.04 | 0.02 | ||||
| α-Linolenic acid | ||||||||
| Q1 | 1.4 | 35/181 (19.3) | 1.00 | 0.820 (0.796–0.844) | 1.5 | 35/324 (10.8) | 1.00 | 0.776 (0.761–0.791) |
| Q2 | 1.8 | 26/182 (14.3) | 0.62 (0.34–1.14) | 0.812 (0.788–0.835) | 1.8 | 39/324 (12.0) | 1.21 (0.72–2.04) | 0.775 (0.761–0.791) |
| Q3 | 2.1 | 36/182 (19.8) | 0.79 (0.44–1.39) | 0.811 (0.788–0.835) | 2.1 | 26/324 (8.0) | 0.78 (0.44–1.37) | 0.756 (0.742–0.771) |
| Q4 | 2.5 | 42/182 (23.1) | 1.18 (0.68–2.06) | 0.831 (0.807–0.856) | 2.5 | 25/325 (7.7) | 0.63 (0.35–1.12) | 0.749 (0.734–0.763) |
| p for trend | 0.41 | 0.54 | 0.054 | 0.003 | ||||
| Eicosapentaenoic acid | ||||||||
| Q1 | 0.1 | 21/181 (11.6) | 1.00 | 0.825 (0.801–0.850) | 0.1 | 30/324 (9.3) | 1.00 | 0.777 (0.762–0.793) |
| Q2 | 0.2 | 34/182 (18.7) | 1.47 (0.79–2.76) | 0.813 (0.790–0.837) | 0.2 | 33/324 (10.2) | 0.84 (0.48–1.48) | 0.767 (0.752–0.782) |
| Q3 | 0.3 | 37/182 (20.3) | 1.23 (0.66–2.33) | 0.806 (0.783–0.830) | 0.3 | 26/324 (8.0) | 0.53 (0.29–0.95) | 0.749 (0.735–0.764) |
| Q4 | 0.5 | 47/182 (25.8) | 1.72 (0.95–3.21) | 0.830 (0.806–0.854) | 0.4 | 36/325 (11.1) | 0.57 (0.32–1.00) | 0.763 (0.748–0.778) |
| p for trend | 0.13 | 0.88 | 0.02 | 0.09 | ||||
| Docosahexaenoic acid | ||||||||
| Q1 | 0.3 | 22/181 (12.2) | 1.00 | 0.825 (0.801–0.850) | 0.2 | 33/324 (10.2) | 1.00 | 0.775 (0.760–0.790) |
| Q2 | 0.4 | 30/182 (16.5) | 1.21 (0.65–2.30) | 0.804 (0.781–0.827) | 0.4 | 27/324 (8.3) | 0.69 (0.38–1.23) | 0.768 (0.753–0.783) |
| Q3 | 0.6 | 39/182 (21.4) | 1.39 (0.75–2.59) | 0.817 (0.794–0.841) | 0.5 | 28/324 (8.6) | 0.52 (0.29–0.93) | 0.748 (0.734–0.763) |
| Q4 | 0.9 | 48/182 (26.4) | 1.73 (0.96–3.19) | 0.828 (0.804–0.852) | 0.7 | 37/325 (11.4) | 0.56 (0.32–0.96) | 0.765 (0.750–0.780) |
| p for trend | 0.06 | 0.70 | 0.03 | 0.16 | ||||
| n-6 Polyunsaturated fatty acids | ||||||||
| Q1 | 9.8 | 32/181 (17.7) | 1.00 | 0.829 (0.805–0.854) | 10.6 | 38/324 (11.7) | 1.00 | 0.780 (0.765–0.795) |
| Q2 | 12.1 | 32/182 (17.6) | 0.87 (0.48–1.56) | 0.797 (0.774–0.820) | 12.4 | 31/324 (9.6) | 0.87 (0.51–1.49) | 0.765 (0.750–0.780) |
| Q3 | 14.0 | 30/182 (16.5) | 0.71 (0.39–1.30) | 0.809 (0.786–0.833) | 13.7 | 32/324 (9.9) | 1.01 (0.59–1.73) | 0.761 (0.747–0.776) |
| Q4 | 16.7 | 45/182 (24.7) | 1.45 (0.82–2.57) | 0.839 (0.815–0.864) | 15.9 | 24/325 (7.4) | 0.60 (0.33–1.05) | 0.750 (0.736–0.765) |
| p for trend | 0.27 | 0.45 | 0.13 | 0.008 | ||||
| Linoleic acid | ||||||||
| Q1 | 9.5 | 32/181 (17.7) | 1.00 | 0.827 (0.803–0.852) | 10.3 | 37/324 (11.4) | 1.00 | 0.779 (0.764–0.794) |
| Q2 | 11.8 | 33/182 (18.1) | 0.89 (0.50–1.60) | 0.800 (0.777–0.823) | 12.1 | 31/324 (9.6) | 0.91 (0.53–1.56) | 0.766 (0.752–0.782) |
| Q3 | 13.6 | 29/182 (15.9) | 0.68 (0.37–1.25) | 0.806 (0.783–0.830) | 13.4 | 33/324 (10.2) | 1.09 (0.64–1.86) | 0.763 (0.748–0.778) |
| Q4 | 16.3 | 45/182 (24.7) | 1.48 (0.84–2.64) | 0.842 (0.817–0.867) | 15.6 | 24/325 (7.4) | 0.62 (0.34–1.09) | 0.748 (0.734–0.763) |
| p for trend | 0.26 | 0.37 | 0.19 | 0.005 | ||||
| Arachidonic acid | ||||||||
| Q1 | 0.1 | 31/181 (17.1) | 1.00 | 0.826 (0.802–0.850) | 0.1 | 36/324 (11.1) | 1.00 | 0.774 (0.759–0.789) |
| Q2 | 0.2 | 30/182 (16.5) | 0.82 (0.45–1.48) | 0.804 (0.781–0.827) | 0.2 | 34/324 (10.5) | 0.96 (0.56–1.64) | 0.766 (0.752–0.782) |
| Q3 | 0.2 | 29/182 (15.9) | 0.76 (0.42–1.39) | 0.803 (0.780–0.826) | 0.2 | 28/324 (8.6) | 0.70 (0.40–1.21) | 0.759 (0.744–0.773) |
| Q4 | 0.2 | 49/182 (26.9) | 1.53 (0.88–2.69) | 0.842 (0.818–0.866) | 0.2 | 27/325 (8.3) | 0.61 (0.34–1.06) | 0.757 (0.743–0.772) |
| p for trend | 0.13 | 0.40 | 0.045 | 0.09 | ||||
a Adjustment for age, smoking status, alcohol consumption, leisure time physical activity, hypertension, dyslipidemia, diabetes mellitus, body mass index, waist circumference, employment, and education.
No significant relationships were found between intake of total fat, saturated fatty acids, or monounsaturated fatty acids and carotid wall thickening or maximum CIMT regardless of sex (data not shown).
To the best of our knowledge, this is the first study to show significant inverse associations between fish intake and the prevalence of carotid wall thickening and the maximum CIMT; however, such inverse associations were detected in Japanese women, but not men. Higher n-3 PUFA intake was independently associated with a higher prevalence of carotid wall thickening in men and a lower prevalence of carotid wall thickening in women; a significant inverse relationship between n-3 PUFA intake and the maximum CIMT was found in women although the association was not significant in men. Only in women, a higher intake of α-linolenic acid, EPA, DHA, n-6 PUFA, linoleic acid, and arachidonic acid was independently inversely associated with carotid wall thickening or the maximum CIMT.
Higher intake levels of n-3 PUFA, EPA, and DHA, but not saturated fatty acids, monounsaturated fatty acids, linolenic acid, or n-6 PUFA, were significantly inversely related to the mean CIMT in a cross-sectional study of 1902 Japanese adults aged 40 years or older9). Total linolenic acid intake was inversely associated with the CIMT of the internal and bifurcation segments of the carotid arteries but not with the common carotid artery in a cross-sectional study of 1575 white participants in the National Heart, Lung, and Blood Institute Family Heart Study, which was conducted in the USA, who had no coronary artery disease, stroke, hypertension, or diabetes mellitus10). A cross-sectional study of 686 Alaskan Eskimos over 35 years of age showed that intake levels of n-3 PUFA, C20–22 fish oil, palmitic acid, and stearic acid were inversely related to the mean CIMT8). In a cross-sectional study of 108 elderly Italian women, there were significant inverse associations between intake of PUFA and linoleic acid and the mean CIMT6). Intake levels of saturated fatty acids and trans fatty acids, but not monounsaturated fatty acids or PUFA, were significantly positively associated with the mean CIMT in a cross-sectional study of 620 people of Aboriginal, South Asian, Chinese, or European origin who were aged 35–75 years and lived in Canada7). For serum data, higher phospholipid proportions of oleic acid, α-linolenic acid, and DHA, but not linoleic acid or EPA, showed inverse associations with the mean CIMT in a Spanish cross-sectional study of 451 asymptomatic adults with primary dyslipidemia24). A cross-sectional study showed a significant inverse relationship between serum EPA plus DHA levels and mean CIMT among the 281 Japanese men (defined as born and living in Japan), but not among the 306 white (defined as white men born and living in the USA) or the 281 Japanese–American men (defined as Japanese men born and living in the USA) aged 40 to 49 years25). The findings are in partial agreement with our results.
Among women in the present study, after additional adjustment for EPA or DHA intake, the significant inverse exposure–response relationship between fish intake and carotid wall thickening completely disappeared; thus, the beneficial relationship may be partially ascribed to EPA or DHA in fish. The observed protective associations with intake of EPA and DHA might be due to their beneficial modulation of several known risk factors for cardiovascular disease, such as blood lipids, blood pressure, heart rate, heart rate variability, platelet aggregation, endothelial function, and inflammation3). The preventive relationship with intake of α-linolenic acid that was identified in this study might be ascribed to its cardioprotective effects via reducing plaque calcification, lowering lipids, maintaining endothelial function, and exhibiting antithrombotic, antiarrhythmic, and anti-inflammatory effects26). Higher linoleic acid intake might reduce LDL cholesterol and increase insulin sensitivity, and it might be associated with lower concentrations of C-reactive protein, interleukin (IL)-6, and IL-1β27). Among men in the present study, n-3 PUFA intake was positively associated with age, leisure time physical activity, diabetes mellitus, BMI, and waist circumference and inversely related to alcohol consumption and employment. Although statistical adjustments were made for BMI and waist circumference, part of the observed independent positive relationship between n-3 PUFA intake and carotid wall thickening may be ascribed to an uncontrolled confounding effect related to obesity. Alternatively, such a positive relationship in men may be a chance phenomenon. Higher intake of fish, n-3 PUFA, and n-6 PUFA may have a minor impact on atherosclerosis because men had more advanced atherosclerosis due to smoking, alcohol consumption, hypertension, and diabetes mellitus.
The strengths of the present study include the homogeneity of participants with respect to their residential area, use of an automated onscreen carotid ultrasound system, and the comprehensive assessment of potential confounding factors.
Some weaknesses of the present study should be clarified. The nature of the cross-sectional studies prevents conclusions from being drawn about causality. For selection bias to be present, the participation rate must have been low, and the participation rate could not be estimated in this study because the exact number of eligible subjects was unknown. Therefore, the present participants may not be representative of the Japanese general population. For example, the educational levels of the present subjects were higher than those of the general population. According to a population census that was conducted in 2010 in Ehime Prefecture28), the proportion of people aged 60–69 years with low, medium, and high educational levels and an unknown educational level were 28.2%, 48.6%, 19.0%, and 4.2%, respectively, in men and 26.7%, 56.4%, 12.9%, and 4.0%, respectively, in women. The corresponding figures in the present study for participants aged 60–69 years were 13.2%, 52.7%, 34.1%, and 0.0%, respectively, in men and 16.4%, 51.8%, 31.9%, and 0.0%, respectively, in women.
A potential limitation of the automated onscreen carotid ultrasonography device that was used in the present study is the lack of quantified plaque that is present in the images. CIMT and plaque are phenotypically distinct findings that both indicate increased vascular risk, but CIMT without plaque remains a significant marker of increased risk of vascular events and significantly predicts plaque occurrence29). We cannot rule out the possibility of diet as a residual or unmeasured confounding variable, the potential for measurement error from the FFQ, or environmental factors related to fatty acids.
The present cross-sectional study is the first to find significant inverse associations between fish intake and carotid wall thickening and the maximum CIMT although significant inverse associations were detected in Japanese women but not men. In Japanese women only, significant inverse relationships were observed between intake of n-3 PUFA, α-linolenic acid, EPA, DHA, n-6 PUFA, linoleic acid, and arachidonic acid and the prevalence of carotid wall thickening and/or the maximum CIMT. There was a significant positive association between n-3 PUFA intake and carotid wall thickening in Japanese men only. More work is needed to better understand the present findings.
There are no linked research data sets for this paper. It is not possible to anonymize the data.
The authors would like to thank the Yawatahama City Government, the Uchiko Town Government, the Seiyo City Government, the Ainan Town Government, the Ehime Prefecture Medical Association, Seiyo Municipal Hospital, Seiyo Municipal Nomura Hospital, the Minamiuwa Medical Association (Ito Clinic, Takemoto Clinic, Okazawa Clinic, Matsumoto Clinic, Nakaura Clinic, Hamaguchi Clinic, Kokuho Ipponmatsu Hospital, Uchiumi Clinic, Kogawa Family Clinic, and Ehime Prefectural Minamiuwa Hospital), and Hidehiko Onoue, MT (Junpu Health Care Center) for their valuable support.
This study was supported by the Research Unit of Ehime University and JSPS KAKENHI Grant Numbers 20K10249 and 21H03199.
The authors declare no competing interests.
YM, KT, and EK contributed to the study concept and design and the data acquisition. CN contributed to the dietary assessment. HS, YH, TM, TH, BM, and RK contributed to the data acquisition. YM was responsible for the analysis and interpretation of data and the drafting of the manuscript. All authors read and approved the final manuscript.
