Relationship between premenopausal dietary intake and postmenopausal subclinical atherosclerosis

Relationship between premenopausal dietary intake and postmenopausal subclinical atherosclerosis

Atherosclerosis 186 (2006) 420–427 Relationship between premenopausal dietary intake and postmenopausal subclinical atherosclerosis Hyun Ah Park, Jun...

171KB Sizes 0 Downloads 47 Views

Atherosclerosis 186 (2006) 420–427

Relationship between premenopausal dietary intake and postmenopausal subclinical atherosclerosis Hyun Ah Park, Jung Sun Lee, Lewis H. Kuller ∗ Department of Epidemiology, Graduate School of Public Heath, University of Pittsburgh, Pittsburgh, 130 N. Bellefield Avenue, Room 550, Pittsburgh, PA 15213, USA Received 7 May 2005; received in revised form 16 July 2005; accepted 1 August 2005 Available online 22 September 2005

Abstract Objectives: This study examines the association of premenopausal dietary intake with postmenopausal subclinical atherosclerosis. Design: A prospective population-based cohort of 401 premenopausal women from the Healthy Women Study. Baseline premenopausal dietary intake was determined with the use of single 24-h recall. Coronary and aortic calcium scores were measured by electron beam computed tomography 8 years after menopause, and carotid plaque index was measured by carotid ultrasound scan 5 years after menopause. Results: Prevalence of coronary, aortic, and carotid subclinical atherosclerosis were 47.3%, 75.4%, and 52.1%, respectively. In unadjusted analysis, the relative risks (RR) of saturated fat intake for coronary calcification in second, third and highest quartile groups compared to the lowest were 1.82 (95% confidence interval, 1.00–3.30), 1.49 (0.82–2.70), and 1.99 (1.09–3.62), while those of carbohydrate intake were 0.48 (95% confidence interval, 0.26–0.88), 0.47(0.26–0.86), and 0.35 (0.19–0.64), respectively. None of the dietary components were significantly associated with aortic calcification or carotid plaque. Conclusion: The association between premenopausal dietary intake and postmenopausal subclinical atherosclerosis supports the recommended premenopausal dietary intervention for the prevention of cardiovascular disease. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Diet; Subclinical atherosclerosis; Calcification; Plaque; Coronary artery; Aorta; Carotid artery; Women

1. Introduction The availability of new technology for measuring subclinical atherosclerosis such as coronary calcium and carotid intima-media thickness provides new approaches to evaluate risk factors in atherosclerosis. Though plaque instability and risk of thrombosis are important determinants of clinical cardiovascular disease in the presence of atherosclerosis, the extent of atherosclerosis is a key determinant of clinical disease. The subclinical atherosclerosis measures allow us non-invasive measurement of the extent of the atherosclerotic disease [1] with well-evaluated prognostic values [2–6]. Because of relatively long incubation period, risk factors evaluated at the same time as the measure of subclinical disease may have limited utility as markers of risk of atheroscle∗

Corresponding author. Tel.: +1 412 383 1895; fax: +1 412 383 1956. E-mail address: [email protected] (L.H. Kuller).

0021-9150/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2005.08.002

rosis unless they are highly correlated with previous risk factor levels. This long incubation period is especially problematic in the study of early postmenopausal women, i.e., in the first 5–15 years after the menopause because of the substantial changes in risk factors and lifestyles of women through the menopause. Our previous study found that the closer the measurements of risk factors were done with regard to the measurement of subclinical disease, the weaker the association [7]. However, the long term relationship between premenopausal dietary risk factors and subsequent postmenopausal subclinical atherosclerosis has not been examined yet. The Healthy Women Study (HWS) is a population-based prospective cohort study specially designed to understand the changes in cardiovascular risk factors in healthy premenopasual women as they go through the menopause. We have previously reported from the HWS that premenopausal cardiovascular risk factors such as blood pressure, serum

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427

lipids, body mass index, and smoking status are predictors of the amount of coronary and aortic calcification and carotid intima-media thickness at postmenopause [7–11]. Using the same data from the HWS, we have further explored the possibility of the dietary intake measured at premenopause as a modifiable risk factor of subclinical atherosclerosis. Specific questions examined in this study included: (1) whether premenopausal dietary intake would be associated with premenopausal serum lipids, and (2) premenopausal dietary intake would be associated with postmenopausal subclinical atherosclerosis measured approximately 13–15 years later. 2. Method 2.1. Study overview Detailed description of population recruitment and characteristics in HWS has been published elsewhere [12]. In brief, 541 premenopausal women living in Allegheny County, Pennsylvania were recruited between 1983 and 1984 from driver’s license list. The study eligibility criteria included the following: age 42–50 years; menstrual bleeding within the last 3 months (no hormone replacement therapy induced menstruation); diastolic blood pressure less than 100 mmHg; no medications known to influence cardiovascular risk factors (e.g., estrogen, insulin, lipid-lowing drugs, thyroid, antihypertensive, and psychotropic medication). After eligibility assessment, study participants were evaluated for a baseline examination. Fasting blood samples were collected to measure insulin, glucose, and serum lipids levels. Total cholesterol, high density lipoprotein (HDL) cholesterol, and triglycerides were assessed in the lipid laboratory of the Graduate School of Public Health using the standards of Centers for Disease Control [13]. Low density lipoprotein (LDL) cholesterol concentrations were estimated by the Friedewald equation [14]. Plasma insulin and glucose levels were determined by enzymeimmunoassay and radioimmuoassay, respectively. Blood pressure and anthropometric measures (height and weight) were made. A questionnaire was used to assess health-related behaviors including physical activity level, alcohol consumption, and smoking amount. Participants reported their menstrual status on postcards on a monthly basis after baseline examination. Menopause was defined as the cessation of cycling and/or using of hormone replacement therapy in combination for 12 months, at which time they were considered 1 year postmenopausal and reevaluated. Evaluations were also performed at 2, 5, 8, and 12 years after menopause. 2.2. Dietary assessment At the baseline clinic visit, dietary intake was assessed with the use of 1-day 24-h dietary recall. Interview was administrated by a trained nutritionist with extensive experience in collecting nutritional data using three-dimensional models of food portions in a clinical setting. Nutrients intake

421

was calculated using computerized nutrient database, which was a compilation of nutrient data mainly from the USDA’s Handbook No. 8 and data from MRFIT [15]. Nutrient intakes were expressed as the percentage of total energy intake, i.e., nutrient density. We also calculated the Keys score as a composite indicator of fat intake using Key et al.’s equation [16]. This score indicates the projected change in serum cholesterol (mg/dl) from dietary intake of cholesterol, saturated fat and polyunsaturated fat. 2.3. Subclinical atherosclerosis measures 2.3.1. Plaque index Carotid ultrasound scan measures were added to the protocol for women who were evaluated at 5 or 8 years after menopause from 1993. A Toshiba SSA-270A scanner (Toshiba American Medical Systems, New York, New York) was used. Certified sonographers scanned the right and left common carotid arteries, the carotid bifurcation, and the first 1.5 cm of the internal and external carotid arteries. For each location, the sonographers imaged the vessel in multiple planes and then focused on the largest area of focal plaque. Trained readers scored plaque index. Plaque was defined as a distinct area identified with either a focal area of hyperechogenicity and/or a focal protrusion into the vessel lumen with 50% greater thickness than surrounding areas. For each segment the degree of plaque was graded as follows: (0) no observable plaque; (1) one small plaque <30% of the vessel diameter; (2) one medium plaque between 30% and 50% of the vessel diameter or multiple small plaques; (3) one large plaque >50% of the vessel diameter or multiple plaques with at least one medium plaque. The plaque index summed the grades across the right and left carotid arteries to create as an overall measure of the extent of focal plaque. Reproducibility of the plaque index was evaluated in 15 subjects who had their scans performed by two sonographers. The intraclass correlation was 0.93. 2.3.2. Coronary and aortic calcium score Electron beam (EB) CT scans were performed using an Imatron C-150 Scanner (Imatron, South San Francisco, CA). For evaluation of the coronary arteries, 30–40 contiguous, 3 mm thickness transverse images from the level of the aortic root to the apex of heart were taken. Images were obtained during maximal breath-holding by using ECG triggering so that each 100 ms exposure was obtained during the same phase of the cardiac cycle at 80% of the RR interval to minimize the effect of cardiac motion. During a second pass, 6 mm contiguous images of the aorta of 300 ms exposure were obtained from the aortic arch to the iliac bifurcation. Calcium score was calculated by the method of Agatston et al. [17] with a densitometric software program available on the Imatron C-150 scanner. The reproducibility of both coronary and aortic scan was very high. The intraclass correlation for the coronary score was 0.99 and that of aortic score was 0.98.

422

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427

2.4. Analytic sample During the 8 years of postmenopausal follow-up, there were 15 deaths and 59 drop-outs. Of the remaining 467 women, 357 women had EBCT examination 8 years after menopause, 382 women had carotid ultrasound scan 5 years after menopause. Those who were not measured for subclinical indexes were more likely to be single, black, and smoker and to have higher blood pressure, fasting insulin, and lower HDL cholesterol than those who were measured. The analytic sample for coronary artery calcium was 357 women, for aortic calcification was 346 women, and for carotid plaque was 382 women. 2.5. Statistical analysis Because of highly skewed distribution of subclinical atherosclerosis measures, we categorized calcification scores into two groups of those without calcification and with calcification, respectively. The plaque index was also dichotomized into two groups of those without plaque and with plaque. Differences of baseline premenopausal cardiovascular risk factors by postmenopausal subclinical disease status were analyzed using t-test and Mann–Whitney test for continuous variables, and chi-square test for categorical variables. The associations between baseline premenopausal dietary intake and premenopausal serum lipids were examined by multivariate linear regression analysis while adjusting for potential confounders. After univariate logistic regression analysis to assess the relationship between baseline premenopausal dietary intake and postmenopausal subclinical measures, we adjusted relative risks (RR) for nondietary cardiovascular risk factors to test independent associations. Because serum lipids level was postulated as a mediator between dietary intake and subclinical atherosclerosis, it was not included as a covariate in the multivariate models. For these analyses, nutrient intakes as the percentage of total energy intake (nutrient density) and serum lipids (total cholesterol, LDL cholesterol, HDL cholesterol, and triglyceride) were divided into quartile groups. The lowest quartile group was used as the reference group. All statistical analysis was conducted using the SPSS 11.5 statistical packages, and statistical significance was set at P < 0.05. All P values were two-tailed.

3. Result At the baseline examination, mean age of participants who had EBCT and carotid ultrasound scan were 62.1 and 58.6 years, respectively. Average follow-up time (mean ± S.D.) since the baseline examination for participants who had EBCT and carotid ultrasound scan were 14.5 ± 1.3 and 11.0 ± 1.6 years, which were on average 10.1 ± 2.5 and 6.7 ± 2.1 years after menopause, respectively.

Fig. 1. Distribution of subclinical atherosclerosis measures.

The distributions of three subclinical atherosclerosis measures were presented in Fig. 1. Coronary calcium score ranged from 0 to 1175 with median values of 0 and the aortic calcium score ranged from 0 to 8417 with a median value of 84. Carotid plaque index ranged from 0 to 13 (median, 1). The measures of subclinical atherosclerosis were correlated with one another (0.208–0.422, P < 0.001). The strongest correlation was found between the coronary and aortic calcification scores (0.422, P < 0.001). The prevalence of any measurable disease was 47.3% (169 participants) for the coronary arteries (coronary calcium score > 0), 75.4% (261 participants) for aorta (aortic calcium score > 0), and 52.1% (199 participants) for carotid plaque (plaque index > 0). Baseline premenopausal cardiovascular risk factors were associated with each measure of subclinical atherosclerosis (Table 1). Persons with subclinical atherosclerosis were more likely to be less educated, and to have higher LDL cholesterol, triglyceride, and blood pressure, and to have lower HDL cholesterol than those without subclinical atherosclerosis. Fasting glucose level was associated with only aortic calcification (P = 0.005), and body mass index was associated only with coronary calcification (P < 0.001). Parental history of stroke was associated with carotid plaque (P = 0.013).

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427

423

Table 1 Baseline premenopausal risk factors by postmenopausal subclinical atherosclerosis

Age at entry, year (mean, S.D.) White, n (%) College graduate, n (%) Married (%) Fasting glucose (mg/dl)a Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Pulse pressure (mmHg) Cholesterol (mg/dl) LDL-C (mg/dl)b HDL-C (mg/dl)b Triglyceride (mg/dl)a Body Mass Index (wt/ht2 )a Physical activity (kcal/w)a Baseline smoking, n (%) Alcohol intake >30 g/day, n (%) Parental history of CHD, n (%)b Parental history of stroke, n (%) fifth menopausal HRT use, n (%)b

Coronary calcification

Aortic calcification

Carotid plaque index

No (N = 188)

Yes (N = 169)

No (N = 85)

Yes (N = 261)

No (N = 183)

Yes (N = 199)

47.6 ± 1.6 178 (94.7) 106 (56.4) 145 (77.1) 85.4 ± 10.0 106.0 ± 9.9 71.2 ± 7.8 34.8 ± 7.2 178.9 ± 26.3 102.0 ± 24.4 62.1 ± 13.2 74.1 ± 37.3 23.7 ± 3.3 1509.1 ± 1600.8 36 (19.1) 7 (3.7) 71 (37.8) 37 (19.7) 95 (52.5)

47.5 ± 1.6 156 (92.3) 72 (42.6)* 127 (75.1) 88.8 ± 25.2 110.4 ± 12.9*** 72.6 ± 7.9 37.8 ± 9.0** 191.5 ± 35.7*** 116.2 ± 33.1*** 57.5 ± 14.4** 89.1 ± 52.4** 25.8 ± 5.1*** 1596.2 ± 2055.6 60 (35.5)** 10 (5.9) 73 (43.2) 47 (27.8) 73 (48.3)

47.3 ± 1.7 77 (90.6) 52 (61.2) 59 (69.4) 87.1 ± 33.9 103.4 ± 9.4 69.4 ± 8.1 34.0 ± 7.2 173.9 ± 27.7 96.8 ± 25.4 62.6 ± 11.2 72.2 ± 36.7 23.9 ± 3.5 1873.1 ± 2995.9 10 (11.8) 4 (4.7) 30 (35.3) 17 (20.0) 47 (58.8)

47.7 ± 1.6 246 (94.3) 119 (45.6)* 204 (78.2) 87.7 ± 10.3** 109.4 ± 11.8*** 72.5 ± 7.5** 37.0 ± 8.5** 188.2 ± 32.4*** 112.4 ± 30.0*** 58.8 ± 14.4* 84.6 ± 48.6 24.9 ± 4.6 1463.6 ± 1246.6 80 (30.7)** 13 (5.0) 111 (42.5) 63 (24.1) 115 (47.7)

47.6 ± 1.6 169 (92.3) 108 (59.0) 141 (77.0) 85.4 ± 9.7 106.5 ± 10.3 71.2 ± 7.7 34.7 ± 7.5 177.9 ± 27.3 102.0 ± 25.0 61.2 ± 12.7 73.2 ± 36.4 24.4 ± 4.0 1385.9 ± 1556.6 34 (18.6) 8 (4.4) 70 (38.3) 32 (17.5) 89 (51.1)

47.8 ± 1.5 184 (92.5) 86 (43.2)** 149 (77.0) 88.8 ± 23.4 109.8 ± 11.8** 72.4 ± 8.1 37.4 ± 8.2** 189.6 ± 32.0*** 112.9 ± 30.5*** 57.2 ± 14.9 87.7 ± 52.0** 24.7 ± 4.5 1597.5 ± 1981.2 73(36.7)*** 13 (6.5) 88 (44.2) 56 (28.1)* 84 (44.4)

a

Mann–Whitney test. LDL cholesterol, low density lipoprotein cholesterol; HDL cholesterol, high density lipoprotein cholesterol; CHD, coronary heart disease; HRT, hormone replacement therapy. * P < 0.05. ** P < 0.01. *** P < 0.001. b

The distribution of total energy and macronutrients intake of study participants are presented in Table 2. The median of daily energy intake was 1720.2 kcal. The median percentage of energy intake from carbohydrate, protein and fat was 43.4%, 16.0% and 38.6%, respectively. At baseline, premenopausal dietary intake was associated with serum lipids (Table 3). Total energy intake was positively associated with total cholesterol (P for trend = 0.016), LDL cholesterol (P = 0.025), and triglyceride (P = 0.016) after controlling for age, physical activity, alcohol, smoking, and body mass index. The percentage of energy from carbohydrate was negatively related to HDL cholesterol (P = 0.001), and posTable 2 The distribution of baseline macronutrients intake, keys score, and fiber intake for 401 HWS participants

Total calori (kcal) Protein (%) Carbohydrate (%) Fat (%) Saturated fat (%) Cholesterol (mg/1000 kcal) Keys scorea Fiber (g/1000 kcal)

Median

Interquarile range 25 percentile

75 percentile

1720.2 16.0 43.4 38.6 12.5 145.3 44.7 2.1

1369.8 13.0 35.8 31.9 9.8 96.6 35.7 1.3

2114.2 19.3 50.2 43.7 15.8 244.5 56.5 3.1

Keys score = 1.26(2S − P) + 1.5 root C, where S and P are the percentages of total energy from saturated and polyunsaturated fats, respectively, and C is the daily cholesterol intake in mg/1000 kcal. Higher scores indicate higher projected changes in serum cholesterol (mg/dl).

itively to triglyceride (P = 0.021). The percentage of energy from fat, saturated fat and Keys score were positively associated with total cholesterol and LDL cholesterol, respectively. The association between premenopausal dietary intake and subsequent postmenopausal atherosclerosis is shown in Table 4. In unadjusted analyses, total energy intake, the percentage of energy from protein, saturated fat, cholesterol (nutrient density), and Keys score were associated with an increased risk of coronary calcification, and the percentage of energy from carbohydrate and fiber intake were associated with a decreased risk of coronary calcification. The unadjusted relative risks for the second, third, and highest quartile of saturated fat intake compared to the lowest quartile were 1.82 (95% confidence interval (CI): 1.00–3.30), 1.49 (0.82–2.70), and 1.99 (1.09–3.62), respectively. The comparable RRs for carbohydrate intake were 0.48 (95% CI: 0.26–0.88), 0.47 (0.26–0.86), and 0.35 (0.19–0.64), respectively. These associations remained even after controlling for other baseline premenopausal cardiovascular risk factors. Baseline premenopausal dietary intake, however was not significantly associated with postmenopausal aortic calcification or carotid plaque.

4. Discussion

a

Few studies examined the relationship between diet and subclinical atherosclerosis. Available studies on this subject focused on single nutrient [18,19] and/or assessed only

424

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427

Table 3 Mean values (±S.D.) of baseline serum lipids by quartiles of baseline macronutrients intake, keys score, and fiber intake for 401 HWS participants Total cholesterol (mg/dl)

HDL-Ca (mg/dl)

LDL-Ca (mg/dl)

Triglyceride (mg/dl)

Calories (kcal) Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

178.8 ± 30.5 180.9 ± 28.9 189.9 ± 32.1 188.6 ± 32.3 0.011 0.016

60.6 ± 13.0 59.7 ± 13.7 60.6 ± 13.8 58.8 ± 14.4 0.393 0.435

103.9 ± 29.1 104.5 ± 28.0 112.8 ± 28.5 112.0 ± 30.3 0.020 0.025

71.5 ± 34.5 83.6 ± 46.3 82.4 ± 50.9 89.0 ± 51.2 0.012 0.016

Protein % energy Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

183.0 ± 30.0 183.4 ± 31.6 185.2 ± 29.9 186.7 ± 33.6 0.357 0.498

61.8 ± 13.7 57.5 ± 13.3 60.6 ± 13.7 59.8 ± 14.0 0.646 0.681

105.6 ± 28.5 108.4 ± 29.0 108.8 ± 28.3 110.4 ± 31.2 0.269 0.526

77.9 ± 38.6 87.2 ± 55.4 78.9 ± 42.7 82.5 ± 47.7 0.760 0.818

Carbohydrate % energy Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

187.7 ± 32.4 187.4 ± 33.0 182.0 ± 28.1 181.2 ± 31.1 0.080 0.126

63.0 ± 15.3 60.5 ± 13.6 58.8 ± 12.5 57.3 ± 12.9 0.002 0.001

109.3 ± 29.5 111.0 ± 30.0 106.7 ± 28.6 106.4 ± 28.8 0.342 0.339

77.2 ± 40.9 79.3 ± 41.0 82.5 ± 45.5 87.4 ± 56.9 0.106 0.021

Fat % energy Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

178.9 ± 31.5 181.7 ± 23.4 189.3 ± 33.1 188.2 ± 34.9 0.012 0.032

58.4 ± 13.8 60.2 ± 13.0 61.6 ± 14.7 59.5 ± 13.3 0.423 0.029

102.6 ± 29.4 106.1 ± 22.7 111.6 ± 30.3 113.0 ± 32.6 0.005 0.027

89.8 ± 63.9 77.2 ± 35.4 80.8 ± 41.3 78.7 ± 39.9 0.131 0.008

Saturated fat % energy Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

180.8 ± 32.9 181.9 ± 27.1 185.8 ± 28.8 189.6 ± 35.1 0.032 0.033

58.9 ± 13.5 61.1 ± 13.9 59.1 ± 13.7 60.6 ± 13.8 0.565 0.270

105.0 ± 31.5 104.7 ± 26.6 109.6 ± 25.9 113.9 ± 31.7 0.017 0.023

84.6 ± 58.1 80.8 ± 43.9 85.3 ± 42.6 75.8 ± 39.5 0.264 0.178

Cholesterol (mg/1000 kcal) Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

181.3 ± 29.1 184.3 ± 32.7 187.7 ± 33.1 184.8 ± 29.9 0.648 0.666

61.3 ± 14.1 58.8 ± 13.9 60.2 ± 13.1 59.4 ± 13.8 0.586 0.723

104.2 ± 28.1 108.4 ± 29.7 111.7 ± 30.4 109.0 ± 28.4 0.316 0.623

79.1 ± 42.6 86.0 ± 50.4 79.0 ± 45.0 82.5 ± 48.1 0.927 0.572

Keys scorec Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

183.5 ± 33.6 186.4 ± 30.7 180.0 ± 25.0 188.3 ± 33.7 0.024 0.030

59.5 ± 13.0 59.2 ± 15.0 60.7 ± 13.5 60.2 ± 13.3 0.903 0.302

106.8 ± 31.7 110.6 ± 28.2 103.2 ± 24.0 112.8 ± 30.9 0.010 0.023

86.0 ± 58.4 83.0 ± 36.0 80.9 ± 46.0 76.6 ± 42.8 0.502 0.252

Fiber (g/1000 kcal) Q1 Q2 Q3 Q4 Unadjusted P Adjusted Pb

190.0 ± 31.5 183.7 ± 32.1 181.2 ± 29.7 183.4 ± 31.4 0.183 0.583

58.7 ± 14.2 58.9 ± 12.8 61.7 ± 14.2 60.3 ± 13.7 0.275 0.975

112.9 ± 29.2 108.7 ± 29.8 103.9 ± 27.7 107.8 ± 29.6 0.208 0.777

91.8 ± 58.9 80.2 ± 43.8 78.2 ± 40.0 76.3 ± 40.1 0.032 0.324

a

LDL cholesterol, low density lipoprotein cholesterol; HDL cholesterol, high density lipoprotein cholesterol. Adjusted for age (continuous), physical activity (<1000, 1000–1999, or ≥2000 kcal/week), alcohol (<30 or ≥30 g/day), smoking (yes or no), and body mass index (<25 kg/m2 or ≥25 kg/m2 ). c Keys score = 1.26(2S − P) + 1.5 root C, where S and P are the percentages of total energy from saturated and polyunsaturated fats, respectively, and C is the daily cholesterol intake in mg/1000 kcal. Higher scores indicate higher projected changes in serum cholesterol (mg/dl). b

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427

425

Table 4 The relative risks of subclinical atherosclerosis by quartiles of baseline macronutrients intake, keys score, and fiber intake for 401 HWS participants Coronary calcification (N = 357)

Aortic calcification (N = 346)

Carotid plaque index (N = 382)

Unadjusted RR (95% CI)

Adjusted RR (95% CI)a

Unadjusted RR (95% CI)

Adjusted RR (95% CI)a

Unadjusted RR (95% CI)

Adjusted RR (95% CI)a

Calories (kcal) Q1 1 Q2 1.45 (0.80, 2.64) Q3 1.86 (1.02, 3.37) Q4 1.99 (1.09, 3.62)

1 1.89 (0.95, 3.74) 1.90 (0.96, 3.75) 2.47 (1.25, 4.88)

1 0.90 (0.45, 1.78) 0.80 (0.41, 1.58) 1.41 (0.68, 2.94)

1 0.98 (0.44, 2.18) 0.64 (0.29, 1.41) 1.58 (0.68, 3.66)

1 0.90 (0.51, 1.59) 0.94 (0.53, 1.66) 1.29 (0.73, 2.29)

1 1.11 (0.58, 2.12) 0.87 (0.45, 1.66) 1.92 (0.99, 3.73)

Protein % energy Q1 1 Q2 1.00 (0.56, 1.80) Q3 1.02 (0.57, 1.84) Q4 1.20 (0.67, 2.16)

1 1.15 (0.59, 2.23) 1.33 (0.68, 2.62) 2.20 (1.07, 4.53)

1 1.55 (0.78, 3.08) 1.36 (0.70, 2.67) 1.53 (0.77, 3.04)

1 1.78 (0.82, 3.86) 1.84 (0.85, 3.98) 2.25 (0.98, 5.18)

1 0.56 (0.32, 1.00) 0.61 (0.34, 1.09) 0.68 (0.38, 1.21)

1 0.52 (0.28, 1.01) 0.58 (0.30, 1.11) 0.74 (0.37, 1.50)

Carbohydrate % energy Q1 1 Q2 0.48 (0.26, 0.88) Q3 0.47 (0.26, 0.86) Q4 0.35 (0.19, 0.64)

1 0.47 (0.24, 0.92) 0.44 (0.22, 0.88) 0.37 (0.18, 0.74)

1 1.02 (0.49, 2.09) 0.63 (0.32, 1.25) 0.94 (0.46, 1.91)

1 1.10 (0.49, 2.51) 0.60 (0.27, 1.34) 1.20 (0.52, 2.79)

1 0.59 (0.33, 1.05) 0.79 (0.45, 1.40) 0.84 (0.48, 1.50)

1 0.59 (0.30, 1.15) 0.98 (0.50, 1.92) 1.08 (0.55, 2.15)

Fat % energy Q1 Q2 Q3 Q4

1 0.96 (0.53, 1.73) 1.17 (0.65, 2.11) 1.57 (0.87, 2.84)

1 0.90 (0.46, 1.74) 0.99 (0.51, 1.92) 1.24 (0.63, 2.45)

1 0.79 (0.39, 1.59) 0.67 (0.34, 1.32) 1.15 (0.55, 2.40)

1 0.72 (0.33, 1.57) 0.44 (0.20, 0.97) 0.78 (0.33, 1.87)

1 0.62 (0.35, 1.09) 1.06 (0.60, 1.89) 0.71 (0.40, 1.26)

1 0.63 (0.33, 1.20) 0.95 (0.50, 1.82) 0.47 (0.24, 0.93)

Saturated fat % energy Q1 1 Q2 1.82 (1.00, 3.30) Q3 1.49 (0.82, 2.70) Q4 1.99 (1.09, 3.62)

1 1.74 (0.89, 3.40) 1.32 (0.68, 2.59) 1.70 (0.87, 3.32)

1 1.15 (0.57, 2.31) 0.76 (0.39, 1.48) 1.50 (0.73, 3.11)

1 1.05 (0.48, 2.29) 0.61 (0.29, 1.29) 1.27 (0.57, 2.86)

1 0.94 (0.53, 1.66) 0.79 (0.45, 1.40) 1.04 (0.59, 1.85)

1 0.97 (0.51, 1.85) 0.75 (0.39, 1.43) 0.86 (0.45, 1.65)

Cholesterol (mg/1000 kcal) Q1 1 Q2 1.67 (0.91, 3.04) Q3 2.23 (1.22, 4.06) Q4 1.74 (0.96, 3.17)

1 1.69 (0.86, 3.31) 2.53 (1.27, 5.04) 1.73 (0.87, 3.42)

1 1.08 (0.55, 2.13) 1.15 (0.58, 2.28) 1.29 (0.64, 2.59)

1 0.92 (0.43, 1.97) 0.88 (0.40, 1.94) 1.20 (0.55, 2.65)

1 1.06 (0.60, 1.88) 1.02 (0.58, 1.80) 1.00 (0.57, 1.77)

1 1.25 (0.66, 2.38) 0.94 (0.49, 1.82) 0.99 (0.52, 1.88)

Keys scoreb Q1 Q2 Q3 Q4

1 1.21 (0.66, 2.20) 2.12 (1.16, 3.84) 1.73 (0.95, 3.14)

1 1.16 (0.59, 2.28) 2.21 (1.12, 4.34) 1.65 (0.84, 3.24)

1 1.08 (0.55, 2.11) 1.02 (0.52, 1.98) 1.99 (0.95, 4.18)

1 1.03 (0.48, 2.19) 0.75 (0.35, 1.61) 2.05 (0.89, 4.72)

1 0.70 (0.40, 1.24) 0.90 (0.51, 1.59) 1.04 (0.59, 1.85)

1 0.78 (0.41, 1.47) 0.80 (0.42, 1.53) 0.94 (0.49, 1.81)

Fiber (g/1000 kcal) Q1 1 Q2 0.58 (0.32, 1.05) Q3 0.54 (0.30, 0.98) Q4 0.44 (0.24, 0.80)

1 0.53 (0.27, 1.05) 0.58 (0.30, 1.13) 0.49 (0.24, 0.98)

1 0.80 (0.40, 1.58) 0.80 (0.40, 1.58) 1.23 (0.59, 2.55)

1 0.73 (0.33, 1.59) 0.89 (0.41, 1.94) 2.00 (0.84, 4.75)

1 0.90 (0.51, 1.59) 1.21 (0.68, 2.13) 1.18 (0.67, 2.10)

1 1.04 (0.54, 2.00) 1.92 (0.99, 3.68) 1.71 (0.86, 3.40)

a Adjusted for age (continuous), education (college graduate, under college graduate), systolic blood pressure (continuous), fasting glucose (continuous), physical activity (<1000, 1000–1999, or ≥2000 kcal/week), alcohol intake (<30 or ≥30 g/day), smoking (yes or no), parental history of coronary heart disease (yes or no), parental history of stroke (yes or no), total energy intake (continuous), and fifth menopausal year hormone therapy (yes or no). b Keys score = 1.26(2S − P) + 1.5 root C, where S and P are the percentages of total energy from saturated and polyunsaturated fats, respectively, and C is the daily cholesterol intake in mg/1000 kcal. Higher scores indicate higher projected changes in serum cholesterol (mg/dl).

cross–sectional relationships [20,21]. To our best knowledge, this is the first study reporting the longitudinal relationship between dietary intake of healthy premenopausal women and subsequent postmenopausal subclinical atherosclerosis covering three vascular beds over 11–14 years follow-up. We observed a positive association of saturated fat intake (as a percentage of total energy) and Keys score with subclinical coronary calcification, respectively. Both saturated

fat and Keys score were also found to be positively associated with total cholesterol and LDL cholesterol in premenopausal cross-sectional analysis. These results suggest that serum cholesterol level could mediate the effects of dietary fat on coronary atherosclerosis risk. Similar findings were reported from large prospective studies in which incident coronary heart disease was used as an outcome measure instead of coronary calcification. The Nurses’ Health Study and Health Pro-

426

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427

fessionals Follow-up study showed the association of dietary saturated fat with coronary heart disease [22,23]. Especially Ascherio et al. found that saturated fat influence the coronary heart disease risk as predicted by their effects on serum cholesterol in men [23]. Carbohydrate intake (as a percentage of total energy) showed a significant protective effect on coronary calcification despite its unfavorable effect on serum lipids (decrease of HDL cholesterol and increase of triglyceride). Further adjustment for serum lipids level increased the protective effect of carbohydrate intake. (data not shown) One possible explanation for this seemingly paradoxical effect is that higher carbohydrate intake (as the percentage of total energy) is associated with lower intake of fat and energy, and higher intake of dietary fiber (data not shown) which is similar to the findings from the nationally representative samples [24,25]. Total fat intake was not a significant predictor of subclinical atherosclerosis in our study even though it has been one of the major target nutrients in nutrition and public health campaign. Because different subtypes of fat have different effects on the risk of atherosclerosis [26], the summed effect could vary according to the composition of subtype fat intakes. Contrary to the coronary results, no dietary nutrient intake was proved to be significantly associated with aortic or carotid atherosclerosis risk. We think the effect of serum lipids on coronary atherosclerosis might be more evident than that on aortic and carotid atherosclerosis. For these vessels, hypertension may be a more dominant risk factor [27]. As Anderson pointed out, the effects of cardiovascular risk factors on different cardiovascular disease endpoints may be differential [28]. The underlying mechanisms for this need further exploration. Adjusting nondietary cardiovascular risk factors into the final models did not substantially change the association between the dietary intake and subclinical atherosclerosis (Table 4), suggesting their minimal confounding effect in our sample. This may be also partly due to the relatively homogeneous cardiovascular risk factor profiles among the HWS participants according to study eligibility criteria. The minimal changes in estimates could also mean that residual confounding is unlikely. This study has limitation. Dietary intake was assessed with the use of 1-day 24-h recall. As with any other dietary assessment methods, single 24-h recall may have limitations to estimate the long term dietary intake of premenopausal women. Even though our assessment was done by a trained nutritionist with extensive experience in assessing the dietary pattern and eating behaviors of middle aged women using the best available approaches, there remains the possibility of measurement error. One of the strengths of our study is the use of markers of subclinical atherosclerosis as outcome variables instead of clinical diagnosis of cardiovascular disease. It provides an opportunity to identify a link between diet and atherosclerosis at the very early stage of the long disease process and hence to intervene at the asymptomatic disease stage. An aggressive risk factor modification exerted at premenopause might

have a potential to reduce the risk of atherosclerosis and the incidence of heart attack at postmenopause. In summary, our results suggest that premenopausal dietary intake affects serum lipids level and risk of postmenopausal subclinical atherosclerosis. Since menopausal transition is associated with accelerated subclinical atherosclerosis progression [29], premenopause, especially perimenopause is a critical period for risk factor modification through lifestyle intervention.

Acknowledgement This study was supported by National Institutes of Health grant HL-28266.

References [1] American Heart Association, American College of Cardiology/American Heart Association Expert Consensus document on electron-beam computed tomography for the diagnosis and prognosis of coronary artery disease. Circulation 2000;102:126–40. [2] Raggi P, Callister TQ, Cooil B, et al. Identification of patients at increased risk of first unheralded acute myocardial infarction by electron-beam computed tomography. Circulation 2000;101:850–5. [3] Wong ND, Hsu JC, Detrano RC, et al. Coronary artery calcium evaluation by electron beam computed tomography and its relation to new cardiovascular events. Am J Cardiol 2000;86:495–8. [4] Chan SY, Mancini GB, Kuramoto L, et al. The prognostic importance of endothelial dysfunction and carotid atheroma burden in patients with coronary artery disease. J Am Coll Cardiol 2003;42:1037–43. [5] Stork S, van den Beld AW, von Schacky C, et al. Carotid artery plaque burden, stiffness, and mortality risk in elderly men: a prospective, population-based cohort study. Circulation 2004;110:344–8. [6] Honda O, Sugiyama S, Kugiyama K, et al. Echolucent carotid plaques predict future coronary events in patients with coronary artery disease. J Am Coll Cardiol 2004;43:1177–84. [7] Matthews KA, Kuller LH, Sutton-Tyrrell K, Chang YF. Changes in cardiovascular risk factors during the perimenopause and postmenopause and carotid artery atherosclerosis in healthy women. Stroke 2001;32:1104–11. [8] Kuller LH, Matthews KA, Sutton-Tyrrell K, Edmundowicz D, Bunker CH. Coronary and aortic calcification among women 8 years after menopause and their premenopausal risk factors: the healthy women study. Arterioscler Thromb Vasc Biol 1999;19:2189–98. [9] Sutton-Tyrrell K, Lassila HC, Meilahn E, et al. Carotid atherosclerosis in premenopausal and postmenopausal women and its association with risk factors measured after menopause. Stroke 1998;29:1116–21. [10] Lassila HC, Tyrrell KS, Matthews KA, Wolfson SK, Kuller LH. Prevalence and determinants of carotid atherosclerosis in healthy postmenopausal women. Stroke 1997;28:513–7. [11] Sutton-Tyrrell K, Kuller LH, Matthews KA, et al. Subclinical atherosclerosis in multiple vascular beds: an index of atherosclerotic burden evaluated in postmenopausal women. Atherosclerosis 2002;160:407–16. [12] Matthews KA, Meilahn E, Kuller LH, et al. Menopause and risk factors for coronary heart disease. N Engl J Med 1989;321:641–6. [13] Allain CC, Poon LS, Chan CS, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem 1974;20:470–5. [14] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499–502.

H.A. Park et al. / Atherosclerosis 186 (2006) 420–427 [15] Dolecek TA, Stamler J, Caggiula AW, Tillotson JL, Buzzard IM. Methods of dietary and nutritional assessment and intervention and other methods in the multiple risk factor intervention trial. Am J Clin Nutr 1997;65:196S–210S. [16] Keys A, Parlin RW. Serum cholesterol response to changes in dietary lipids. Am J Clin Nutr 1966;19:175–81. [17] Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15:827–32. [18] Simon JA, Murtaugh MA, Gross MD, et al. Relation of ascorbic acid to coronary artery calcium: the coronary artery risk development in young adults study. Am J Epidemiol 2004;159:581–8. [19] Kritchevsky SB, Tell GS, Shimakawa T, et al. A carotenoid intake and carotid artery plaques: the atherosclerosis risk in communities study. Am J Clin Nutr 1998;68:726–33. [20] Tell GS, Evans GW, Folsom AR, et al. Dietary fat intake and carotid artery wall thickness: the Atherosclerosis risk in communities (ARIC) study. Am J Epidemiol 1994;139:979–89. [21] Oh KW, Nam CM, Jee SH, Choe KO, Suh I. Coronary artery calcification and dietary cholesterol intake in Korean men. Acta Cardiol 2002;57:5–11. [22] Hu FB, Stampfer MJ, Manson JE, et al. Dietary fat intake and the risk of coronary heart disease in women. N Engl J Med 1997;337:1491–9.

427

[23] Ascherio A, Rimm EB, Giovannucci EL, et al. Dietary fat and risk of coronary heart disease in men: cohort follow up study in the United States. BMJ 1996;313:84–90. [24] Bowman SA, Spence JT. A comparison of low-carbohydrate vs. high-carbohydrate diets: energy restriction, nutrient quality and correlation to body mass index. J Am Coll Nutr 2002;21:268– 74. [25] Ursin G, Ziegler RG, Subar AF, et al. Dietary patterns associated with a low-fat diet in the national health examination follow-up study: identification of potential confounders for epidemiologic analyses. Am J Epidemiol 1993;137:916–27. [26] Hu FB, Manson JE, Willett WC. Types of dietary fat and risk of coronary heart disease: a critical review. J Am Coll Nutr 2001;20:5–19. [27] Pinto A, Tuttolomondo A, Di Raimondo D, Fernandez P, Licata G. Cerebrovascular risk factors and clinical classification of strokes. Semin Vasc Med 2004;4:287–303. [28] Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular disease risk profiles. Am Heart J 1991;121:293–8. [29] Wildman RP, Schott LL, Brockwell S, Kuller LH, Sutton-Tyrrell K. A dietary and exercise intervention slows menopause-associated progression of subclinical atherosclerosis as measured by intimamedia thickness of the carotid arteries. J Am Coll Cardiol 2004;44: 579–85.