TABLE 3.
Relative risk (RR) estimates for the association of saturated fat intake and risk of coronary heart disease (CHD) or stroke in 21 unique prospective epidemiologic studies1
| Study | Sex | Cases | Median or mean saturated fat intake | Adjusted covariates | Multivariate adjusted RR (95% CI) |
| Coronary heart disease studies | |||||
| Shekelle et al, 1981 (17) (Western Electric Study) | Male | Fatal CHD: 215 | 16.6% of total energy | Age, SBP, cigarettes per day, serum cholesterol, alcoholic drinks per month, BMI, geographic origin | β = 0.031, P = 0.144 For 1-unit increase in saturated fat |
| McGee et al, 1984 (9) (Honolulu Heart Study)2 | Male | Total CHD: 1177 | 12.7% of total energy (age-adjusted)3 | Age, total energy intake, SBP, BMI, smoking, family history of MI, physical activity, intakes of PUFA, alcohol, protein, carbohydrate, vegetable, and cholesterol | RRmen < 60 y = 0.92 (0.68, 1.23)2 RRmen ≥ 60 y = 0.70 (0.41, 1.20)2 Pooled RR = 0.86 (0.67, 1.12) P for test of heterogeneity = 0.39 For fifth vs first quintile |
| Kushi et al, 1985 (13) (Ireland Boston Diet Heart Study) | Male | Fatal CHD: 110 | 16.8% of total energy3 | Age, SBP, serum cholesterol, cigarette smoking, alcohol intake, cohort | β = 0.061, P = 0.05 For 1-unit increase in saturated fat |
| Posner et al, 1991 (16) (Framingham Study) | Male | Total CHD: 213 | 45–55 y old: 15.2% of total energy3 56–65 y old: 14.8% of total energy3 | Variable of interest, energy intake, physical activity, serum cholesterol, SBP, left ventricular hypertrophy, cigarette smoking, glucose intolerance, Metropolitan relative weight | RR45–55 y = 0.78 (0.61, 1.00) RR≥56 y = 1.06 (0.86, 1.30) Pooled RR = 0.92 (0.68, 1.24) P for test of heterogeneity = 0.06 For recommended versus actual intake (15.2% vs 10%) |
| Fehily et al, 1993 (28) (Caerphilly Study) | Male | Total CHD: 21 | 17.3% for CHD-free subjects and 18.1% for CHD cases | None | RR = 1.57 (0.56, 4.42) For third vs first tertile |
| Goldbourt et al, 1993 (35) (Israeli Ischemic Heart Disease Study) | Male | Fatal CHD: 1070 | NR | Age, blood pressure, serum cholesterol, ever-smoking, diabetes prevalence in 1963 | RRmen <60 y = 1.05 (0.87, 1.27)4 RRmen ≥60 y = 0.66 (0. 44, 1.00)4 Pooled RR = 0.86 (0.56, 1.35) P for test of heterogeneity = 0.05 For fourth vs first quartile |
| Ascherio et al, 1996 (4) (Health Professionals Follow-Up Study) | Male | Total CHD: 1702 | Fifth quintile: 14.8% of total energy First quintile: 7.2% of total energy | Age, BMI, smoking, physical activity, history of hypertension or high blood cholesterol, history of MI < age 60 y, energy intake, fiber | RRmen <60 y = 1.24 (0.87, 1.77)4 RRmen ≥60 y = 1.01 (0.73, 1.41)4 Pooled RR = 1.11 (0.87, 1.42) P for test of heterogeneity = 0.42 For fifth vs first quintile |
| Esrey et al, 1996 (6) (Lipid Research Clinics Study) | Both | Fatal CHD: 92 | 30–59 y old: 16.8% for CHD deaths and 15.1% for non-CHD deaths3 60–79 y old: 13.8% for CHD deaths and 14.3% for non-CHD deaths3 | Age, sex, energy intake, serum lipids, SBP, cigarette smoking status, BMI, glucose intolerance | RR<60 y = 1.11 (1.04, 1.18)5 RR≥60 y = 0.96 (0.88, 1.05) Pooled RR = 0.97 (0.80, 1.18) P for test of heterogeneity = 0.40 For 1-unit increase in saturated fat |
| Mann et al, 1997 (32) | Both | Fatal CHD: 45 | Men: Third tertile, 41.0 g/d; First tertile, 14.6 g/d Women: Third tertile, 38.1 g/d; First tertile, 13.7 g/d | Age, sex, smoking, social class | RR = 2.77 (1.25, 6.13)5 For third vs first tertile |
| Pietinen et al, 1997 (15) (Alpha-Tocopherol, Beta-Carotene Study) | Male | Total CHD: 635 | Fifth quintile: 67.5 g/d First quintile: 34.7 g/d | Age, treatment group, smoking, BMI, blood pressure, education, intakes of energy, alcohol, fiber, physical activity, intakes of linoleic acid and trans and monounsaturated fats | RR = 0.93 (0.60, 1.44) For fifth vs first quintile |
| Boniface and Tefft, 2002 (5) (Health and Lifestyle Survey) | Both | Fatal CHD: 155 | Men: 47.0 g/d3 Women: 34.4 g/d3 | Age, alcohol, smoking, exercise, social class | RRmen <60 y = 1.51 (0.69, 3.31)4 RRmen ≥60y = 1.01 (0.57, 1.80)4 RRwomen <60 y = 1.32 (0.38, 4.57)4 RRwomen ≥60y = 2.34 (1.02, 5.40)45 Pooled RR = 1.37 (1.17, 1.65)5P for test of heterogeneity = 0.44 For third tertile vs first tertile |
| Jakobsen et al, 2004 (8) | Both | Total CHD: 326 | Men: 19.7% of total energy Women: 19.5% of total energy | Fat intake as % total energy intake, total energy intake, cohort identification, % energy protein, % energy other fatty acids, family history of MI, smoking, physical activity, education, alcohol, fiber, cholesterol, SBP, BMI | RRwomen <60 y = 4.78 (0.95, 24.10)4 RRwomen ≥60 y = 1.03 (0.53, 2.00)4 RRmen <60 y = 1.01 (0.48, 2.14)4 RRmen ≥60 y = 0.79 (0.48, 1.29)4 Pooled RR = 1.03 (0.66, 1.60) P for test of heterogeneity = 0.61 For third tertile vs first tertile |
| Leosdottir et al, 2007 (14) (Malmo Diet and Cancer Study) | Both | Total CHD: 908 | Men: Fourth quartile, 22.3% of total energy; First quartile, 12.3% of total energy Women: Fourth quartile, 21.8% of total energy; First quartile, 12.2% of total energy | Age, smoking habits, alcohol consumption, socioeconomic status, marital status, physical activity, BMI, fiber intake, and blood pressure. | RRwomen = 0.81 (0.53, 1.24) RRmen = 1.02 (0.76, 1.37) Pooled RR = 0.95 (0.74, 1.21) P for test of heterogeneity = 0.38 For fourth vs first quartile |
| Oh et al, 2005 (33) (Nurses’ Health Study) | Female | Total CHD: 1766 | Fifth quintile: 17.6% of total energy; First quintile: 10.1% of total energy | Age, BMI, cigarette smoking, alcohol intake, parental history of MI, history of hypertension, menopausal status, hormone use, aspirin use, multivitamin use, vitamin E supplement use, physical activity, intakes of energy, protein, cholesterol, MUFAs, PUFAs, trans fat; α-linolenic acid, marine n−3 fatty acids, cereal fiber, and fruit and vegetables | RR = 0.97 (0.74, 1.27) For fifth vs first quintile |
| Tucker et al, 2005 (18) (Baltimore Longitudinal Study of Aging) | Male | Fatal CHD: 71 | Survivors: 12.3% CHD deaths: 13.8% Other deaths: 14.0 % of total energy3 | Age, total energy intake, BMI, smoking, alcohol use, physical activity score, supplement use, fruit and vegetable intakes, secular trend | RRmen <60 y = 0.57 (0.14, 2.30)4 RRmen ≥60 y = 2.31 (0.73, 7.27)4 Pooled RR = 1.22 (0.31, 4.77) P for test of heterogeneity = 0.13 For third tertile vs first tertile |
| Xu et al, 2006 (10) (Strong Heart Study) | Both | Total CHD: 138 | Fourth quartile: 16.5% of total energyFirst quartile: 7.5% of total energy | Variable of interest as % of energy, sex, age, study center, diabetes status, BMI, HDL, LDL, TG, smoking, alcohol consumption, hypertension, energy from protein, total energy intake | RR<60 y = 5.17 (1.60, 16.4)5RR≥60 y = 0.80 (0.41, 1.54)Pooled RR = 1.91 (0.31, 11.84)P for test of heterogeneity = 0.006For fourth quartile vs first quartile |
| Stroke studies | |||||
| McGee et al, 1984 (9) (Honolulu Heart Study)2 | Male | Total stroke: 492 | 12.7% of total energy (age-adjusted)3 | Age, total energy intake, SBP, BMI, smoking, family history of MI, physical activity, intakes of PUFAs, alcohol, protein, carbohydrate, vegetables, and cholesterol | RRmen <60 y = 0.95 (0.60, 1.50)2RRmen ≥60 y = 1.23 (0.66, 2.29)2Pooled RR = 1.04 (0.72, 1.50)P for test of heterogeneity = 0.52For fifth vs first quintile |
| Goldbourt et al, 1993 (35) (Israeli Ischemic Heart Disease Study) | Male | Fatal stroke: 362 | NR | Age, body height, blood pressure, smoking, diabetes | RRmen <60 y = 0.75 (0.54, 1.05)4RRmen ≥60 y = 1.26 (0.70, 2.29)4Pooled RR = 0.92 (0.56, 1.51)P for test of heterogeneity = 0.13For fourth quartile vs first quartile |
| Gillman et al, 1997 (11) (Framingham Study) | Male | Ischemic stroke: 61 | 15.0% of total energy3 | Age, total energy, SBP, cigarette smoking, glucose intolerance, BMI, physical activity, left ventricular hypertrophy, and intakes of alcohol and fruit and vegetables | RRischemic stroke = 0.90 (0.83, 0.96)5For 1% increase in saturated fat |
| Iso et al, 2001 (31) (Nurses’ Health Study) | Female | Hemorrhagic stroke: 74 | Fifth quintile: 36 g/dFirst quintile: 20 g/d | Age, smoking, time interval, BMI, alcohol intake, menopausal status, postmenopausal hormone use, vigorous exercise, usual aspirin use, multivitamins, vitamin E, n−3 fatty acids, calcium, total energy intake, quintiles of cholesterol, MUFAs, PUFAs (linoleic), vegetable protein, trans or unsaturated fat, animal protein, history of hypertension, diabetes, and high cholesterol | RR = 1.05 (0.33, 3.39)For fifth quintile vs first quintile |
| He et al, 2003 (29) (Health Professionals Follow-Up Study) | Male | Total stroke: 598 | Fifth quintile: 31 g/dFirst quintile: 17 g/d | BMI, physical activity, history of hypertension, smoking status, aspirin use, multivitamin use, alcohol consumption, potassium, fiber, vitamin E, fruit and vegetables, total energy, hypercholesterolemia, other fats (MUFAs, PUFAs, and trans fats) | RRmen <60 y = 0.72 (0.35, 1.51)4RRmen ≥60 y = 0.82 (0.49, 1.36)4Pooled RR= 0.79 (0.52, 1.19)P for test of heterogeneity = 0.79For fifth quintile vs first quintile |
| Iso et al, 2003 (30) | Both | Hemorrhagic stroke: 67 | Fourth quartile: 17.1 g/d First quartile: 5.2 g/d | Age, sex, total energy intake, BMI, hypertension, diabetes, total cholesterol, smoking status, ethanol intake, menopausal status (for women) | RR = 0.30 (0.13, 0.71)5 For fourth quartile vs first quartile |
| Sauvaget et al, 2004 (34) (Adult Health Study) | Both | Ischemic stroke: 60 | NR | Age and sex stratified and adjusted for radiation dose, city, BMI, smoking, alcohol, history of hypertension and diabetes | HR = 0.58 (0.28, 1.20) For third tertile vs first tertile |
| Leosdottir et al, 2007 (14) (Malmo Diet and Cancer Study) | Both | Ischemic stroke:648 | Men: Fourth quartile, 22.3% of total energy; First quartile, 12.3% of total energy Women: Fourth quartile, 21.8% of total energy; First quartile, 12.2% of total energy | Age, smoking habits, alcohol consumption, socioeconomic status, marital status, physical activity, BMI, fiber intake, blood pressure | RRwomen = 1.26 (0.81, 1.96) RRmen = 1.19 (0.80, 1.77) Pooled RR = 1.22 (0.91, 1.64) P for test of heterogeneity = 0.85 For fourth vs first quartile |
1
MI, myocardial infarction; MUFA, monounsaturated fatty acid; NR, not reported; PUFA, polyunsaturated fatty acid; SBP, systolic blood pressure; HR, hazard ratio; TG, triglycerides.
2
RR estimates were derived from a provided data set (36).
3
Values represent the mean intake of saturated fatty acid.
4
Data were provided by study investigators on request.
5
Statistically significant relation.