Butter, margarine & total cardiovascular disease (CVD)
Data about this relation was provided by only one cohort (Panagiotakos D). The combined consumption of butter & margarine did not significantly affect CVD risk.
Butter, margarine & coronary heart disease (CHD).
Butter.
Data about 4 different cohorts was found, including 3,250+ cases (the amount of cases was not defined in one cohort).
Results: A significantly increased disease risk was found in one cohort of small size among individuals from all ethnic groups, except Caucasians (Gartside PS).
No other associations were found.
Effect modification: Data about effect modification was provided by one cohort (Gartside PS). Except for a modifying effect of ethnicity, no interactions
were found with serum cholesterol, physical activity, BMI, and several other variables.
Conclusion: Few associations were found. No evidence was found for an association between butter and CHD.
Margarine.
Data about 5 different cohorts was found, including 1,783+ cases (the amount of cases was not defined in one cohort).
Results: A significantly increased disease risk was found in one cohort (Willet WC), while a significant protective effect against mortality was found in another
cohort (Kushi LH). No other associations were found.
Effect modification: No data was found.
Conclusion: Inconsistent findings were done. No evidence was found for an association between margarine and CHD (Average RR = 1.20, but the protective effect
from Kushi LH [8] could not be included in this analysis, because no amount of cases was defined).
Butter vs Margarine.
For this analysis, data was included from within-cohort comparisons of the effect of butter vs margarine. 6 cohorts compared the effect of butter
vs margarine.
Results:
- Holmberg S (2009) OR = 1.10 (0.78-1.56) for low fat margarine or no fat vs butter.
- Iso H (2007) Among men, risk was 0.81 (0.56-1.15) for butter, and 0.89 (0.67-1.17) for margarine. Among women, risk was 0.77 (0.48-1.24) for butter, and 0.90 (0.66-1.22) for margarine.
- Soinio M (2003) Risk was 6.9% in men who used margarine, and 14.6 in men who used butter (P = 0.043). No significant association was found among women.
- Whiteman D (1999) RR = 1.81 (0.87-3.79) for use of hard margarine vs butter (The RR for polyunsaturate vs butter was: 1.23).
- Willet WC (1993) No association was found with butter (no data shown). RR = 1.66 (1.10-2.49; P = 0.02) for margarine)
- Shaper AG (1991) RR = 0.87 (0.79-1.06) for use of butter vs margarine.
Risk was lower from butter consumption than from margarine consumption in one cohort, including 431 cases (Willet WC). In contrast, risk was lower from margarine
than from butter in another cohort among men only, including 41 male cases (Soinio M). No other associations were found, but the RRs were lower for butter than for
margarine in all 4 remaining cohorts.
5 cohorts including 2.056 cases provided RR's, but no RR's were available from one cohort including 65 cases in which opposite findings were done (Soinio M. 2003).
The average RR for use of margarine instead of butter = 1.28.
Conclusion: No significant differences were found between the effects of butter vs margarine on heart disease, but the RRs were lower for butter in 5 out of 6
cohorts. Within-cohort comparisons showed an average RR of 1.28 for use of margarine instead of butter.
| Author | Cohort name | Cases | End point | Relative Risk (RR) |
|---|---|---|---|---|
| 28) Warensjö E (2010) | The NSHDS | Not defined | Risk | No significant association. |
| 19) Iso H (2007) | The JACC Study | 521 men, and 340 women | Mortality | Men: HR = 0.81 (0.56-1.15). Women: HR = 0.77 (0.48-1.24). |
| 12) Gartside PS (1998) | The NHANES I | 1,958 | Risk | A significantly increased risk among all ethnic groups, except Caucasians. |
| 5) Willet WC (1993) | The Nurses' Health Study | 431 | Risk | No significant association. |
| Total number of cases: 3,250 + X |
| Author | Cohort name | Cases | End point | Relative Risk (RR) |
|---|---|---|---|---|
| 19) Iso H (2007) | The JACC Study | 517 men, and 333 women | Mortality | Men: HR = 0.89 (0.67-1.17). Women: HR = 0.90 (0.66-1.22). |
| 8) Kushi LH (1996) | The Iowa Women's Health Study | Not defined | Mortality | RR = 0.59 (0.36-0.97; P = 0.048). |
| 6) Knekt P (1994) | The Finnish Mobile Clinic Health Cohort | 186 men, and 58 women | Mortality | Men: RR = 0.85 (0.59-1.24; P = 0.39). Women: RR = 0.88 (0.44-1.74; P = 0.77). |
| 5) Willet WC (1993) | The Nurses' Health Study | 431 | Risk | RR = 1.66 (1.10-2.49; P = 0.02). |
| 4) Shaper AG (1991) | The British Regional Heart Study | 258 | Risk | RR = 0.57 (0.22-1.46) for no use vs use. |
| Total number of cases: 1,783 + X | Average RR = 1.20 |
Margarine, butter & stroke.
4 articles, providing information about 4 different cohorts were found. The relation with total stroke mortality was examined in 2 cohorts (Yochum LA;
Iso H 2007), and the relation with subtypes of stroke risk was examined in the other 2 cohorts (Iso H 2001; Larsson SC).
Results: The association with margarine was examined in 3 cohorts, including 2,184 cases (Iso H 2001; Yochum LA; Iso H 2007). Protective
effects were found among women in all 3 cohorts, which were significant twice (Iso H 2001; Iso H 2007). No association was found among men in the only
available cohort (Iso H 2007).
The association with butter was examined in 3 cohorts, including 5,269 cases (Iso H 2001; Iso H 2007; Larsson SC). Increased risks of intracerebral hemorrhage
were found in both cohorts which examined the association (Iso H 2001; Larsson SC). In one cohort the risk, but not the trend was significant (Larsson SC).
No other associations were found.
Inclusion of intermediate levels of consumption:
Significant protective effects of margarine consumption against total stroke mortality among women were found at consumption ≥ 1 time/week in one cohort
(Iso H 2007), and at the intermediate level of consumption 4-7 times/week in the other cohort (Yochum LA).
Effect modification: No data was found.
Conclusion: Protective effects of margarine against stroke mortality were found among women in 2 cohorts of moderate-large size. Margarine possibly protects
against total stroke mortality among women at consumption 4-7 times/week (Average RR = 0.73).
Increased intracerebral hemorrhage risks of butter were found in 2 cohorts of moderate-large size. And no other associations were found. Butter possibly increases
risk of intracerebral hemmorhage. The level of consumption for - and strength of - this effect could not be defined.
| Author | Cohort name | Cases | End point | Relative Risk (RR) |
|---|---|---|---|---|
| 23) Larsson SC (2009) | The ATBC Study | 2,702 ischemic stroke risk. 383 intracerebral hemorrhage risk. 196 subarachnoid hemorrhage risk | Stroke risk | Ischemic stroke: RR = 1.00 (0.87-1.14; P = 0.99). Intracerebral hemorrhage: RR = 1.44 (1.01-2.07; P = 0.19). Subarachnoid hemorrhage: RR = 0.98 (0.59-1.64; P = 0.87). |
| 19) Iso H (2007) | The JACC Study | 1,040 men, and 874 women | Stroke mortality | Men: HR = 0.96 (0.76-1.22). Women: HR = 0.96 (0.75-1.24). |
| 5) Iso H (2001) | The Nurses' Health Study | 74 | Intracerebral hemorrhage risk | A positive association. |
| Total number of cases: 5,269 | Average RR = 1.02 |
| Author | Cohort name | Cases | End point | Relative Risk (RR) |
|---|---|---|---|---|
| 19) Iso H (2007) | The JACC Study | 1,038 men, and 857 women | Stroke mortality | Men: HR = 0.99 (0.82-1.19). Women: HR = 0.77 (0.63-0.95; P = < 0.05). |
| 8) Yochum LA (2000) | The Iowa women's Health Study | 215 | Stroke mortality | RR = 0.64 (0.41-1.00; P = 0.10). |
| 5) Iso H (2001) | The Nurses' Health Study | 74 | Intracerebral hemorrhage risk | A significant inverse association. |
| Total number of cases: 2,184 | Average RR = 0.86 |
| Author | Cohort name | Subjects | Years of follow-up | Cases | End point | Consumption of | Relative Risk (RR) | Adjustments |
| 22) Panagiotakos D (2009) | The ATTICA Study | 2,101 men and women aged 18-89 without CVD. (Greece) | 5 (2001-2002 to 2006) | 170 (108 men and 62 women) | Cardiovascular disease incidence (fatal or non-fatal acute myocardial infarction, ustable angina, or any other type of CVD) | Margarine/butter | A similar % of cases used margarine/butter in cooking (26%) as noncases (20%; P = 0.28). | Age. |
| Author | Cohort name | Subjects | Years of follow-up | Cases | End point | Consumption of | Relative Risk (RR) | Adjustments | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 28) Warensjö E (2010) | The Northern Sweden Health and Disease Study (NSHDS). (Including 3 cohorts: the VIP, the MONICA, and the MSP) | 73,000 subjects from 3 subcohorts, and whithout previous MI, stroke or malignant disease. (Nested case-control) | (1987-1999) | See variables | First MI | Butter |
Case-control match by sex, age, date of health-survey, and geographic region. |
26) Holmberg S (2009) | No cohort name defined | 1,752 farmers and non-farming rural men without a history of CHD. | (Sweden) 12 | (1990-91 to 2002-03) 138 | Coronary heart disease risk (hospitalized or deceased due to CHD) | Spread on sandwiches | OR = 1.10 (0.78-1.56) for low fat margarine or no fat vs butter. | Unadjusted. |
19) Iso H (2007) | The JACC Study. | 39,528 men, and 54,172 women. | (Japan) Not defined. | 521 men, and | 340 women. Ischemic heart disease mortality | Butter |
|
Age and study area. |
19) Iso H (2007) | The JACC Study. | 39,507 men, and 54,297 women. | (Japan) Not defined. | 517 men, and | 333 women. Ischemic heart disease mortality | Margarine |
|
Age and study area. |
16) Soinio M (2003) | No cohort name defined. | 366 men and 295 women with diabetes type 2, but free from CHD, aged 45-64 from Kuopio. | (Finland) 7 | (1982-84 to 1989) See variables | CHD death | Use of margarine vs butter |
|
Age and diabetes duration. |
13) Whiteman D. (1999) | The OXCHECK Study. | 10,522 men and women aged 35-64 without a previous history of angina. | (UK) 9 | (1989-1997) 107 | IHD mortality | Spread | RR = 2.00 (1.14-3.50; No P-value) for hard margarine (e.g. stork), and RR = 1.23 (0.79-1.93; No P-value) for polyunsaturate (e.g. flora, sunflower) (reference category = butter). | The association with margarine remained after additional adjustment for fresh or frozen green vegetables/salad, pudding/cakes/biscuits/sweets, and fresh or frozen red meat: RR = 1.81 (0.87-3.79).
Gender, smoking and age. |
12) Gartside PS (1998) | The NHANES I | 5,811 men and women aged 40-74 with and without CHD and not using any special diets. | 16 | (1971-1987) 1,958 | CHD risk (morbidity and mortality) | Butter | There was an ethnic group-butter interaction such that in African-Americans and in American Indians, Hispanics, and Asians, CHD events increased as butter intake rose (p=.026, p=.015). Butter intake was not positively associated with CHD event rates in Caucasians | Effect modification: No significant interactions were found with Age, gender, geographic region, serum cholesterol, physical exercise, physical activity, education, smoking, BMI, alcohol, fish, bread, cheese, and dessert intake. Not defined, but possibly: Age, race, gender, geographic region, serum cholesterol, physical exercise, physical activity, education, smoking, BMI, alcohol, fish, bread, cheese, and dessert intake. |
8) Kushi LH (1996) | The Iowa Women's Health Study. | 19,411 postmenopausal women aged 55-69 who did not take vitamin supplements. | (USA) 7 | (1986-1992) Not defined. | CHD death | Margarine |
RR = 0.59 (0.36-0.97; P = 0.048) for the highest vs lowest quartile of consumption. | Amount specific data (times/wk): 0: RR = 1. 0.5-3: RR = 0.73 (0.41-1.30). 4-7: RR = 0.60 (0.36-1.02). > 7: RR = 0.59 (0.36-0.97). Age, total energy intake, BMI, waist-to-hip ratio, pack-years of cigarette smoking, hypertension, diabetes mellitus, oral-contraceptive use, estrogen-replacement therapy, physical activity, alcohol intake, marital status, and level of educational attainment. |
6) Knekt P (1994) | The Finnish Mobile Clinic Health Cohort. | 5,133 men and women aged 30-69. | 14 | (1968-72 to 1984) 186? men, and 58? women. | CHD mortality | Margarine |
|
Age, smoking, serum cholesterol, hypertension, BMI, and energy intake. |
5) Willet WC (1993) | The Nurses' Health Study. | 85,095 women. | (USA) 8 | (1980-1988) 431? | CHD risk | (nonfatal MI or death from CHD) Margarine |
RR = 1.66 (1.10-2.49; P = 0.02) for the highest vs lowest sextile of consumption. | Amount specific data (teaspoon; 5 ml): < 1 per month: RR = 1. 1-4 per month: RR = 0.83 (0.50-1.38). 2-4 per wk: RR = 1.03 (0.71-1.50). 5-7 per wk: RR = 1.07 (0.80-1.42). 2-3 per day: RR = 1.03 (0.77-1.38). ≥ 4 per day: RR = 1.66 (1.10-2.49). Similar positive trends were seen for margarine in tub form or stick form separately (no data shown). Age, time period, BMI, alcohol, smoking, menopausal status, postmenopausal hormone use, family history of MI before age 60, history of hypertension, multivitamin use. |
5) Willet WC (1993) | The Nurses' Health Study. | 85,095 women. | (USA) 8 | (1980-1988) 431? | CHD risk | (nonfatal MI or death from CHD) Butter | Not significantly associated with risk of CHD (no data shown). | Age, time period, BMI, alcohol, smoking, menopausal status, postmenopausal hormone use, family history of MI before age 60, history of hypertension, multivitamin use. |
4) Shaper AG (1991) | The British Regional Heart Study | 7,735 men aged 40-59 without a diagnosis of IHD at screening. | 9.5 | See variables | Ischaemic heart disease incidence | Margarine and butter |
|
Age, social class, cigarette smoking, and blood cholesterol concentration. |
|