Associations According to Risk Factor Status

To examine the possibility that a positive association with egg intake is limited to certain subgroups, we conducted additional multivariate analyses stratified by risk factor status including hypercholesterolemia, diabetes, hypertension, smoking, alcohol use, body mass index, age, vitamin supplement use, parental history of MI, and intakes of saturated fat, polyunsaturated fat, and carbohydrates. We found no evidence of a positive association with higher consumption of eggs in any subgroup except a suggestion that the risk might be elevated among individuals with diabetes. Among diabetic men, the multivariate RRs of CHD across categories of intake were less than 1 per week (1.0), 1 per week (1.0), 2 to 4 per week (1.16), 5 to 6 per week (1.16), and 1 or more eggs per day (2.02); (95% CI, 1.05-3.87; P=.04 for trend and P=.18 for interaction between egg consumption and diabetes status). The corresponding RRs for diabetic women were 1.0, 0.91, 1.05, 1.87, and 1.49 (95% CI, 0.88-2.52; P=.008 for trend and P=.07 for interaction). To investigate the possibility that an effect of egg consumption may be more apparent among those with a low-background cholesterol diet,36 we conducted analyses stratified by dietary cholesterol from foods other than eggs. The RRs for more than 1 eggs per day were 1.05 (95% CI, 0.61-1.79) for the men and 0.97 (95% CI, 0.13-7.10) for the women whose non-egg cholesterol intake was low (mean intakes were 88.4 mg/4200 kJ for men and 118.8 mg/4200 kJ for women), compared with 0.97 (95% CI, 0.64-1.46) for the men and 0.79 (95% CI, 0.57-1.11) for the women whose non-egg cholesterol intake was relatively high (mean intakes were 135.9 mg/4200 kJ for men and 175.9 mg/4200 kJ for women).

The Association With Stroke

We documented 258 incident cases of stroke in men during 8 years of follow-up and 563 cases of stroke in women during 14 years of follow-up. In multivariate analyses (Table 3 <../fig_tab/joc81683_t3.html>), egg consumption was not significantly associated with risk of total stroke; the RRs for 1 egg or more per day were 1.07 (95% CI, 0.66-1.75; P for trend=.50) for men and 0.89 (95% CI, 0.60-1.31; P for trend=.77) for women. Adjustment for intake of bacon, vegetables, and fruit further attenuated the association for men (RR for 1 egg per day, 1.00; 95% CI, 0.57-1.78; P for trend = 0.95). In both cohorts, no significant association was observed between egg consumption and risk of ischemic or hemorrhagic stroke.

COMMENT


In these 2 large prospective cohort studies of men and women, we found no overall significant association between egg consumption (up to 1 egg per day) and risk of CHD or stroke. We specifically found no evidence for a significant increase in risk with either recent or relatively long-term (over the past decade) egg consumption. Despite somewhat different patterns of egg consumption in men and women, the results from the 2 cohorts were remarkably consistent.

In subgroup analyses, egg consumption appeared to be associated with increased risk of CHD among individuals with diabetes. This result should be interpreted cautiously due to numerous subgroup analyses, but the consistency of the association in the 2 cohorts argues against a chance finding. The increased risk may be related to abnormal cholesterol transport due to decreased levels of apolipoprotein E37 and increased levels of apolipoprotein C-III38 among patients with diabetes.

We considered the possibility that inaccurate self-reports of egg consumption or confounding by intake of other foods could explain the observed null results. Egg consumption was reported on food frequency questionnaires with relatively high accuracy (correlations were 0.8 between the self-report and multiple week dietary records in our validation studies28, 29). Also, egg consumption was assessed several times in both cohorts so that our analyses using updated dietary information could dampen measurement error and take into account changes in eating behavior.

Several metabolic studies have suggested a hypocholesterolemic effect of decholesterolized eggs (ie, Eggbeaters) on blood cholesterol levels compared with whole eggs.39, 40 We were not able to examine the effect of such products on the risk of CHD because they were not included in the dietary questionnaires on the printed form. However, our questionnaires permit respondents to report other foods that are frequently consumed. In NHS, only 48 women recorded consuming Eggbeaters in the 1984 dietary questionnaire and 105 recorded this in the 1990 questionnaire.

Egg consumption was positively associated with smoking, lower physical activity, and a generally unhealthy eating pattern (ie, more whole milk, red meat, and bacon and less skim milk, vegetables, and fruits) in men. Confounding due to these factors would artifactually produce an elevated risk for egg consumption. As expected, an apparent positive association with higher egg consumption in the age-adjusted analysis in men was attenuated after adjustment for smoking and other covariates. After further adjusting for bacon intake, which was positively associated with risk of CHD in our cohorts, the RRs became weakly inverse. This speaks to the importance of considering overall eating patterns when examining the effects of egg consumption.

It is possible that participants with high serum cholesterol levels were more likely to reduce their egg intake than others, which might obscure a positive association between egg consumption and risk of CHD. However, in our primary analyses, we excluded subjects with diagnosed hypercholesterolemia at baseline. Also, in our analyses by using updated dietary information, we stopped updating egg consumption at the beginning of the time interval during which individuals reported hypercholesterolemia to avoid confounding due to change in diet during follow-up.41 Finally, we did not measure blood cholesterol levels in our cohorts. However, blood cholesterol should not be controlled in the analyses as it is an intermediate variable when assessing the relationship between dietary cholesterol and CHD.

In controlled metabolic studies, ingestion of cholesterol by eating egg yolks or whole eggs raises serum total and LDL cholesterol levels.5-7, 42 In most egg feeding studies, intakes of other nutrients such as fatty acids, carbohydrates, and protein were balanced between egg and no egg groups so that only dietary cholesterol varied. In our cohorts, participants who consumed more eggs had lower intakes of carbohydrates, suggesting that, in reality, people often substitute eggs for carbohydrate-rich foods such as breakfast cereals. The effects of egg cholesterol on blood cholesterol can be predicted from well-established equations derived from metabolic studies. A 50-g egg contains about 213 mg of cholesterol, 6 g of protein, and 5 g of fat.43 Of the fat, nearly half is monounsaturated fat and 16% is polyunsaturated fat. The equation derived by Keys and Parlin4 predicts that adding 1 egg to an average diet (assume 200 mg background cholesterol and 7560 kJ/d) will result in about a 4% increase in total serum cholesterol for a normocholesterolemic person (assume total blood cholesterol of 5.17 mmol/L [200 mg/dL]). If we assume that raising cholesterol levels is the only effect of egg consumption, this would translate into about an 8% increase in CHD risk,44 an effect generally too small to be detectable in this and most epidemiologic studies or clinical trials. In NHS, dietary cholesterol (but not eggs) was nonsignificantly associated with CHD,17 raising the possibility that eggs contain other nutrients that may be beneficial in preventing CHD.

The equation developed by Mensink and Katan45 predicts that substituting fatty acids from 1 egg for carbohydrates would raise HDL cholesterol by about 2% (assume HDL level of 1.03 mmol/L [40 mg/dL]) and decrease triglycerides also by about 2% (assume triglyceride level of 2.82 mmol/L [250 mg/dL]). The effects of egg consumption on raising HDL levels have been observed in some metabolic studies,19, 20 but not in others.36, 40 Also, in egg feeding studies, triglyceride levels were significantly reduced in some,19, 20 but not other studies.36, 46 In particular, Schnohr et al19 observed a reduction of triglycerides by 0.09 mmol/L (7.97 mg/dL) by adding 2 eggs to the usual diet. Packard et al20 observed a reduction of triglycerides by 0.19 mmol/L (16.8 mg/dL) by adding 6 eggs to the diet. In addition, egg intake decreased blood glycemic and insulinemic responses,47 especially when egg yolk was ingested (compared with whole egg or egg white).21 This might result from further delayed gastric emptying after yolk ingestion. Moreover, adding eggs to pasta produced lower insulin and C-peptide responses.48 Holt et al49 tested 38 common foods and found that eggs were among the foods that have the lowest glycemic and insulin indexes.