Abstract

OBJECTIVES:

• To review all prospective studies which published information about eggs in relationship with all-cause mortality.
• To define the amount of consumption found to be related with possible effects on all-cause mortality.
• To define possible effect modification by confounders.

DATA SOURCE: The Pubmed database was searched (No start date - May 25th, 2010) for relevant articles using the keywords "dietary cholesterol, dietary fat, egg, or eggs" combined with "prospective, cohort, follow-up, or longitudinal". The exact search term is described Here.
Prospective studies published in the English language were included. Reference lists were searched for additional articles.
RESULTS: 15 articles were found which provided information about 12 different cohorts. Of these, 0 articles were excluded.

• Overall analysis.
  13 articles were found providing data about 12 different cohorts, and including 41,892 deaths. For high vs low consumption, significant associations in both directions were found.
• High- vs median level consumption.
  Some protective effects were found at the level of consumption of 1-6 eggs/week, while increased risks were almost restricted to higher consumption.
• Analysis stratified by sex or geographic area.
  No consistent difference in effect was found between men and women. When both Asian cohorts were excluded, significantly increased risks of high vs low consumption were found in 3 cohorts of moderate-large size, and including 62% of all cases.
• Effects among subjects with pre-existing disease.
  A significantly increased death risk of high vs low egg consumption was found among subjects with existing diabetes mellitus in 2 out of 3 cohorts.

CONCLUSION: Inconclusive evidence was found for an overall association between eggs and all-cause mortality at high vs low consumption. Possible heterogeneity may exist between high- vs median level egg consumption. Suggestive evidence was found for a weak (- 6%) protective effect against death at median levels of egg consumption (1-6 eggs/week), while increased risks were almost restricted to higher consumption.
No consistent difference in effect was found between men and women. High consumption of eggs (≥ 7 eggs/week) possibly increases risk of all-cause mortality among US & European subjects, but the strength of the effect is weak (+ 3%). Also, high consumption possibly increases risk of all-cause mortality among subjects with diabetes mellitus (+ 61%).
LIMITATIONS: Though some evidence was found for effects between egg consumption and death, results were not very consistent. Therefore, future research should confirm these findings before any conclusions may be drawn.

Eggs and all-cause mortality.

15 articles, providing information about 12 different cohorts were found. Survival was the end point in one of these cohorts (3).

Use of data: For the overall analysis, results from 2 articles were not included. Data about these 2 articles can be found in the "extended version of the table". About the 2 excluded articles:

• Trichopoulou A (2003). Examined 22.043 subjects aged 20-86 from The Greek part of The EPIC Study, which included 275 deaths.
• Trichopoulou A (2006). Examined 1,013 subjects with diabetes mellitus from The Greek part of The EPIC Study, which included 80 deaths.

Reason for exclusion in this analysis is as follows:
In 2005 Trichopoulou published results from the EPIC-Elderly Study, examining subjects aged ≥ 60 from 9 European countries, and including data from EPIC-Greece (6b). Most deaths from the first publication (2003) about The Greek part of The EPIC Study will probably have occured among subjects aged ≥ 60, and therefore inclusion of the data would have led to using the same data twice. In addition, subjects with a history of diabetes mellitus were not excluded from analysis in the EPIC-Elderly Study, so inclusion of the subjects with diabetes mellitus from EPIC-Greece (2006) would also have led to using the same data twice.
But the EPIC-Elderly study excluded subjects with prevalent CVD. A third publication about the EPIC-Greece cohort included subjects with prevalent CHD only (6a), and was therefore included in the analysis.

Results: For the overall analysis, data from 13 articles about 12 different cohorts was used. A total of 41,892 deaths occured among these cohorts.

• A significantly increased risk was found in 3 cohorts (2, 6b, 12), and a significantly increased risk (trend) was found in a 4th cohort among women, but not men (7). In addition, a nonsignificantly increased risk was found in a 5th cohort among men, but not women (3). The amount of cases in these analysis was 16,595.
• A significant protective effect was found in another cohort among both men and women, including 16,073 cases (10).
• No associations were found in the remaining cohorts (1, 4, 5, 6a, 8, 9, 11).

Conclusion: For high vs low consumption, significant associations in both directions were found. And the total amount of cases from the cohorts in which an increased risk was found, was appoximately equal to the amount of cases from the cohort in which a protective effect was found. Inconclusive evidence was found for an overall association between eggs and all-cause mortality at high vs low consumption (Excluding incomplete data from Dawber TR [1]: Average RR = 1.04).

Inclusion of intermediate levels of consumption. And analysis stratified by sex.

Protective effects were found in the following cohorts:

• Nube M (1,190 cases). Both men and women had higher survival rates at consumption of 2-3 eggs/week.
• Mann JI (381 cases). In analysis among men and women combined, a significant protective effect was found at consumption of 1-5 eggs/week.
• Nakamura Y (562 female cases) A significant protective effect among women only was found at consumption of 1-2 eggs/week.
• Iso H (16,073 cases). A significant protective effect was found among men at ≥ 3 eggs week, and among women at ≥ 5 eggs/week.

Increased risks were found in the following cohorts:

• Kahn HA (6,075 cases). A significantly increased risk was found at consumption of ≥ 6 eggs/week. In stratified analysis, the association was found among women only.
• Nube M (742 male cases). Men only had lower survival rates at the highest tertile of consumption: 4-16 eggs/week.
• Nakamura Y (562 female cases). Among men only, a significantly increased risk was found at consumption 1-2 eggs/week. Among women only, the trend, but not the RR was significantly increased. Risk increased at consumption of ≥ 14 eggs/week.
• Djoussé L (5,169 male cases). Among men, risk increased significantly at consumption of ≥ 7 eggs/week.

Significantly increased risks were found at levels approximating the level of consumption of 1 egg/day in 4 cohorts, including 12,548 cases. While protective effects were mostly found at the level of consumption of 1-5 eggs/week. The latter analysis included 4 cohorts, and 18,206 cases (44% of all cases).

Calculation of average RR's for intermediate levels of consumption: An attempt was made to calculate RR's for several intermediate levels of consumption: 1-2 eggs/week, 3-4 eggs/week, and 1-6 eggs/week. For the latter analysis, a simplistic approach was used: If RR's were provided for different units of consumption at the range of 1-6 eggs/week, the average RR for these units per cohorts was included in the analysis. This was done to increase the statistical power of the analysis, because 2 cohorts (5, 9) provided RR's at wide intermediate ranges of consumption only (1-5/6 eggs/week).

Conclusion: Possible heterogeneity may exist in the effect of egg consumption. This is created by the level of consumption: Some protective effects were found at the level of consumption of 1-6 eggs/week, while increased risks were almost restricted to higher consumption. Suggestive evidence was found for a protective effect against death at median levels of egg consumption (Average RR = 0.94 for 1-6 eggs/week), while the evidence for higher levels of consumption (≥ 7 eggs/week) remains contradictive. No consistent difference in effect was found between men and women.

Analysis stratified by geographic area.

Significant protective effects were found only in one Japanese cohort of fairly large size (Iso H). When both Asian cohorts (Iso H; Nakamura Y) are excluded from the analysis, 10 cohorts remain among US and European subjects, and including 25,807 deaths.
In this analysis, significantly increased risks of high vs low consumption were found in 3 cohorts (Kahn HA; Trichopoulou A [2005]; Djoussé L) and in a 4th cohort among men (Nube M), including a total of 16,033 cases (62% of all cases). And no opposing effects were found. But this analysis shows the average RR was similar to the average RR for all cohorts (Excluding incomplete data from Dawber TR [1]: RR = 1.03).

Conclusion: When Asian cohorts were excluded, significantly increased risks of high vs low consumption were found in 3 cohorts of moderate-large size, and including 62% of all cases. Among US and European cohorts, high consumption of eggs possibly increases risk of all-cause mortality (+ 3%).

Effects among subjects with existing disease.

Subjects with diabetes: The relation between egg consumption and death among subjects with diabetes mellitus was examined in 3 cohorts, including 596 cases.
Significantly increased risks were found in 2 cohorts, including 343 cases (6, 12). No significant association was found in the remaining cohort, but the RR was above 1 (9).
Subjects with CHD: The relation between egg consumption and death among subjects with CHD was examined in 1 cohort, including 131 cases (See extended table: Trichopoulou A. 2005b). No association was found (MR = 0.93; 95% CI = 0.75-1.15).

Conclusion: A significantly increased death rate of high vs low egg consumption was found among subjects with existing diabetes mellitus in 2 out of 3 cohorts. Among subjects with diabetes mellitus, high consumption of eggs possibly increases risk of all-cause mortality (+ 61%).