Type 2 diabetes mellitus is increasing in prevalence globally( Reference Zimmet, Alberti and Shaw 1 ) and is becoming a world pandemic. The total number of people with diabetes is projected to rise from 171 million in 2000 to 366 million in 2030( Reference Wild, Roglic and Green 2 ). The number of people with diabetes is increasing due to population growth, ageing, urbanization and the increasing prevalences of obesity and physical inactivity( Reference Wild, Roglic and Green 2 ). The prevalence of type 2 diabetes mellitus is 4·5 % among people aged 35–65 years in Lithuania( Reference Norkus, Domarkienė and Šulcaitė 3 ).
Type 2 diabetes appears to involve an interaction between susceptible genetic backgrounds and environmental factors including highly calorific diets( Reference Naser, Gruber and Thomson 4 ). Eggs are a major source of dietary cholesterol, with an average egg containing about 200 mg of cholesterol. On the other hand, eggs contain other nutrients such as minerals, folate, B vitamins, fat-soluble vitamins A, D, E and K, carotenoids and saturated, monounsaturated and polyunsaturated fatty acids( Reference Song and Kerver 5 – Reference Herron and Fernandez 7 ).
In the last decade, usage of eggs increased in Lithuania (from 176 to 191 eggs per capita)( 8 ). In food ration eggs are often used, because poultry farming is highly developed and eggs are one of the cheapest food types in Lithuania.
As consumption of eggs is related to the development of some chronic non-infectious diseases( 9 ), we decided to investigate if an increase in incidence of type 2 diabetes mellitus could partly be related to habits of egg consumption. Thus we performed a case–control study in an out-patient clinic in Kaunas, Lithuania, with the aim to determine the relationship between egg consumption and risk of type 2 diabetes mellitus.
Materials and methods
A case–control study was carried out at an out-patient clinic in Kaunas, Lithuania, in 2001. The study included 234 cases aged 35–86 years with a newly confirmed diagnosis of type 2 diabetes mellitus according to WHO criteria( 10 ). In total, 468 controls who had neither impaired fasting glucose level nor type 2 diabetes mellitus after a glucose tolerance test were recruited from patients of the same clinic. They were individually matched to the diabetic patients by gender and age (± 5 years). The ratio of cases to controls was 1:2.
Information on age, gender, family history of diabetes, education, occupational and marital status, egg consumption, alcohol consumption, cigarette smoking, physical activity and stress level was assessed by a special questionnaire designed by our research group. All study participants were asked to fill out the questionnaire by themselves.
Participants were asked to fast for 12 h and to avoid smoking and heavy physical activity for at least 2 h before the examinations. After a 5 min rest in a quiet room, systolic and diastolic blood pressure levels were measured twice at an interval of a few minutes on the right arm with a standard mercury sphygmomanometer with 2·0 mmHg column accuracy. Anthropometric measurements were made according to the guidelines of WHO( 11 ). Height and weight were measured twice. Height was measured without shoes in centimetres (0·1 cm accuracy). Weight was measured with light clothing in kilograms (0·5 kg accuracy). BMI (kg/m2) was calculated as weight (in kilograms) divided by the square of height (in metres)( Reference Garrow and Webster 12 ). Waist circumference was measured by holding a non-stretchable measuring tape snugly around the waist, defined as the midpoint between the bottom rib and the tip of the hipbones, and hip circumference was measured at the level of the great femur trochanters; both were measured in centimetres (0·1 cm accuracy).
Laboratory blood tests included fasting blood samples drawn from participants’ elbow vein and venous plasma samples analysed for glucose and TAG levels. Plasma glucose was estimated by the GOD–PAP method (EBIO compact analyser; Eppendorf, Germany). Plasma TAG was estimated by the GPO–PAP method (RA-50 clinical chemistry analyser; Bayer Diagnostics, Ireland). According to the recommendations of WHO( 10 ), a 75 g oral glucose tolerance test for assessing carbohydrate disorders was performed and evaluated in each study participant.
Study variables were categorized as follows.
-
1. Egg consumption: <1 egg/week, 1–1·9 eggs/week, 2–2·9 eggs/week, 3–4·9 eggs/week and ≥5 eggs/week.
-
2. BMI: 18·5–24·9 kg/m2, 25·0–29·9 kg/m2 and ≥30·0 kg/m2.
-
3. Family history of diabetes: first-degree relatives with diabetes and first-degree relatives without diabetes.
-
4. Cigarette smoking: non-smoker, ex-smoker, infrequent smoker and current smoker.
-
5. Level of education (number of years): ≤10 years, 11–13 years and ≥14 years.
-
6. Morning exercise (of at least 30 min duration) during the last 12 months: no, sometimes and yes.
-
7. Plasma TAG: <1·7 mmol/l and ≥1·7 mmol/l.
Conditional logistic regression was used to calculate odds ratios and corresponding 95 % confidence intervals for diabetes mellitus in relation to exposures of interest. Variables such as BMI, family history of diabetes, cigarette smoking, education, morning exercise and plasma TAG level were retained in models as confounders when their inclusion changed the value of the odds ratio by more than 10 % in any exposure category.
All reported trend test significance levels (P values) were two-sided( Reference Woodward 13 ). The χ 2 test was utilized to calculate the difference between proportions. The level of significance was set at 5 %. All calculations were performed with the STATA statistical software package version 7 (StataCorp., College Station, TX, USA).
Results
The characteristics of the case (type 2 diabetic patients) and control groups are displayed in Table 1. In our study there were 28·21 % men and 71·79 % women. The mean age was 64·09 (sd 7·85) years (range: 39–86 years) for men and 65·23 (sd 8·3) years (range: 34–86 years) for women. The cases had significantly lower education level compared with the controls. Their BMI was higher than in the controls’. There were more controls without a family history of diabetes (first-degree relatives without diabetes) than cases.
Univariate regression modelling showed that participants who consumed ≥5 eggs/week had a higher risk of type 2 diabetes mellitus (crude OR = 1·27; 95 % CI 1·10, 1·46) than those who consumed <1 egg/week. A dose–response relationship was defined between risk of this disease and egg consumption (P for trend = 0·001).
Variables such as BMI, family history of diabetes, cigarette smoking, education, morning exercise and plasma TAG level were retained in multivariate logistic regression models as confounders because their inclusion changed the value of the odds ratio by more than 10 % in any exposure category. Multivariate logistic regression data showing the relationship between type 2 diabetes mellitus and egg consumption are presented in Table 2. After adjusting for BMI, family history of diabetes, smoking and education, it was found that the excess risk of type 2 diabetes mellitus because of egg consumption changed a little, but remained statistically significant. After further controlling for BMI, family history of diabetes, smoking, education, morning exercise and plasma TAG level, we found a threefold increased risk of type 2 diabetes mellitus in participants who consumed ≥5 eggs/week (OR = 3·02; 95 % CI 1·14, 7·98) compared with those eating <1 egg/week.
*OR adjusted for BMI, family history of diabetes, smoking and education level.
†OR adjusted for BMI, family history of diabetes, smoking, education level, morning exercise and plasma TAG level.
Discussion
The present study has demonstrated an increased risk of type 2 diabetes mellitus related to egg consumption. We found that consumers of ≥5 eggs/week had three times greater risk of type 2 diabetes mellitus than those eating <1 egg/week. Prospective studies have shown similar results. Djoussé et al.( Reference Djoussé, Gaziano and Buring 14 ) found that multivariable-adjusted hazard ratios for type 2 diabetes among individuals who consumed ≥7 eggs/week were 1·58 in men and 1·77 in women, compared with no eggs consumed. In their study of 2849 Chinese adults (aged ≥20 years), Shi et al.( Reference Shi, Yuan and Zhang 15 ) found that egg consumption was positively associated with the risk of diabetes. Participants who consumed ≥1 egg/d had 2·28 (95 % CI 1·14, 4·54) times greater risk of type 2 diabetes mellitus than those who consumed <2 eggs/week, and women who consumed ≥1 egg/d had three times greater risk( Reference Shi, Yuan and Zhang 15 ). Djoussé et al.( Reference Djoussé, Kamineni and Nelson 16 ) studying 3898 older men and women from the Cardiovascular Health Study did not find the association between egg consumption or dietary cholesterol and increased risk of incident type 2 diabetes.
Salmeron et al.( Reference Salmeron, Hu and Manson 17 ) found that total fat, SFA and MUFA intakes were not associated with risk of type 2 diabetes in women, but trans-fatty acids increased the risk and PUFA reduced the risk.
Eggs are a source of high-quality protein, with little total fat. Compared with other animal protein sources, eggs contain proportionately less saturated fat, which has generally been recognized as a strong dietary determinant of elevated LDL levels( Reference Mente, de Koning and Shannon 18 ). Epidemiological studies suggest that, among hyper-responders, dietary cholesterol from eggs leads to a modest increase in serum LDL cholesterol (LDL-C) and HDL cholesterol (HDL-C) concentrations and no effect on the ratio of LDL-C to HDL-C( Reference Weggemans, Zock and Katan 19 – Reference McNamara 23 ). It is estimated that each additional 100 mg of dietary cholesterol intake results in an increase of 1·9 mg/dl in LDL-C and 0·4 mg/dl in HDL-C( Reference McNamara 23 ). Dietary cholesterol increases the susceptibility of LDL to oxidation, increases postprandial lipidaemia and potentiates the adverse effects of dietary saturated fat. Dietary cholesterol, including egg yolks, is harmful to the arteries( Reference Spence, Jenkins and Davignon 24 ). Endothelial function refers to arterial vasomotor responses mediated predominantly by the release of nitric oxide (vasodilating) and endothelin (vasoconstricting) from the vascular endothelium, and plays an important role in the pathogenesis of atherosclerosis, hypertension, CVD and diabetes( Reference Vogel 25 , Reference Njike, Faridi and Dutta 26 ).
Feskens and Kromhout( Reference Feskens and Kromhout 27 ) studied 394 non-diabetic men (aged 50–70 years) in the Zutphen Study and found that saturated fat and dietary cholesterol intakes may affect glucose tolerance detrimentally. However, in their investigation of twenty-eight overweight or obese male patients (aged 40–70 years) on a carbohydrate-restricted diet, Mutungi et al.( Reference Mutungi, Ratliff and Puglisi 28 ) found that use of 3 eggs/d had no effects on fasting glucose compared with no egg consumption. On the other hand, Pearce et al.( Reference Pearce, Clifton and Noakes 29 ) studied sixty-five participants with type 2 diabetes or impaired glucose tolerance and investigated the effect of a hypoenergetic high-protein/high-cholesterol diet v. a diet with a similar amount of animal protein (high-protein/low-cholesterol) on plasma lipids, glycaemic control and cardiovascular risk markers. They found that the high-protein energy-restricted diet high in cholesterol from eggs improved glycaemic and lipid profiles, blood pressure and apo-B in individuals with type 2 diabetes( Reference Pearce, Clifton and Noakes 29 ).
Djoussé and Gaziano( Reference Djoussé and Gaziano 30 ) from the Physicians’ Health Study reported that egg consumption was positively related to total mortality in male physicians. This association was stronger among diabetic participants, who had a twofold increased risk of death (2·01; 95 % CI 1·26, 3·20) when comparing the highest with the lowest category of egg consumption v. non-diabetic participants.
Our work has some limitations that need to be acknowledged. First, egg consumption was self-reported by the participants and we cannot exclude reporting bias in the present study. The second limitation is that we did not calculate detailed nutrient and energy intakes from dietary habits.
Conclusion
Our data support a possible relationship of egg consumption and increased risk of type 2 diabetes mellitus.
Acknowledgements
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. No conflict of interest was declared. L.R. researched the data, contributed to the analysis, wrote the manuscript and edited the manuscript. R.O. researched the data, contributed to the analysis and edited the manuscript. The authors express sincere thanks to the staff of the Kaunas Dainavos Outpatient Clinic for providing affable conditions for the study.