Subjects were enrolled from a prospective epidemiologic study on the prevalence of ischemic heart disease (IHD) and effect of non-pharmalogical intervention on IHD risk 25. Details of subject recruitment and characteristics have been described thoroughly previously 25. In brief, participants in the Inter99 Study were drawn on a random basis from the Danish Civil Registration System which comprises all subjects residing permanently in Denmark. Participants were recruited in 7 birth cohorts (years of birth 1939/1940, 1944/1945, 1949/1950, 1954/1955, 1959/1960, 1964/1965, and 1969/1970).

The present study was part of The Inter99 Eye Study which consists of invited subjects from the Inter99 Study. The Inter99 Eye Study population was invited in two main groups: 1) an age- and sex-stratified control group polled to match the background population and 2) high risk subjects, i.e. subjects with high risk of ischemic heart disease or subjects with diabetes mellitus or impaired glucose tolerance. The present study includes only subjects from the control group (potentially eligible, n = 502). The study population thus represents the entire spectrum of IHD risk and glucose tolerance. Thirty-six subjects were excluded as they did not have lens fluorescence measured due to technical difficulties. Furthermore, 2 subjects were excluded due to previous corneal grafting (keratoplasty), and 10 eyes of 6 subjects were excluded due to cortical lens opacities exceeding 2.5 on the LOCS III grading scale as such opacities may interfere with the fluorescence measurements because they reach into the central 2 mm of the optical axis where fluorescence measurements take place 26. Only subjects who had an eye examination within 6 months of the general health assessment were included, this criteria lead to exclusion of 39 subjects. Thus, a total of 421 subjects were included in the present study.

The present study population matched the background population which was a random sample of the Danish suburban population as assessed on the basis of a number of important characteristics such as age and IHD risk (table 1).

The study protocol was reviewed and approved by the local medical ethics committee of Copenhagen County. Written informed consent was obtained from all participants after the nature and possible consequences of the study had been carefully explained.

All subjects underwent a general ophthalmic examination including determination of visual acuity and assessment of fundus status. Subjects were asked about present or past history of ophthalmic disorders, surgery, or ocular medication, family history of cataract and glaucoma.

Lens fluorescence was measured on the undilated eye using a non-invasive ocular fluorophotometer, the Fluorotron (OcuMetrics, Mountain View, CA, USA). Excitation was from 430–490 nm and detection from 530 to 630 nm. The Fluorotron was fitted with an anterior segment adaptor that concentrates the 149 measurement steps in intraocular intervals of 0.125 mm in the anterior segment of the eye. Lens fluorescence was measured along the optical axis of the eye and the measurements were corrected for loss of light due to absorption as described previously. 27 Six scans, three from each eye, were recorded and the mean of these 6 scans was used in the analyses. All fluorescence values were calibrated to an external solution of fluorescein and are reported in ng fluorescein equivalents per ml [ng f-eq/ml].

Presence and severity of lens opacities was determined from retroilluminated lens photographs focused on the anterior lens surface taken after dilation of the pupil using one drop of 0.5% tropicamide and one drop of 10% phenylephrine hydrochloride on 35 mm Kodak Ektachrome 64 colour slides (Kodak Inc, Rochester, NY, USA) using a Canon 60UVi camera (Canon Inc, Japan). The degree of cortical lens opacities was graded by one of the authors (LK) according to the LOCS III scale. 28 Cortical lens opacities exceeding 2.5 on the LOCS III scale lead to exclusion of that eye. Nuclear lens aging was assessed by lens autofluorometry using the Fluorotron. Lens fluorescence measurements have previously been shown to correlate well to the LOCS III nuclear grading 29.

Furthermore, all subjects underwent a thorough general health examination and questionnaire concerning past and present medical history, and use of medication. A physical examination was performed including determination of height, weight, waist circumference, and ECG. Venous blood samples were drawn for measurement of plasma cholesterol, HDL-cholesterol, triglycerides, and HbA1c. An oral glucose tolerance test was used and evaluated according to WHO 1999 criteria 30.

Ischemic heart disease (IHD) risk was calculated using the Precard ^® software. The Precard software calculates the risk (in percentage) of having an ischemic heart event based on age, sex, cholesterol (total cholesterol and HDL-cholesterol), systolic blood pressure, smoking, body mass index, diabetes mellitus, family history of cardiovascular disease and previous heart disease. 21 As the risk of ischemic heart disease increases dramatically with age, an age-standardized measure was produced by projecting the ischemic heart event to occur between the ages of 60 and 70 years.

Smoking was assessed as number of pack years ever smoked. One pack year equals 20 cigarettes per day for one year.

All statistical procedures were performed using the SAS software (version 8.2, SAS Institute, Cary, NC, USA). The level of statistical significance was set at 0.05. Lens fluorescence was transformed to logarithmic values to obtain a normal distribution with a constant variance, all fluorescence values reported have been backtransformed to the geometrical mean with 95% confidence intervals. Relations between variables were calculated using linear or multiple regression analysis.

A significant positive association was found between lens fluorescence (LF, log10 values) and ischemic heart disease risk (IHD risk, in %) (p = 0.001, adjusted R²= 51.4%), eq. 1:

LF = 0.0126·Age + 0.0024·IHD risk + 2.0214     (eq. 1)

Risk of IHD was calculated as a subject's risk of having an IHD event projected to occur between the age of 60 and 70 years. Figure 1 shows the relationship between lens fluorescence and IHD risk.

Risk of having an ischemic heart disease (IHD) event, projected to occur between the age of 60 and 70 years, was calculated using the Precard^® software. Since lens fluorescence was significantly increased in subjects with high IHD risk profile we wanted to examine whether this was due to one or more of the specific variable(s) that were associated to IHD risk.

Lens fluorescence was significantly related to age (p < 0.0001) and number of pack years ever smoked (p < 0.0001), and it was significantly increased in subjects with diabetes mellitus compared to normoglycemic subjects (p < 0.0001). No relation was found for lens fluorescence and sex (p = 0.37), or lens fluorescence and total cholesterol (p = 0.13), HDL-cholesterol (p = 0.44), systolic blood pressure (p = 0.99), or body mass index (p = 0.17). Lens fluorescence was not increased in subjects who had previously had a myocardial infarction (n = 11, p = 0.28) or in subjects with a family history of cardiovascular disease (n = 34, p = 0.99).

In order to quantify the effect of dysglycemia on lens fluorescence, three different glucose variables were tested: HbA1c, fasting p-glucose and 120-min post oral glucose tolerance test p-glucose. All three variables showed a significant positive association to lens fluorescence (p-values < 0.0001, Pearsons correlation coefficients ranged from 0.30 to 0.34) as well as to IHD risk assessed using the Precard ^® software (p-values < 0.0001, Pearsons correlation coefficients ranged from 0.40 to 0.49). Equation 2 comprises the single IHD risk factors that showed a significant relationship to lens fluorescence (p < 0.0001 for all three explanatory variables, adjusted R² = 54.0%). Fasting p-glucose was chosen to represent dysglycemia.

LF = 0.01191·Age + 0.01637·Fasting glucose + 0.00127·Pack years + 1.96338     (eq. 2)

Thus, association between lens fluorescence and the Precard ^® IHD risk estimate (eq. 1) was as strong as the relationship between lens fluorescence and the known single risk factors for IHD, hyperglycemia and tobacco smoking (eq. 2), adjusted R² 0.51 and 0.54, respectively.