Genome-wide association studies of age-related macular degeneration (AMD) have been very successful, as recently described. Already in year 2006, the known disease loci explaining some 50% of the empirical familial risk of AMD was identified, which was assumed to vary between 3 and 6 among siblings. This percentage is among the highest of all common diseases. As for most diseases, the concern has been the low proportion of empirical familial risk explained by the detected variants, referred to as missing heritability. The estimation of the proportion of familial risk explained by the identified genes depends critically on how accurately the empirical familial risk has been estimated in the original family studies. Three studies on AMD, reporting familial risks of 19.3, 4.2, and 2.4, have been published. All these studies have been small, and the highest risk was reported in a study in which only one AMD case was found in the control group.

We used the nation-wide Swedish data sets to assess the familial risk of AMD. The familial relationships (offspring, maximum age 75 years and no age limit for parents) were available from the Multigeneration Register, and AMD diagnosis was obtained through the Hospital Discharge Register, based on the International Classification of Diseases version 10 code H35.3 (1997–2007); the code does not give diagnostic details, but considering that the patients were hospitalized, a relatively severe disease can be assumed. Standardized incidence ratios and 95% confidence intervals (CIs) were calculated as a relative risk for offspring with AMD when a parent was diagnosed with the same disease, compared with offspring whose parents were unaffected. Similar calculations were carried out for siblings. The start of the follow-up was January 1, 1997, and it was continued until hospitalization for AMD, death, emigration, or closing date, December 31, 2007. The standardized incidence ratios were standardized for age (5-year groups), sex, period (5-year groups), region, and socioeconomic status.

The number of hospitalized persons for main diagnosis AMD in Sweden was 7290, of whom 4610 (63.2%) were female. The age group with the largest number of patients was 60- to 79-years. Familial standardized incidence ratios for offspring AMD were 2.24 (n = 29; 95% CI, 1.50 to 3.23) when any first-degree relative was diagnosed with AMD. The risk was somewhat higher (2.59) when parents were the probands compared with sibling probands (2.12). Diagnosis before age 65 years was associated with a risk of 2.55 (n = 19; 95% CI, 1.53 to 4.00) compared with 1.82 (n = 10; (%5 CI, 0.87 to 3.37) at an older age. Because the number of patients in the offspring generation was 3227, the proportion of familial cases was 0.9 % (29/3227).

The present study was nationwide, covering all patients hospitalized for AMD in Sweden. Yet the number of familial cases was small, as in the previous studies. The present familial risk of 2.24 was close to the United States estimate of 2.4, which may suggest that the newly discovered genetic loci cover well over 50% of this empirical risk. If correct, there would not be much room for yet-undiscovered genes, nevertheless keeping disease heterogeneity in mind.