While increasing estimates of prevalence have largely been attributed to advances in imaging and detection, partly driven by the boom in refractive surgery, we may be witnessing a true increase in the incidence of keratoconus for other reasons. Keratoconus is thought to be caused by a complex interplay of environmental and genetic factors, as well as biomechanical and biochemical disorders [14, 22–24]. Whilst their exact nature remains unclear, the relevance of environment and genetic factors explains the wide variation in prevalence across geographic areas. Varying prevalence among groups of different ethnicity living in the same geographic location suggests a genetic basis for disease. For example, higher prevalence than the British average has been found in Indian, Pakistani and Bangladeshi communities living in the United Kingdom [20, 21]. Further evidence of a genetic basis to the disease includes a significant association with consanguinity [25], autosomal dominant patterns of familial inheritance [26], higher concordance between monozygotic than dizygotic twins [27] and an association with other genetic disorders [28]. In one study, 10 % of patients with keratoconus had a family history of the disease, compared with just 0.05 % of the age-matched control group [29]. While both dominant and recessive patterns of autosomal inheritance have been proposed [9, 26], most cases of keratoconus to date have been deemed sporadic [28].

Current thinking is that geographic variations in prevalence can be explained by specific environmental factors promoting the expression of genetic factors related to ethnicity [15]. This may occur through epigenetic modifications including DNA methylation, which alters gene expression and subsequent phenotype [30]. Epigenetic modifications may result from environmental stressors including toxins and microbial exposure [30], but the most widely discussed are ultraviolet light exposure and eye rubbing [15].

Higher prevalence of keratoconus has been identified in Saudi Arabia [18], Iran [31], New Zealand [32], Israel [9] and some Pacific Island populations [5]. One explanation for this distribution is that these are areas with high ultraviolet (UV) light exposure—an environmental factor widely implicated in keratoconus [15]. Excess UV exposure may be geographical in origin (latitudinal or altitudinal), or related to outdoor pursuits including work and leisure activities. It is proposed that UV light increases the production of reactive oxygen species within the cornea [33] and that keratoconic corneas lack the ability to process excess reactive oxygen species [34] which leads to oxidative stress, cytotoxicity and corneal thinning [35].

While UV exposure and other environmental factors may play a role in determining the prevalence of keratoconus, the observation that Asians living in the United Kingdom have a prevalence of KC 7.6 times that of Caucasians suggests that non-environmental (i.e. genetic) factors are predominant [20]. Similarly, the prevalence of keratoconus is much higher in non-Persians (Arabs, Turks and Kurds) living in Tehran (7.9 %) than Persians (2.5 %) [31]. An argument against the role of excess UV exposure is that natural corneal collagen cross-linking is induced by UV light, which might be expected to reduce the prevalence and rate of progression of keratectasia in these areas [36].

Another environmental stressor widely studied in the context of keratoconus is eye rubbing and its relation to atopic or allergic disease. The association between eye rubbing and keratoconus was first described in 1956 [37]. While some studies have found similar rates of eye rubbing among patients with keratoconus and normal controls [10, 38], the association with eye rubbing is now widely accepted [8].

Recurrent epithelial trauma, similar to that caused by contact lens wear, results in the release of matrix metalloproteinases 1 and 13, interleukin-1 and tumour necrosis factor-alpha that lead to stromal remodelling and keratocyte apoptosis [39–41]. In turn, this may cause epigenetic modifications that facilitate the gene expression required for development of keratoconus [15]. Raised intraocular pressure caused by eye rubbing has also been cited as a contributory factor [42]. Interestingly, the duration of eye rubbing in patients with keratoconus appears longer than that associated with allergic eye disease not associated with keratoconus [43], possibly explaining why the majority of atopic patients do not go on to develop keratectasia. In hot and dry climates, high levels of dust may induce frequent eye rubbing, providing another potential explanation for the higher prevalence in these areas [8].

Some of the most compelling evidence to support a role for eye rubbing comes from reports of asymmetric keratoconus attributed to asymmetric eye rubbing [44, 45]. In 1984, Coyle described an 11-year-old boy who could stop his paroxysmal atrial tachycardia by rubbing his left eye, thus eliciting the oculo-cardiac reflex —a manoeuvre he performed up to 20 times a day. Although initially he had a normal ocular examination, when examined four years later he was found to have unilateral keratoconus [3]. It is possible that if this patient had been assessed with modern topography, he would have been found to have bilateral, albeit highly asymmetrical involvement.

The prevalence of atopic/allergic disease in developed countries has risen in recent years [46], and similar increases in keratoconus could be related to this [47]. Similar to keratoconus, the aetiology of atopy is thought to be a combination of genetic and environmental factors, linked via epigenetic modifications [15]. There is controversy as to whether there is a true association between atopy and keratoconus, and if there is, to what extent this might be. Whilst allergic eye disease causes itch that leads to the urge for patients to rub their eyes, atopy is common in the general population as well as the population of keratoconics. Some studies have recorded low correlations between atopy and keratoconus in large series [48–50] but others have reported strong associations [51–53].

More recently, using univariate analysis, Bawazeer et al found that keratoconus was associated with eye rubbing, atopy and family history [54]. However, multivariate analysis of the same data by the same group revealed eye rubbing as the only significant predictor of disease [54]. So, while atopy may contribute to keratoconus, it is thought to be more through the promotion of eye rubbing than the atopic process itself [54]. Evidence supporting this theory also comes from other conditions in which eye rubbing of a non-atopic origin is a feature. For example, both Leber’s congenital amaurosis and Down syndrome are associated with repeated eye rubbing and keratoconus [55]. However, whether keratoconus arises in these patients due to eye rubbing, or a genetic link, is yet to be elucidated.

Results concerning gender preponderance vary between studies. Jonas [6], Laqua [56], Amsler [57] and Hammerstein [58] demonstrated female preponderance of 53 %, 57 %, 65 % and 66 % female dominance, respectively. Ertan [59], Street [60], Fatima [61], Pouliquen [62] and Owens [32] demonstrated male preponderance of 62 %, 62 %, 53 %, 57 % and 59 %, respectively. Others have demonstrated no significant gender differences [7] and overall keratoconus is not considered to favour one gender over the other. In the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study [63], the progression rates of keratoconus were found to be equivalent in both men and women.

Keratoconus is a disease of adolescence and young adulthood typically presenting between the ages of 20 and 30 years [64], and diagnosis uncommon after the age of 35 years [4]. An exception to this is the diagnosis of older patients when presenting for other reasons, e.g. as candidates for cataract or keratorefractive surgery, where ectasia went undetected in earlier life due to either mild symptoms or less sophisticated imaging. Younger age of diagnosis may imply different aetiological factors. In a Japanese study, HLA antigen association was found to be higher in keratoconics diagnosed under the age of 20 years, in particular HLA-A26, B40 and DR9 antigens [65].

It should be noted that age of diagnosis is quite different from age of onset, and the latency between the two remains unclear. Younger age of onset predicts greater severity [66], faster progression and/or shorter time to penetrating keratoplasty [66]. Early diagnosis of keratoconus is crucial, as treatment including corneal collagen cross-linking can now be offered to arrest disease progression, and this has been facilitated by recent advances in imaging. In a Finnish cohort, Ihalainen (1986) reported that 73 % of patients were aged 24 years or below at the first onset of symptoms, with a mean age of 18 years [67]. Olivares Jimenez et al [68] reported a mean age of symptom onset of 15.39 years in a Spanish cohort. Again, ethnic differences are apparent, with Asians having a significantly lower age (4–5 years less) of first presentation compared with Caucasians [20, 21, 69]. As estimates of prevalence increase, estimates of age of onset decrease for the same reasons of earlier and more sensitive diagnosis [8].

Low numbers of patients reported as diagnosed with keratoconus aged over 50 years are somewhat surprising given the chronic nature of the disease [8] and several explanations have been proposed for this. Some have pointed to associations with conditions that reduce life expectancy, including mitral valve prolapse [70], obesity [71] obstructive sleep apnoea [71, 72] and Down syndrome, although the mortality rate in keratoconics has not been found to differ from that of the general population [73]. A genuine increase in prevalence due to rising rates of allergy in young people could be the explanation, although it could simply be that more young patients are being diagnosed due to technological advances [8].

Keratoconus has been associated with several other syndromic conditions. This has helped understand both the epidemiology and pathophysiology of the disease.