The Salisbury Eye Evaluation was a population-based study of visual function in the elderly on the Eastern Shore of Maryland. The dry-eye component of this investigation studied 2420 men and women 65 years of age or older (5). Fifty-eight percent of the dry eye study cohort was female, and 26% was black. DES was assessed by a questionnaire that queried subjects about six symptoms including dryness, grittiness/sandiness, burning, redness, crusting on the eyelashes, and eyes being stuck shut in the morning. Each subject was asked to rate the frequency with which she or he experienced each symptom, with choices of all of the time, often, sometimes, or rarely. A subject was considered
to have DES if she or he experienced a minimum of one of the six symptoms at least often.

The Melbourne Visual Impairment Project dry-eye study included a population-based sample of 926 men and women at least 40 years of age (mean, 59.2 years) (7). Forty-seven percent of subjects were male. Subjects were interviewed to rate the severity of six dry-eye symptoms: discomfort, foreign-body sensation, itching, tearing, dryness, and photophobia. Definitions were provided for the possible ratings of absent/none, mild, moderate, or severe. Severe symptoms generally had to present constantly. A subject was considered to have DES if at least one of the six symptoms was considered by the subject to be “severe,” with the additional condition that the symptom was not attributable by the patient to hay fever.

DES was assessed during the second Beaver Dam Eye Study examination in 3722 men and women at least 48 years of age (mean, 65 years) (6). Ninety-nine percent of subjects were white, and 43% were male. DES was assessed by the question, “For the past 3 months or longer have you had dry eyes?” If needed, the subject was given a description of “foreign body sensation with itching, burning, sandy feeling, not related to allergy.” Subjects who answered this question affirmatively were considered to have DES.

The WHS, a randomized trial to assess the benefits and risks of aspirin and vitamin E in the prevention of cardiovascular disease and cancer in healthy women, comprises 39,876 female health professionals (13), aged 49 years and older at the time of dry-eye assessment on the 4-year follow-up questionnaire (8). The PHS was a randomized trial of aspirin and beta-carotene in the prevention of cardiovascular disease and cancer conducted among 22,071 U.S. male physicians (14), who were aged 55 years and older at the time of dry-eye ascertainment at 14-years of follow-up (8). DES was ascertained by three questions pertaining to diagnosis or symptoms of DES: (a) “Have you ever been diagnosed by a clinician as having dry-eye syndrome?” (b) “How often do your eyes feel dry (not wet enough)?” and (c) “How often do your eyes feel irritated?” Possible answers to the latter two questions included “constantly,” “often,” “sometimes,” or “never.” These two questions were found to have a sensitivity of approximately 60% coupled with a specificity of 94% compared with clinical diagnosis of dry eye (15), and nearly the same predictability as a longer 14-item questionnaire. DES was defined as the presence of clinically diagnosed DES or severe dry-eye symptoms of both dryness and irritation either constantly or often.

The Salisbury study was the first large-scale epidemiologic study to report an estimate of the prevalence of DES, finding that 14% of subjects reported one or more symptoms to be present at least often (5). In the Melbourne study, DES had been previously diagnosed in less than 1% of subjects. Overall, 5.5% of subjects reported at least one severe symptom not attributable to hay fever (7). Of the 3703 subjects in the second Beaver Dam Eye Study examination with data on dry eye, the prevalence of dry-eye symptoms was 14.4% [95% confidence interval (CI), 13.3%-15.6%] (6).

In the WHS, the age-adjusted prevalence of DES was 7.8%, or 3.23 million women aged 50 years or older in the United States (16). The prevalence of diagnosed DES was 6.1%, and that of symptoms of both dryness and irritation constantly or often was 3.4%. In preliminary findings from the PHS, the age-standardized prevalence was 3.5% among men aged 55 years and older. Based on this estimate, 900,000 men in this age group are predicted to have DES.

Clinicians have long suspected that DES becomes more common with aging, and there is some evidence of an age-related decrease in tear production (16,17). Until recently, however, epidemiologic data showing an age-related increase in DES were lacking. Figure 11-1 summarizes data on the relationship of the prevalence of DES according to age in the large epidemiologic studies described previously. Inspection of Fig. 11-1 suggests that the results of all studies taken together are most consistent with a trend toward a higher prevalence of DES in the older age groups. There is currently little information on the incidence of DES. Moss and colleagues have estimated the 5-year incidence of DES in the Beaver Dam Study as 13%, with a significantly higher incidence in older subjects (17.9% for those 80 years of age and older, vs. 10.7% for 48- to 59-year-olds) (18).

Basic research points to a possible role for sex steroid hormones in the pathogenesis of DES, including both androgens (as a protective factor) and estrogens (as a risk factor) (19, 20, 21, 22, 23, 24, 25, 26, 27). Apparently consistent with the basic research, clinical and epidemiologic observations have long suggested that DES is more common in women. In the Melbourne study, women were nearly two times more likely to report severe symptoms of dry-eye [odds ratio (OR), 1.85] (7). Similarly, in the Beaver Dam study, the prevalence of DES was significantly higher in women (17.0%) compared with men (11.1%) (6). Based on data from the WHS and PHS, the estimated age-standardized prevalence of DES in the United States appears to be higher in women than in men (Fig. 11-2). In contrast, investigators of the Salisbury study observed no significant sex-related difference in the prevalence of dry-eye symptoms among women (15.6%) and men (13.3%) aged 65 years and older (5).

Only two studies have been able to address the issue of potential racial/ethnic differences in DES. The Salisbury study included 650 blacks and 1832 whites from the same community in Maryland and observed a prevalence of DES symptoms of 13.5% among blacks and 15.0% among whites, with no statistically significant difference between these groups (5). In the WHS, 95% of the women indicated they were white, 2.2% black, 1.4% Asian/Pacific islander,
1% Hispanic, and 0.3% Native American; 0.2% were of other or unknown race/ethnicity (13). The prevalence of DES was similar among these groups of women, and there were no significant differences in the prevalence DES among blacks and whites. However, both Hispanic and Asian women reported a higher prevalence of severe dry-eye symptoms compared with whites (16). The reasons for this are unclear but may be related to several factors. For example, some groups of women with DES may be getting less relief from dry-eye symptoms. In addition, the higher prevalence of dry-eye symptoms in some minority groups could be related to existence of a greater number of health problems in these women, which either themselves or by virtue of their treatments may be associated with a higher prevalence of dry-eye symptoms. There is also the possibility of underdiagnosis of DES among some minority groups. Further studies are needed to clarify these issues.

Current information on other risk factors for DES is limited. Future studies should examine compelling hypotheses in more detail. In the Melbourne study (7), investigators found no significant associations with the use of contact lenses, cigarette smoking, history of previous eye injury, definite symptoms of dry mouth, menopause, or the use of hormone therapy among postmenopausal women. Arthritis was found to be significantly associated with a higher prevalence of clinical signs of dry eye. However, the usefulness of these analyses was limited because some of these exposures were rare in the population (e.g., only 21 subjects wore contact lenses) and no estimates of the magnitude or confidence intervals for these relationships were provided. In the Salisbury study, dry-eye symptoms were reported to be significantly associated with self-reported poor health and with the number of medical comorbidities (5). Further, a combined end point of either dry-eye or dry-mouth symptoms was also associated with use of certain classes of systemic medications, including nonsteroidal antiinflammatory drugs (OR, 1.30), diuretics (OR, 1.25), vasodilators (OR, 1.37), analgesics and antipyretics (OR, 1.28), antiulcer agents (OR, 1.44), anxiolytics and benzodiazepines (OR, 2.35), antiinfectives (OR, 1.88), antidepressants and antipsychotics (OR, 2.54), antihypertensive
agents (OR, 1.98) and antihistamines (OR, 1.67). The number of medication classes associated with drying symptoms used was also associated with a monotonic increase in risk of dry-eye or dry-mouth symptoms (OR of 2.5 for use of three; and OR of 4.2 for use of six or more vs. none). On the other hand, no relationship was found for rheumatoid arthritis, smoking, alcohol consumption, use of estrogen, presence of anti-Ro/SS-A or La/SS-B autoantibodies, antinuclear antibody positivity of at least 1:320, or rheumatoid factor positivity (28). Investigators in the Beaver Dam study (6) performed a non-hypothesis-driven type of analysis of a long list of factors to determine whether any would show a statistical relationship with dry-eye. In these analyses, statistically significant independent associations were observed for arthritis history (OR, 1.91); past (OR, 1.22) and current (OR, 1.82) cigarette smoking; caffeine use (OR, 0.75); thyroid disease (OR, 1.41); history of gout (OR, 1.42); total/high-density lipoprotein cholesterol ratio (OR, 0.93); diabetes mellitus (OR, 1.38); and past (OR, 1.35) and current (OR, 1.41) use of multivitamins. These findings must be interpreted cautiously because many statistical tests were performed, increasing the probability of spurious findings.

Given the basic research supporting the hypothesis that sex steroid hormones modulate the production and release of tear components, the potential relationship of hormone therapy with DES is of particular interest. In the WHS, compared with women who never used hormone therapy, the multivariate-adjusted ORs for DES were 1.69 (95% CI, 1.49-1.91) for women who used estrogen alone and 1.29 (95% CI, 1.13-1.48) for women who used estrogen plus progesterone or progestins (29). Thus, use of estrogen replacement therapy appears to increase the risk of DES by approximately 70% if used alone, and by 30% if used in combination with progesterone or progestins. In this study, there was also a significant association with the duration of hormone therapy such that for each 3-year increase in the duration of hormone therapy, the risk of DES increased by 15% (95% CI, 11%-19%). There is also emerging clinical evidence that androgen deficiency in humans (e.g., in patients taking antiandrogen medication) is associated with meibomian gland dysfunction and DES (25).

Recent evidence from the WHS suggests that dietary factors may also play a role in the development of DES (30). In this study, 32,470 female health professionals aged between 45 and 84 years who provided information on diet and DES were studied. Intake of omega-3 fatty acids was assessed by a validated food frequency questionnaire. After adjusting for age, other demographic factors, hormone therapy, and total fat intake, there was a significant reduction in risk of DES of approximately 20% for the highest versus the lowest fifth of dietary intake of omega-3 fatty acids. In addition, consumption of two to four 4-ounce servings of tuna fish per week was association with a 20% reduction in risk of DES, whereas consumption of five to six 4-ounce servings was related to a nearly 70% reduction in risk of DES. However, there may be other factors associated with individuals habitually consuming tuna on an almost daily basis.

Allergic conjunctivitis comprises one of the most common forms of conjunctivitis and is a significant public health problem in the United States. Most reports consistently indicate that the syndrome is usually part of a larger complex related to seasonal allergy, with as much as 95% of sufferers having associated allergic conditions, particularly allergic rhinitis (31,32). Because of this tendency toward association with other disorders, which also include asthma and sinusitis, the overall economic burden of allergic rhinoconjunctivitis is large. Indeed, overall direct medical expenditures attributable to allergic rhinoconjunctivitis have been estimated at $5.9 billion for 1996, with 38% of these expenditures accounted for by care of children 12 years of age and younger (33).

Although the condition can affect those of any age, the prevalence of allergic rhinoconjunctivitis has been most extensively studied in children, though even for children there is a paucity of recently published data from the United States. The International Study of Asthma and Allergies in Childhood (ISAAC) is a large-scale, multinational, collaborative study to evaluate the epidemiology of allergic diseases in children around the world. An ISAAC protocol study of 27,507 children throughout the United Kingdom reported a prevalence of rhinoconjunctivitis of 18.2% (34). In 37,592 children studied in 6 ISAAC centers in New Zealand, the prevalence of allergic rhinoconjunctivitis was 10% among 6- to 7-year-olds and 19% among 12- to 13-year-olds (35), whereas among a random sample of 3058 children in Nigeria, the prevalence of rhinoconjunctivitis was 39.2% (36). A nationwide survey of a sample of 38,955 Korean children revealed a prevalence of approximately 10% for rhinoconjunctivitis (37). A study from Sweden reported that 7% of school children and 15% of adolescents have seasonal allergic conjunctivitis (38). The prevalence of allergic rhinoconjunctivitis was 20.8% among an unselected group of 1175 Italian school children (39). Studies in Turkey have estimated a prevalence of allergic conjunctivitis of between 4.6% and 7.1% in primary school children (40,41). The most recent estimates
from Sweden indicate that approximately 17% of 12- and 13-year-olds suffer from allergic rhinoconjunctivitis (42). The suggestion has been made that the prevalence of allergic rhinoconjunctivitis has been rising in recent decades (43, 44, 45).

Family history appears to be a significant risk factor for allergic conjunctivitis. In one study of 445 patients with allergic conjunctivitis, 65% had a family history of allergy (31). Evidence for genetic linkage of allergic conjunctivitis has been identified in one study for chromosomes 5, 16, and 17, further supporting the idea that a genetic predisposition exists (46). Environmental factors are also likely to play a key role in the development of allergic conjunctivitis, and as with asthma there is some evidence suggesting that exposure to high antigen loads at a young age may offer a protective effect (42,47,48). Others have suggested that poorly ventilated homes, rather than outdoor pollution, may play a larger role in the incidence of allergic conjunctivitis (49). Early research suggests that diet may also play a role in allergic conjunctivitis, with one study demonstrating an increased risk with higher intake of n-6 polyunsaturated fatty acids (50). An ecological study of per capita dietary consumption and prevalence of allergic rhinoconjunctivitis in the ISAAC population identified a consistent inverse trend between the prevalence of allergic rhinoconjunctivitis and the intake of starch, cereals, and vegetables (51), and an increased risk associated with higher per capita consumption of trans fatty acids (52). However, these studies are far from conclusive, and further study of these issues is needed.

GPC was first described in 1974 (53). Since this early report, this entity has been shown to be closely associated with soft contact lenses, longer contact lens wearing times, and less frequent lens replacement. Nonetheless, there are no reliable estimates of the prevalence or incidence of GPC among the contact lens-wearing population. Although the syndrome is clearly associated with soft contact lenses, it also occurs in wearers of hard contact lenses (54). For example, in one study of 221 patients with GPC, 15.4% were wearers of hard contact lenses (55).

The possibility that the lens material itself is antigenic has been debated, but studies of potential lens antigenicity in animals dispute this theory (56). Nonetheless, different lens materials may be more prone to absorbing ocular proteins onto their surfaces, and this has been proposed as a potential cause of GPC-related allergy (57). Despite this, there is no convincing evidence to show that one type of soft lens is more prone to lead to the development of GPC than any other (58,59). The frequency of contact lens replacement appears to play a greater role than these factors in the pathogenesis of GPC, with one study showing that patients who replaced their lenses after 4 or more weeks had an incidence of GPC of 36%, versus 4.5% for those who replaced them after fewer than 4 weeks (60). Although there are few, if any additional data to support this theory, if it is true the recent shift toward more frequent replacement of contact lenses may reduce the impact of this condition.

The epidemiology of VKC has been little studied, though it appears clearly related to atopy (61,62). Case series reports demonstrate a higher frequency in children than adults, and the condition appears to show a male predilection (63,64). One recent publication of a series of 195 patients found the mean age to be 11 years and the male-to-female ratio to be 2.8:1.0, but it is not certain that this would reflect the overall demographic profile of VKC sufferers. Most patients in this study (77%) had seasonal disease (usually springtime), and 41% had a personal history of atopy. The overall prevalence of VKC is not known, although there are thought to be significant geographic variations in its frequency and severity, with it being potentially a greater problem in parts of the Middle East (65,66). Further study of the epidemiology of VKC, although perhaps somewhat difficult to conduct because of the (presumed) relatively low incidence of the condition, would be desirable.

Representing the most severe and clinically least encountered form of ocular allergy, AKC has been termed the adult variant of VKC and is often characterized by an association with dermatitis. The prevalence of AKC is unknown. It appears to affect predominantly men between 30 and 50 years of age and tends to be chronic (67). AKC is usually associated with systemic atopy, and a personal or family history of asthma or eczema is very common (68, 69, 70). One of the most damaging aspects of the disease is its tendency to cause corneal melting and severe vision loss. In one 9-year retrospective study of 45 patients with AKC, 34 had keratopathy and 21 had persistent epithelial defects causing severe visual compromise (69). Penetrating keratoplasty may be required for patients with permanent visual disability, but a very high rate of complications and graft rejection may occur (71,72).

Future studies of the incidence of these less frequent ocular allergic conditions might be most feasible in managed care or other settings with relatively large numbers of regularly followed patients.

A number of issues make study of the incidence and prevalence of keratoconus a difficult undertaking. The condition is relatively rare, making population-based prevalence surveys impractical. Furthermore, the gradual onset of the condition may delay clinical diagnosis, and relatively mild cases may go undetected. Given these limitations, the best population-based data on keratoconus incidence and prevalence in the United States have been published from records of patients in Olmsted County, Minnesota (76). Investigators reviewed medical information systems of the Mayo Clinic for all cases of newly diagnosed keratoconus identified from 1935 through 1982, supplemented with a questionnaire sent to local optometrists who were not covered in the database. Using U.S. census data to derive an appropriate denominator, the overall age-adjusted prevalence of keratoconus was estimated at 54.5 cases per 100,000 population. The age- and sex-adjusted incidence rate was estimated at 2.0 cases per 100,000 person-years of observation, with a rate of 2.4 per 100,000 among men and 1.6 per 100,000 among women. The higher incidence of keratoconus among men in this study was not statistically significant, although there was a statistically significantly higher prevalence among men in the study population (69.5 per 100,000 among men versus 39.2 per 100,000 among women). The incidence of keratoconus was observed to peak between the ages of 15 and 24 for men and 25 to 34 among women. Although the design of this study may have resulted in some underdetection of milder cases, it likely provides the best available estimate of the incidence and prevalence of clinically (and hence visually) important keratoconus. These data are further supported by a similar study in the United Kingdom, in which investigators conducted a retrospective review of ophthalmology records from a large hospital in Leicestershire with a catchment population of approximately 900,000 (77). In this study, the incidence of keratoconus among whites was found to be nearly the same (2.2 per 100,000 person-years) as that in Olmstead County, Minnesota, with a similar finding of a higher prevalence among men. An incidence of 1.4 per 100,000 person-years was observed in a study based at the Oulu University Central Hospital in Finland, where two thirds of cases occurred among men (78). Given these findings, previous reports of larger numbers of female patients among keratoconus series are perhaps more likely to be due to a greater use of health care services among women or some other factor. Studies of racial differences in keratoconus are lacking in the United States, but some data are available from the Leicestershire study in the United Kingdom. The overall incidence of patients with keratoconus presenting to hospital for contact lens fitting or corneal grafting was 11.8 per 100,000 person-years among Asians (primarily of Indian origin), an incidence more than five-fold higher than that among whites.

Certain types of mechanical trauma have long been suspected as risk factors for keratoconus. In particular, studies have focused on the use of hard contact lenses and eye rubbing. Although the latter factor’s association with the development of keratoconus is difficult to prove, some supportive evidence may be derived from animal models (79) and mechanistic research (80,81). In addition, a number of studies have reported a high prevalence of eye-rubbing behavior among individuals with keratoconus (82, 83, 84). Nonetheless, it has thus far not been possible to separate a possible effect of eye rubbing from other possibilities such as a genetic predisposition or other condition (e.g., atopy or use of hard contact lenses) that may be associated with both an increased tendency for eye rubbing as well as keratoconus. Unfortunately, although the theory is not entirely implausible, studies to date of the potential relationship of the use of hard contact lenses with the development of keratoconus suffer from possible biases introduced by flaws in study design that severely limit the usefulness of the findings.