**3. Results**

The prevalence of systemic comorbidities was 48.8% (219/449 cases). The most prevalent comorbidity was hypertension in 108 cases (24.1%), followed by insomnia in 67 cases (14.9%) and depression in 20 cases (4.5%). The 113 cases categorized as "others" included hyperlipidemia (15), Sjogren's syndrome (7), abnormal thyroid function (7), hypercholesteremia (7), angina (4), arrhythmia (4) and so on. Ninety-eight patients (21.8%) were taking anti-hypertensives and 55 (12.2%) were taking sedatives (Table 2).


**Table 2.** Prevalence of systemic comorbidities and oral medicine use.

DM, diabetes mellitus; RA, rheumatoid arthritis; SD, standard deviation. \*1: (+) indicates more than one comorbidity selected among hypertension, insomnia, depression, DM, RA or others. \*2: multiple answers allowed.

Patients with dry eye disease who had any systemic comorbidities exhibited a significantly lower Schirmer value (*p* = 0.004), shorter TBUT (*p* = 0.002), higher keratoconjunctival staining score (*p* = 0.0002) and lower utility value estimated by HUI-3 (*p* = 0.003) than those without systemic comorbidities (Table 3). After adjustment for age, differences in TBUT, keratoconjunctival staining score and HUI-3 utility value remained statistically significant (*p* = 0.011, *p* < 0.001 and *p* = 0.03, respectively). Patients who in addition to dry eye disease also had hypertension or depression exhibited a significantly lower Schirmer value than those without (*p* = 0.008 and *p* = 0.04, respectively), while those who had insomnia or depression exhibited a significantly shorter TBUT than those without (*p* = 0.01 and *p* = 0.04, respectively). The DEQS score was lower in patients with dry eye disease and hypertension (*p* = 0.003). After adjustment for age, the TBUT in patients with depression or insomnia remained significantly shorter (*p* = 0.04 and *p* = 0.04, respectively).

Regarding oral medicine use, patients with dry eye disease who were prescribed antidepressants had a significantly lower Schirmer value (*p* = 0.04) and lower TBUT (*p* = 0.04), while those who used anti-hypertensives had a significantly lower Schirmer value (*p* = 0.009) and lower DEQS score (*p* = 0.03) than those who did not receive such medications. After adjustment for age, antidepressant use remained significantly different for both the Schirmer value and TBUT (*p* = 0.05 and *p* = 0.04, respectively; Table 4).

We examined the relationship between dry eye disease subtypes and systemic comorbidities. We found that patients with hypertension and insomnia had a significantly higher prevalence of MGD (hypertension: odds ratio [OR] 2.65, 95% confidence interval [CI] 1.64–4.28, *p* < 0.01; insomnia: OR 1.94, 95% CI 1.10–3.42, *p* = 0.02). Patients with dry eye disease and hypertension, insomnia or depression had a significantly higher prevalence of FRD (hypertension: OR 1.56, 95% CI 1.00–2.44, *p* = 0.05; insomnia: OR 2.66, 95% CI 1.57–4.51, *p* < 0.01; depression: OR 4.92, 95% CI 1.85–13.08, *p* < 0.01). After the OR was adjusted for age, insomnia and depression, comorbidities with dry eye disease were significantly associated with FRD (insomnia: OR 2.16, 95% CI 1.26–3.72, *p* < 0.01; depression: OR 4.89, 95% CI 1.81–13.22, *p* < 0.01; Table 5).



TBUT, tear film break up time; DEQS, Dry Eye-Related Quality-of-Life Score; HUI, health utilities index; \* *p* ≤ 0.05; "+": presence; "−": absence.


**Table 4.** Oral medicine and dry eye parameters.

TBUT: tear film break up time; DEQS: Dry Eye-Related Quality-of-Life Score; HUI: Health Utilities Index. \* *p* < 0.05. Mean ± SD; "+": presence; "−": absence.


**Table 5.** Association between dry eye subtype and systemic comorbidities.

OR, odds ratio; CI, confidence Interval; ADDE, aqueous-deficient dry eye; DE, dry eye; sTBUT, short tear film break up time dry eye; MGD, meibomian gland dysfunction; FRD, friction-related disease.
