**3. Results**

#### *3.1. Clinical and Characteistics of Participants*

In this study, mean age of participants was 35.56 ± 12.12 years (range 18 to 59). 53.8% were women. 64.2% (95%CI: 61.2–67.1) of participants had dry eye defined as OSDI > 12.

*3.2. Distribution of Strip Meniscometry Readings*

> The distribution is bimodal and not normal (Figure 1). The mean readings were 7.7 ± 3.6.

**Figure 1.** Histogram of strip meniscometry values in factory workers.

#### *3.3. Factors A*ff*ecting Strip Meniscometry*

The factors associated with SM are shown in (Table 1). When performing a T-test to determine the association between MGD and SM, the three grades of MGD were categorized into two categories for the T-test. The smallest two grades of MGD (MGD Types 0 and 1) are combined into one category, whereas MGD Type 2 was used as the other category.


**Table 1.** Summary of strip meniscometry readings in this study.


**Table 1.** *Cont.*

Reduced SM readings were associated with increased OSDI (Figure 2A) (r = −0.72, *p* < 0.001), directly correlated with Schirmer (Figure 2B) (r = 0.71, *p* < 0.001), and associated with increased MG severity (Figure 2C) (*p* < 0.001).

**Figure 2.** Scatter diagrams showing the relationship between meniscometry and (**A**) Ocular Surface Disease Index (OSDI) scores, (**B**) Schirmer test results. (**C**) Bar graph showing the relationship between meniscometry and meibomian gland dysfunction. MG0: clear meibum, MG1: colored meibum with normal consistency, MG2: viscous meibum, MG3: inspissated meibum, and MG4: blocked meibomian gland. \*\*\*: two-tailed p-value for ANOVA and post-hoc tests, *p* < 0.001.

Interestingly, all the 23 subjects with SM < 3 mm had severe dry eye symptoms (OSDI of 33 or more) (*p* < 0.001 on Fischer's exact probability test).

#### *3.4. Multivariate Analysis*

We first performed logistic regression with SM categorized as < 7 mm to be low (abnormal), since the mean SM of this study was 7.7 mm. With age, gender, and the clinical parameters as independent variables, we found low values of SM to be significantly associated with higher OSDI and lower Schirmer's readings (Table 2).

**Table 2.** Multiple logistic regression where dependent variable is the abnormal meniscometry values ( < 7 mm)


† Adjusted by Age and Gender; †† Adjusted by Age, Gender and OSDI; ††† Adjusted by Age, Gender, ocular surface disease index (OSDI), tear breakup time (TBUT), Schirmers, and meibomian gland dysfunction (MGD); \* *p* < 0.05.

However, when we performed logistic regression with SM categorized as <3 mm to be low (abnormal), we found abnormal SM to be significantly associated with higher OSDI and lower TBUT readings after adjustment for the other variables (Model 4 in Table 3). However there were only 23 or 2.19% of participants with SM less than 3 mm, so it may or may not be possible to uncover all the associated factors with this sample size.

**Table 3.** Multiple logistic regression where dependent variable is the abnormal meniscometry values ( < 3 mm).


† Adjusted by Age and Gender; †† Adjusted by Age, Gender and OSDI; †††Adjusted by Age, Gender, ocular surface disease index (OSDI), tear breakup time (TBUT), Schirmers, and meibomian gland dysfunction (MGD); \* *p* < 0.05.
