The effect of environmental stress on the ocular surface in diabetes mellitus

Purpose: To investigate the effects of low relative humidity (RH) on the ocular surface in

individuals with diabetes mellitus.

Methods: A controlled environment chamber (CEC) was used to simulate desiccating

environmental stress in the form of low humidity. 66 participants aged 40 years and over

were recruited, 33 with diabetes and 33 controls. The two groups were examined during

two environmental conditions, 45% RH and 5% RH with constant temperature of 23

degrees Celsius. During 5% RH, participants watched a film for 90 minutes for visual

engagement. A series of dry eye examinations were performed at both 45% RH and 5%

RH. Dry eye symptoms were assessed using the validated Ocular Surface Disease Index

(OSDI) questionnaire and an additional likert questionnaire specifically designed for this

study quantifying symptom severity, the Dry Eye Severity Index (DESI). Participants in the

diabetes and control groups were categorised by dry eye disease (DED) based on final

OSDI score, where 13 points and above indicated a diagnosis of DED. Participants with

diabetes were also categorised based on disease duration, less than 10 years or 10 years

and more. Following dry eye testing at 45% RH and 5% RH, the effects of environmental

stress were analysed as well as possible differences between groups. Comparisons were

made between diabetes and control groups, diabetes duration subgroups as well as no

DED and DED subgroups.

Results: The diabetes group showed significant differences from the control group: greater

corneal fluorescein staining (CFS) at both 45% RH (p ≤ 0.001) and 5% RH (p = 0.002), as

well as less tear production with the Schirmer I test at both 45% RH (p = 0.041) and 5% RH

(p = 0.016). There was no significant difference in final OSDI score between the diabetes

and control groups, however the diabetes group showed significantly greater severity of

visual function symptoms compared to the control group. Following exposure to

environmental stress, both diabetes and control groups showed significant findings:

increased number of complete blinks, increased conjunctival bulbar hyperaemia, increased

CFS, reduced tear film break-up time (TBUT) and reduced central corneal thickness (CCT)

at five measurement locations. At 5% RH, the diabetes group showed significantly more

incomplete blinks than the control group (p = 0.017). Following environmental stress,

participants with diabetes less than 10 years showed a significant increase in the number

of complete blinks (p = 0.009), however this was not seen in participants with diabetes

duration of 10 years and more. Following environmental stress both duration subgroups

showed significantly increased conjunctival bulbar hyperaemia and CFS, with significantly

reduced TBUT. The diabetes no DED participants were the only subgroup demonstrating

significant exacerbation of DESI score (p = 0.028) after exposure. Following environmental

stress all DED and no DED subgroups showed a significant increase in the number of

complete blinks and CFS, with significantly reduced TBUT. Control no DED participants

were the only subgroup to show significantly increased basal tear volume following

environmental stress, measured with the Phenol Red Thread test (p = 0.025), however this

was not seen in the diabetes no DED or diabetes DED subgroups.

Conclusions: This is the first study to investigate the effects of a low humidity

environment with visual engagement in participants with diabetes. The findings show

significant impact of environmental stress on the ocular surface in both diabetes and

control groups with significant differences in the responses demonstrated by participants

with diabetes in comparison to controls.