Introduction
Major depressive disorder is a highly prevalent disorder and is a leading cause of disability worldwide.Reference Liu, Fan and Zeng 1 –Reference McIntyre, Alsuwaidan and Baune 3 Selective serotonin reuptake inhibitors (SSRIs) and selective norepinephrine reuptake inhibitors (SNRIs) are among the most commonly prescribed antidepressant treatments for depressive disorders.Reference Costagliola, Parmeggiani and Semeraro 4 , Reference Sansone and Sansone 5 Other antidepressants that are used to treat mood disorders include, but are not limited to, norepinephrine-dopamine reuptake inhibitors (NDRIs), serotonin-norepinephrine-dopamine reuptake inhibitors (SNDRIs), serotonin modulators and stimulators (SMSs), serotonin antagonists and reuptake inhibitors (SARIs), norepinephrine reuptake inhibitors (NRIs), tricyclic antidepressants (TCAs), tetracyclic antidepressants (TeCAs), and monoamine oxidase inhibitors (MAOIs). 6
Extant literature indicates that depression may be associated with other chronic diseases, including, but not limited to, glaucoma and cataracts.Reference Pasquale 7 –Reference Park, Rosenblat and Brietzke 9 Glaucoma is a group of progressive optic neuropathies, of which symptoms of glaucoma may be attributable to the degeneration of retinal ganglion cells.Reference Weinreb, Aung and Medeiros 10 Different subtypes of glaucoma exist including angle closure glaucoma (ACG) and open angle glaucoma (OAG), and all are characterized by an increased intraocular pressure and vision loss.Reference Weinreb, Aung and Medeiros 10 , Reference Al-Timimi, Huang-Lung and Keay 11
Cataracts are characterized by opacification of the lens and are leading cause of blindness globally.Reference Kingman 12 The association between depression, glaucoma, and cataracts is hypothesized to be mediated by dysregulation of the autonomic nervous and inflammatory systems.Reference Pasquale 7 , Reference Jung, Han and Wang 8 , Reference Lindemann, Kuerten and Koch 13 Notwithstanding the aforementioned findings, the underlying mechanisms between depression, glaucoma, and cataracts remain unclear.Reference Jung, Han and Wang 8 A mechanism which may explain the association between serotonin reuptake inhibitors and an increase in intraocular pressure has been the effects of serotonin on ciliary body muscle activity and pupil dilatation.Reference Costagliola, Parmeggiani and Semeraro 4 , Reference Chen, Ng and Chiu 14 , Reference Costagliola, Parmeggiani and Sebastiani 15
In a separate line of research, results from population-based studies in the United States, Canada, and China indicate a significant positive association between the use of antidepressants, glaucoma, and cataracts.Reference Chen, Ng and Chiu 14 , Reference Erie, Brue and Chamberlain 16 , Reference Etminan, Mikelberg and Brophy 17 Furthermore, convergent evidence indicates that the prescription of antidepressants has been increasing in high-income countries during the past two decades.Reference Bogowicz, Curtis and Walker 18
The widespread prescription of antidepressants highlights the need for surveillance and an up-to-date understanding of the potential risk of glaucoma and cataracts associated with the use of antidepressants. Herein, we aim to evaluate whether there are disproportionate reporting of cataracts and glaucoma with treatment using antidepressant. Furthermore, we also aim to characterize the strength of the aforementioned associations using Bayesian probabilistic analysis.
Methods
Data collection
Data was retrieved from the United States Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS), a public database that contains a repository of post-approval spontaneous adverse event reports submitted to the FDA by healthcare providers, consumers, and the pharmaceutical industry. The investigated antidepressants include SSRIs (ie, citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline), SNRIs (ie, desvenlafaxine, duloxetine, levomilnacipran, milnacipran, venlafaxine, dextromethorphan-bupropion), SNDRIs (ie, nefazodone), SMSs (ie, vilazodone, vortioxetine), SARIs (ie, trazodone), NRIs (ie, reboxetine, viloxazine), NDRIs (ie, bupropion), TCAs (ie, amitriptyline, clomipramine, desipramine, dibenzepin, lofepramine, doxepin, imipramine, melitracen, nortriptyline, noxiptiline, protriptyline, trimipramine), TeCAs (ie, amoxapine, maprotiline, mianserin, mirtazapine, setiptiline), and MAOIs (ie, isocarboxazid, phenelzine, tranylcypromine, selegiline, moclobemide). In addition, agomelatine, brexanolone, esketamine, opipramol, and tianeptine were investigated and are herein grouped together as other antidepressants.
We obtained reports of cataracts and glaucoma from the date each antidepressant was approved by the FDA for the treatment of mood disorders to December 31, 2023. All antidepressants included were identified using generic and product names, and subsequently evaluated to determine their association with cataracts and glaucoma. The foregoing types of glaucoma were selected in alignment with the categories found in the FAERS database, which are consistent with the Medical Dictionary for Regulatory Activities terminology. Antidepressants that were not documented on the FAERS database were excluded from the investigation. Brand names of antidepressants were additionally searched when the metabolites of the antidepressant and/or product name(s) were not reported. For each antidepressant, we presented the number of cases that were recorded for cataracts and glaucoma (n) and total cases of eye disorders (N) (Table 1).
Table 1. Cases of Cataracts, Glaucoma, ACG, and OAG Associated With Antidepressants Identified in the FAERS Database, With Acetaminophen as the Reference Drug
Abbreviations: P, product name; G, generic name; ROR, reporting odds ratio; CI, confidence interval; ACG, angle closure glaucoma; OAG, open angle glaucoma; FAERS, Food and Drug Administration Adverse Event Reporting System; SSRI, selective serotonin reuptake inhibitor; SNRI, selective norepinephrine reuptake inhibitor; SNDRI, serotonin-norepinephrine-dopamine reuptake inhibitor; SMS, serotonin modulators and stimulator; SARI, serotonin antagonists and reuptake inhibitors; NRI, norepinephrine reuptake inhibitor; NDRI, norepinephrine-dopamine reuptake inhibitor; TCA, tricyclic antidepressant; TeCA, tetracyclic antidepressant; MAOI, monoamine oxidase inhibitor.
Antidepressant search terms: Agomelatine = agomelatine (G), agomelatine citric acid (G), Valdoxan (P); Amitriptyline = amitriptyline (G), amitriptyline hydrochloride (G), amitriptyline pamoate (G), elavil (P); Amoxapine = amoxapine (G), Asendin (P); Brexanolone = brexanolone (G), Zulresso (P); Bupropion = Bupropion (G), bupropion hydrobromide (G), bupropion hydrochloride (G), bupropion hydrochloride extended release (P), bupropion hydrochloride SR (P), bupropion hydrochloride XL (P), bupropion XL (P), hydroxybupropion (+/−)− (G), erythrohydrobupropion (G), threohydrobupropion (G); Bupropion\Dextromethorphan = bupropion\dextromethorphan (G), bupropion hydrochloride\dextromethorphan hydrobromide (G); Citalopram = citalopram (P), citalopram hydropromide (G), citalopram hydrochloride (G), desmethylcitalopram (G), didesmethylcitalopram (G), seropram (citalogram hydrobromide) (P), seropram (citalopgram hydrochloride) (P); Clomipramine = clomipramine (G), clomipramine hydrochloride (G), desmethylchlomipramine (G), anafranil (P); Desipramine = desipramine (G), desipramine hydrochloride (G), norpramin (P); Desvenlafaxine: desvenlafaxine (G), desvenlafaxine er (P), desvenlafaxine fumarate (G), desvenlafaxine succinate (G); Dibenzepin = dibenzepin hydrochloride (G); Doxepin = doxepin (G), doxepin hydrochloride (G), doxepin HCl, desmethyldoxepin (G), Sinequan (P); Duloxetine = duloxetine (G), duloxetin hydrochloride (G); Escitalopram = escitalopram (P), escitalopram oxalate (G), seropram (escitalopram oxalate) (P); Esketamine = esketamine (G), esketamine hydrochloride (G), Spravato (P); Fluoxetine = fluoxetine (p), fluoxetine hydrochloride (G), norfluoxetine (G); Fluovoxamine = fluvoxamine (G), fluvoxamine maleate (G), fluvoxamine maleate extended-release (P); Imipramine = imipramine (G), imipramine hydrochloride (G), imipramine pamoate (G), imipramine\imipramine hydrochloride (G); Isocarboxazid = isocarboxazid (G), Marplan (P); Levomilnacipran = levomilnacipran (G), levomilnacipran hydrochloride (G), levomilnacipran. (P); Lofepramine = lofepramine hydrochloride (G); Maprotiline = maprotiline (G), maprotiline hydrochloride (G), Ludiomil (P); Melitracen = melitracen (G), melitracen hydrochloride (G); Mianserin = mianserin (G), mianserin hydrochloride (G), mianserin\mianserin hydrochloride (G); Milnacipran = milnacipran\milnacipran hydrochloride (G), milnacipran hydrochloride (G); Mirtazapine = mirtazapine (G), apo mirtazapine (G), normirtazapine (G), Remeron (P), Remeron soltab (P), Remeronsoltab (P); Nefazodone = nefazodone (G), nefazodone hydrochloride (G), serzone (P); Nortriptyline = nortriptyline (G), nortriptyline hydrochloride (G), Pamelor (P); Opipramol = opipramol (G), opipramol hydrochloride (G), opipramol\opipramol hydrochloride (G); Paroxetine: paroxetine (G), paroxetine hydrochloride (G), paroxetine hydrochloride (P), paroxetine hydrochloride hemihydrate (G), paroxetine hydrochloride\paroxetine hydrochloride anhydrous (G), paroxetine mesylate (G); Phenelzine = phenelzine (G), phenelzine sulfate (G), Nardil (P); Protriptyline = protriptyline (G), protriptyline hydrochloride (G), Vivactil (P); Reboxetine = reboxetine (G), reboxetine mesylate (G), Edronax (P); Selegiline = selegiline (G), selegiline hydrochloride (G), Eldepryl (P), Emsam (P), Zelapar (P); Sertraline = sertraline (P), sertraline hydrochloride (G), desmethylsertraline (G); Tianeptine = tianeptine (G), tianeptine sodium (G); Tranylcypromine = tranylcypromine (G), tranylcypromine sulfate (G), tranylcypromine\tranylcypromine sulfate (G), Parnate (P); Trazodone = trazodone (P), trazodone HCl (P), trazodone hydrochloride (G); Trimipramine = trimipramine (G), trimipramine maleate (G), trimipramine mesylate (G), n-desmethyltrimipramine (G), Surmontil (P); Venlafaxine = venlafaxine (P), venlafaxine HCl (P), venlafaxine hydrochloride (G), venlafaxine hydrochloride ER (P); Vilazodone = vilazodone (G), vilazodone hydrochloride (G), VIIBRYD (P); Viloxazine = viloxazine (G), viloxazine hydrochloride (G), Quelbree (P); Vortioxetine = vortioxetine (G), vortioxetine hydrobromide (G), vortioxetine DL-lactate (G), brintellix (P), trintellix (P).
Statistical analysis
Data was analyzed using Microsoft Excel for Mac 2024 version 16.81 and RStudio 2023.09.1+494 “Desert Sunflower” Release for Windows. We sought to determine the extent of disproportionate reporting between cases and non-cases for cataracts and glaucoma associated with antidepressant use. Acetaminophen was the reference agent, as there is no association between ophthalmological diseases and acetaminophen exposure.Reference Mohamed and Meyer 19
Disproportionate reportings were determined using the reporting odds ratio (ROR) in accordance with the conventional formula: (odds of the event in the exposed group/odds of the event in the non-exposed group).Reference McIntyre, Alsuwaidan and Baune 3 , Reference Tenny and Hoffman 20 , Reference Montastruc, Sommet and Bagheri 21 When the odds of the adverse event in the exposed or unexposed group was zero, 0.5 was added to both the total number of cases (n) and total cases of eye disorders (N) associated with the antidepressant and reference agent, acetaminophen, to obtain an ROR.Reference Pagano and Gauvreau 22 , Reference Deeks and Higgins 23
In addition, the 95% confidence interval (CI) was also ascertained for each ROR and statistical significance was determined when the 95% CIs did not encompass 1.0.Reference Woods 24 Furthermore, using the lower limit of the 95% CI, the Bayesian information components (IC025) were also calculated to statistically determine the strength of the association between the antidepressant and the adverse event investigated and whether the combination of the drug and adverse event is more commonly reported in the FAERS database than anticipated. Significance regarding the association between the antidepressants and cataract as well as glaucoma was determined when IC025 > 0.
Results
Evaluation of SSRI-associated cataracts and glaucoma
Compared to acetaminophen, citalopram (0.50 ROR; 95% CI [0.36, 0.70]; p = 0.00010), escitalopram (0.53 ROR; 95% CI [0.37, 0.76]; p = 0.00050), paroxetine (0.33 ROR; 95% CI [0.23, 0.46]; p < 0.0001), and sertraline (0.56 ROR; 95% CI [0.44, 0.73]; p < 0.0001) were significantly associated with a lower probability of reported cataracts (Table 1, Figure 1). Both fluoxetine (0.83 ROR; 95% CI [0.64, 1.071]; p = 0.15) and fluvoxamine (0.59 ROR; 95% CI [0.29, 1.22]; p = 0.15) were also associated with a lower odds of reported cataract cases; albeit, not statistically significant (Table 1, Figure 1). Notwithstanding the aforementioned findings, we did not find a significant association between the SSRIs investigated and cataract (IC025 < 0) (Table 1).
Figure 1. Reporting odds ratios of cataracts associated with antidepressant treatment.
In addition to cataract, the association between SSRIs and glaucoma was also examined. Escitalopram (1.68 ROR; 95% CI [1.15, 2.46]; p = 0.0074), fluoxetine (1.81 ROR; 95% CI [1.30, 2.51]; p = 0.00040), and sertraline (1.72 ROR; 95% CI [1.27, 2.34]; p = 0.00050) were significantly associated with greater risk of reported glaucoma in comparison to acetaminophen (Table 1, Figure 2). In contrast, the remaining SSRIs were nonsignificantly associated with greater odds of glaucoma (p > 0.05) (Table 1, Figure 2). Furthermore, only fluoxetine (IC025 = 0.048) and sertraline (IC025 = 0.0028) were significantly associated with glaucoma (Table 1).
Figure 2. Reporting odds ratio of glaucoma associated with antidepressants.
Consistent with the aforementioned findings, escitalopram (25.61 ROR; 95% CI [10.27, 63.88]; p < 0.0001), fluoxetine (7.48 ROR; 95% CI [2.91, 19.27]; p < 0.0001), and sertraline (3.56 ROR; 95% CI [1.35, 9.40]; p = 0.011) were significantly associated with greater odds of ACG compared to acetaminophen (Table 1, Figure 3). In addition, citalopram (32.73 ROR; 95% CI [13.29, 80.61]; p < 0.0001), fluvoxamine (27.80 ROR; 95% CI [9.74, 79.36]; p < 0.0001), and paroxetine (3.56 ROR; 95% CI [1.35, 9.40]; p = 0.0001) were also significantly associated with increased risk of ACG in comparison to acetaminophen (Table 1, Figure 3). All SSRIs were significantly associated with ACG with the exception of sertraline (IC025 = −0.16) (Table 1).
Figure 3. Reporting odds ratios of angle closure glaucoma associated with antidepressants treatment.
In contrast to the results observed for ACG, only citalopram (18.65 ROR; 95% CI [1.031, 337.45]; p = 0.047) was significantly associated with a greater risk of reported OAG compared to acetaminophen (Table 1, Figure 4). The ROR for OAG being reported for escitalopram (14.51 ROR; 95% CI [0.75, 281.070]), fluoxetine (5.99 ROR; 95% CI [0.29, 124.80]), fluvoxamine (11.19 ROR; 95% CI [0.22, 564.78]), paroxetine (5.44 ROR; 95% CI [0.26, 113.27]), and sertraline (4.031 ROR; 95% CI [0.19, 84.00040]), although increased, did not reach statistical significance when compared to acetaminophen (Table 1, Figure 4). All SSRIs investigated were not significantly associated with OAG (IC025 < 0) (Table 1).
Figure 4. Reporting odds ratios of open angle glaucoma associated with antidepressant treatment.
Evaluation of SNRI-associated cataracts and glaucoma
Compared to acetaminophen, duloxetine (0.75 ROR; 95% CI [0.59, 0.96]; p = 0.021) and venlafaxine (0.54 ROR; 95% CI [0.44, 0.73]; p < 0.0001) have significantly lower odds of reported cataract cases (Table 1, Figure 1). Desvenlafaxine (0.73 ROR; 95% CI [0.48, 1.11]), levomilnacipran (0.37 ROR; 95% CI [0.092, 1.53]), and bupropion-dextromethorphan (0.35 ROR; 95% CI [0.021, 5.67]) were also associated with decreased risk of cataract; however, this association was nonsignificant (all p > 0.05) (Table 1, Figure 1). In contrast, albeit a nonsignificant association, milnacipran (1.19 ROR; 95% CI [0.60, 2.35]; p = 0.63) was associated with a greater risk of reported cataract (Table 1, Figure 1). Notwithstanding the aforementioned findings, all SNRIs investigated were not significantly associated with the development of cataract (IC025 < 0) (Table 1).
In contrast to the trends observed with cataract, desvenlafaxine (2.74 ROR; 95% CI [1.83, 4.10]; p < 0.0001), duloxetine (1.85 ROR; 95% CI [1.35, 2.52]; p = 0.0001), and venlafaxine (1.53 ROR; 95% CI [1.11, 2.10]; p = 0.010) were significantly associated with increased risk of glaucoma compared to acetaminophen (Table 1, Figure 2). In addition, only desvenlafaxine and duloxetine were significantly associated with glaucoma in comparison to acetaminophen (IC025 = 0.48 and 0.050, respectively) (Table 1). Milnacipran (1.29 ROR; 95% CI [0.46, 3.57]; p = 0.63) was also associated with greater odds of glaucomain compared to acetaminophen; however, this relationship was nonsignificant (Table 1, Figure 2). Contrastingly, compared to acetaminophen, both levomilnacipran (0.23 ROR; 95% CI [0.014, 3.74]) and bupropion-dextromethorphan (0.87 ROR; 95% CI [0.051, 14.25]) were associated with fewer odds of glaucoma; albeit, not statistically significant (p > 0.05) (Table 1).
All SNRIs investigated were associated with a greater risk of ACG compared to acetaminophen; however, this relationship was only significant for desvenlafaxine (9.030 ROR; 95% CI [3.13, 26.051]; p < 0.0001), duloxetine (8.71 ROR, 95% CI [3.46, 21.93]; p < 0.0001), milnacipran (20.0051 ROR; 95% CI [5.74, 69.67]; p < 0.0001), and venlafaxine (6.69 ROR; 95% CI [2.62, 17.07]; p = 0.0001) (Table 1, Figure 3). Consistently, desvenlafaxine, duloxetine, milnacipran, and venlafaxine were also significantly associated with ACG compared to acetaminophen (IC025 = 0.78, 0.31, 1.80, and 0.20, respectively) (Table 1).
Compared to acetaminophen, only duloxetine (28.56 ROR; 95% CI [1.71, 477.48]; p = 0.020), venlafaxine (19.43 ROR; 95% CI [1.14, 331.66]; p = 0.040), and bupropion-dextromethorphan (105.77 ROR; 95% CI [2.071, 5401.69]; p = 0.020) were significantly associated with greater odds of OAG; however, only duloxetine (IC025 = 0.021) and bupropion-dextromethorphan (IC025 = 0.42) were significantly associated with OAG (Table 1, Figure 4). The remaining SNRIs were not significantly associated with OAP (all p > 0.05 and IC025 < 0) (Table 1, Figure 4).
Evaluation of SNDRI-associated cataracts and glaucoma
Compared to acetaminophen, nefazodone (0.38 ROR; 95% CI [0.20, 0.70]; p = 0.0021) was only significantly associated with a lower risk of cataract and not glaucoma, ACG and OAG (Table 1, Figures 1–4). Notwithstanding the aforementioned findings, nefazodone was not significantly associated with any of the parameters of cataract and glaucoma investigated when compared to acetaminophen (Table 1).
Evaluation of SMS-associated cataracts and glaucoma
Both vilazodone (0.21 ROR; 95% CI [0.052, 0.85]; p = 0.029) and vortioxetine (0.26 ROR; 95% CI [0.12, 0.56]; p = 0.0006) were significantly associated with a lower odds ratio of reported cataract cases; however, both were nonsignificantly associated with cataract when compared to acetaminophen (both IC025 < 0) (Table 1, Figure 1). In addition, both vilazodone (48.27 ROR; 95% CI [1.96, 1188.14]; p = 0.018) and vortioxetine (28.76 ROR; 95% CI [1.38, 599.81]; p = 0.030) were also significantly associated with a greater risk of OAG in comparison to acetaminophen (Table 1, Figure 4). Consistently, the strength of vilazodone and vortioxetine’s association with OAG was also statistically significant (IC025 = 0.63 and 0.20, respectively) (Table 1). Only vortioxetine was significantly associated with a greater odds ratio of reported glaucoma (2.11 ROR; 95% CI [1.28, 3.45]; p = 0.0032) and ACG (12.82 ROR; 95% CI [4.44, 37.00]; p < 0.0001) cases, in which the strength of both these associations was significant (IC025 = 0.12 and 1.19, respectively) (Table 1, Figures 2 and 3). In contrast, vilazodone was nonsignificantly associated with a lower risk of glaucoma (0.26 ROR; 95% CI [0.036, 1.89]; p = 0.18) and greater odds ratio of ACG (3.22 ROR; 95% CI [0.38, 27.67]; p = 0.29) (Table 1, Figures 2 and 3). Furthermore, the strength of the association between vilazodone, glaucoma, and ACG was not significant (IC025 = −5.65 and −2.28, respectively) (Table 1).
Evaluation of SARI-associated cataracts and glaucoma
Trazodone (0.58 ROR; 95% CI [0.37, 0.91]; p = 0.017; IC025 = −1.33) was significantly associated with a lower odds ratio of cataract compared to acetaminophen; however, the strength of this association was not significant (Table 1, Figure 1). In contrast, compared to acetaminophen, trazodone (1.91 ROR; 95% CI [1.23, 2.97]; p = 0.0042) is associated with significantly greater odds of glaucoma and the IC025 is statistically significant (IC025 = 0.061) (Table 1, Figure 2). Furthermore, trazodone is also associated with a greater risk of ACG (1.53 ROR; 95% CI [0.30, 7.91]; p = 0.61) and OAG (3.83 ROR; 95% CI [0.076, 192.99]; p = 0.50); albeit, this is nonsignificant and the strength of the relationship is not statistically significant (all IC025 < 0) (Table 1, Figures 3 and 4).
Evaluation of NRI-associated cataracts and glaucoma
In comparison to acetaminophen, reboxetine is nonsignificantly associated with lower odds of both cataracts (0.36 ROR; 95% CI [0.022, 5.83]; p = 0.46) and glaucoma (0.89 ROR; 95% CI [0.053, 14.66]; p = 0.92), whereas viloxazine was nonsignificantly associated with greater odds of cataracts (1.021 ROR; 95% CI [0.060, 17.33]; p = 0.99) and glaucoma (2.67 ROR; 95% CI [0.16, 45.67]; p = 0.52) (Table 1, Figures 1 and 2). Both reboxetine (9.87 ROR; 95% CI [0.54, 181.74]; p = 0.12) and viloxazine (28.81 ROR; 95% CI [1.51, 549.29]; p = 0.025) are also associated with a greater odds ratio of ACG; however, this is only significant for viloxazine (Table 1, Figure 3). Notwithstanding the aforementioned findings, the strength of the association between each drug with cataract, glaucoma and ACG is not statistically significant (all IC025 < 0) (Table 1). Consistently, both reboxetine (108.67 ROR; 95% CI [2.13, 5551.66]; p = 0.020) and viloxazine (317.32 ROR; 95% CI [6.054, 16631.25]; p = 0.0044) are also associated with higher odds of OAG and the strength of this association is statistically significant (IC025 = 0.46 and 1.96, respectively) (Table 1, Figure 4).
Evaluation of NDRI-associated cataracts and glaucoma
Compared to acetaminophen, bupropion was significantly and nonsignificantly associated with less odds of reported cases of cataract (0.31 ROR; 95% CI [0.22, 0.46]; p < 0.0001) and glaucoma (0.88 ROR; 95% CI [0.58, 1.32]; p = 0.53), respectively (Table 1, Figures 1 and 2). Contrastingly, the risk of ACG (5.40 ROR; 95% CI [2.012, 14.47]; p = 0.00080) and OAG (7.064 ROR; 95% CI [0.34, 147.21]; p = 0.21) were both greater with bupropion use compared to acetaminophen; however, this was only significant for ACG (Table 1, Figures 3 and 4). Furthermore, only the association between bupropion and ACG was statistically significant (IC025 = 0.16) (Table 1).
Evaluation of TCA-associated cataracts and glaucoma
Compared to acetaminophen, desipramine (2.31 ROR; 95% CI [1.14, 4.65]; p = 0.020), doxepin (2.86 ROR; 95% CI [1.86, 4.39]; p < 0.0001), imipramine (1.67 ROR; 95% CI [1.019, 2.73]; p = 0.042), and protriptyline (4.66 ROR; 95% CI [1.023, 21.20]; p = 0.050) were significantly associated with a higher risk of reported cataract cases (Table 1, Figure 1). Amitriptyline (1.19 ROR; 95% CI [0.89, 1.60]), dibenzipin (12.81 ROR; 95% CI [0.43, 383.33]), lofepramine (6.40 ROR; 95% CI [0.29, 142.64]); melitracen (12.81 ROR; 95% CI [0.43, 383.33]), and noxiptiline (12.81 ROR; 95% CI [0.43, 383.33]) were also associated with greater odds of cataract; albeit, the associations were nonsignificant (all p > 0.05, IC025 < 0) (Table 1, Figure 1). Only the strength of the association between doxepin and reported cataract cases was significant (IC025 = 0.63) (Table 1). In contrast, clomipramine (0.14 ROR; 95% CI [0.034, 0.56]; p = 0.0057), nortriptyline (0.60 ROR; 95% CI [0.29, 1.23]; p = 0.16), and trimipramine (0.34 ROR; 95% CI [0.047, 2.44]; p = 0.28) were associated with lower odds of reported cataract cases; however, this was only significant for clomipramine (Table 1, Figure 1).
Consistent with the aforementioned trends, amitriptyline (1.96 ROR; 95% CI [1.32, 2.88]; p = 0.00070; IC025 = 0.12), doxepin (2.99 ROR; 95% CI [1.59, 5.62]; p = 0.00070; IC025 = 0.39), and imipramine (3.018 ROR; 95% CI [1.67, 5.46]; p = 0.00030; IC025 = 0.46) were significantly associated with a greater odds ratio of reported cases of glaucoma and the strength of these associations were statistically significant when compared to acetaminophen (Table 1, Figure 2). The remaining TCAs were all nonsignificantly associated with a greater risk of reported glaucoma except for timipramine (0.41 ROR; 95% CI [0.025, 6.69]; p = 0.53), which was nonsignificantly associated with lower odds of glaucoma in comparison to acetaminophen (Table 1, Figure 2).
With respect to ACG, all TCAs (p < 0.05) were significantly associated with greater risk of reported ACG cases compared to acetaminophen except for doxepin (2.96 ROR; 95% CI [0.34, 25.39]; p = 0.32) and notriptyline (2.26 ROR; 95% CI [0.26, 19.43]; p = 0.46) (Table 1, Figure 3). Notwithstanding the aforementioned findings, only amitriptyline (IC025 = 0.17), clomipramine (IC025 = 0.99), dibenzepin (IC025 = 2.07), imipramine (IC025 = 2.30), lofepramine (IC025 = 1.48), melitracen (IC025 = 2.07), and noxiptiline (IC025 = 2.07) were significantly associated with ACG (Table 1).
Similar to the trends observed for ACG, all TCAs (p < 0.05) were significantly associated with a greater odds ratio of reported OAG compared to acetaminophen except for imipramine (12.73 ROR; 95% CI [0.25, 642.87]; p = 0.20) and notriptyline (11.28 ROR; 95% CI [0.22, 569.61]; p = 0.23) (Table 1, Figure 4). In addition, the strength of the association between all TCAs and OAG was significant apart from desipramine (IC025 = −1.09), imipramine (IC025 = −2.52), notriptyline (IC025 = −2.68), and trimipramine (IC025 = −0.60) (Table 1).
Evaluation of TeCA-associated cataracts and glaucoma
Mirtazapine (0.57 ROR; 95% CI [0.39, 0.82]; p = 0.0023) was the only TeCA to be significantly associated with a lower risk of reported cataract cases in comparison to acetaminophen (Table 1, Figure 1). All other TeCAs (p > 0.05) were nonsignificantly associated with fewer odds of cataract compared to acetaminophen except for setiptiline (25.62 ROR; 95% CI [0.51, 1295.55]; p = 0.11) which was nonsignificantly associated with greater risk of cataract (Table 1, Figure 1). For all TeCAs, the IC025 indicated that the strength of the association with cataract was nonsignificant (IC025 < 0) (Table 1).
Compared to acetaminophen, mianserin (3.34 ROR; 95% CI [1.42, 7.88]; p = 0.0059) and setiptiline (63.50 ROR; 95% CI [1.25, 3226.61]; p = 0.038) were significantly associated with a higher odds ratio of glaucoma (Table 1, Figure 2). Notwithstanding the aforementioned finding, it is important to note that there were zero cases of glaucoma and eye disorders reported for setiptiline. Only the strength of the association between mianserin and glaucoma was significant (IC025 = 0.18) (Table 1). Amoxapine (0.77 ROR; 95% CI [0.047, 12.58]), maprotiline (0.53 ROR; 95% CI [0.073, 3.88]), and mirtazapine (0.80 ROR; 95% CI [0.48, 1.31]) were nonsignificantly associated with a lower risk of glaucoma in comparison to acetaminophen (Table 1, Figure 2).
With regard to ACG, only amoxapine (8.68 ROR; 95% CI [0.47, 159.50]; p = 0.15) was non-significantly associated with a higher risk of ACG compared to acetaminophen (Table 1, Figure 3). The remaining TeCAs were not only significantly associated with a greater odds ratio of ACG (p < 0.05), compared to acetaminophen, but the strength of the association was also statistically significant (IC025 > 0) (Table 1, Figure 3).
In contrast to the trends observed for ACG, only amoxapine (99.58 ROR; 95% CI [1.87, 4874.92]; p = 0.023) and setiptiline (7933.00 ROR; 95% CI [65.21, 965003.80]; p = 0.002) were significantly associated with greater odds of OAG (Table 1, Figure 4). It is important to note that there were zero cases of reported OAG and eye disorders associated with setiptiline. Notwithstanding the aforementioned findings, only amoxapine (IC025 = 0.28) and setiptiline (IC025 = 5.65) were significantly associated with OAG (Table 1). The remaining TeCAs investigated were all nonsignificantly associated with a lower odds ratio of OAG compared to acetaminophen, and the strength of the relationships was not statistically significant (IC025 < 0) (Table 1, Figure 4).
Evaluation of MAOI-associated cataracts and glaucoma
All MAOIs investigated were associated with a lower risk of cataract in comparison to acetaminophen; albeit, this relationship was not statistically significant (p > 0.05) (Table 1, Figure 1). In accordance with the aforementioned trend, no MAOIs investigated were associated with cataract (IC025 < 0) (Table 1).
With respect to reported cases of glaucoma, phenelzine (3.37 ROR; 95% CI [1.65, 6.90]; p = 0.00090; IC025 = 0.43) was the only MAOIs to not only be significantly associated with a greater risk of glaucoma in comparison to acetaminophen, but the strength of the association was also statistically significant (Table 1, Figure 2). The remaining MAOIs were associated with lower odds of reported glaucoma; however, this was not statistically significant (p > 0.05) (Table 1, Figure 2).
Isocarboxazid (42.37 ROR; 95% CI [2.17, 827.76]; p = 0.014) was the only MAOI to be significantly associated with greater odds of ACG (Table 1, Figure 3). The remaining MAOIs were associated with lower odds of ACG compared to acetaminophen; however, this was nonsignificant (p > 0.05) (Table 1, Figure 3). None of the MAOIs were significantly associated with ACG (IC025 < 0) (Table 1).
Consistent with the trends observed in ACG, isocarboxazid (466.64 ROR; 95% CI [8.74, 24907.50]; p = 0.0025; IC025 = 2.49) was also significantly associated with OAG and had a significantly greater odds ratio of OAG than acetaminophen (Table 1, Figure 4). Similarly, the same trends were observed for moclobemide (99.58 ROR; 95% CI [1.87, 4874.92]; p = 0.023; IC025 = 0.023) (Table 1, Figure 4). In contrast, phenelzine (21.98 ROR; 95% CI [0.43, 1110.70]; p = 0.12; IC025 = −1.78), selegiline (48.08 ROR; 95% CI [0.95, 2437.92]; p = 0.053; IC025 = −0.69), and tranylcypromine (34.34 ROR; 95% CI [0.68, 1738.42]; p = 0.077; IC025 = −1.16) were nonsignificantly associated with lower risk of OAG in comparison to acetaminophen (Table 1, Figure 4). It is important to note that for all MAOIs, there were zero reports of OAG (Table 1).
Evaluation of other antidepressants and their association with cataracts and glaucoma
Compared to acetaminophen, agomelatine, brexanolone, esketamine, opipramol, and tianeptine were all associated with a lower relative risk of both cataract and glaucoma; however, this was nonsignificant (all p > 0.05) (Table 1, Figures 1 and 2). Furthermore, the association of the aforementioned antidepressant agents with cataract and glaucoma was also not statistically significant (all IC025 < 0) (Table 1). It is important that the reported cases of cataract for agomelatine, brexanolone, and tianeptine were zero (Table 1). In addition, the reported cases for glaucoma for agomelatine, brexanolone, opipranol, and tianeptine were also zero (Table 1).
In contrast to the aforementioned trends, brexanolone (240.09 ROR; 95% CI [8.78, 6562.62]; p = 0.0012) and tianeptine (23.23 ROR; 95% CI [1.23, 438.52]; p = 0.036) were significantly associated with a greater relative risk of reported ACG compared to acetaminophen (Table 1, Figure 3). The odds ratio of ACG was greater with gomeline (3.04 ROR; 95% CI [0.17, 55.28]), esketamine (2.19 ROR; 95% CI [0.12, 39.76]), and gepirone (12.10 ROR; 95% CI [0.71, 242.65]) compared to acetaminophen; however, this was nonsignificant (p > 0.05) (Table 1, Figure 3). Only the strength of the association between brexanolone and ACG was significant (IC025 = 2.07) (Table 1). Notwithstanding the trends observed, there were zero reported cases of ACG associated with agomelatine, brexanolone, esketamine, opipramol, and tianeptine (Table 1).
With regard to the relationship between the aforementioned antidepressants and OCG, there was a significantly greater risk with brexanolone (2,644.33 ROR; 95% CI [38.32, 182475.75]), esketamine (72.78 ROR; 95% CI [2.95, 1793.50]), opipramol (144.24 ROR; 95% CI [2.81, 7400.88]), and tianeptine (55.90 ROR; 95% CI [4.92, 13311.83]) compared to acetaminophen (all p < 0.05) (Table 1, Figure 4). Agomelatine (33.47 ROR; 95% CI [0.66, 1694.23]; p = 0.080) was also associated with a greater odds ratio of OAG but this was nonsignificant (Table 1, Figure 4). Furthermore, the strength of the association between OAG with brexanolone, esketamine, opipramol, and tianeptine was also statistically significant (IC025 > 0) (Table 1).
Discussion
Our analysis revealed that there are mixed results with regard to the association between each antidepressant drug class and cataract and glaucoma. Notwithstanding the nonsignificant findings, a general trend was observed in which SSRIs, SNRIs SMSs, SARIs (trazodone), NDRIs (bupropion), TeCAs, and MAOIs may be associated with a lower risk of reported cataract compared to the reference drug, acetaminophen. In contrast, TCAs (all except clomipramine, nortriptyline, and trimipramine) and other select antidepressants (ie, brexanolone, esketamine, and opipramol) were associated with higher risk of cataract; albeit not all associations were statistically significant.
As it pertains to glaucoma, SSRIs, SNRIs, SARIs, TCAs, MAOIs, and other select antidepressants investigated (ie, brexanolone, esketamine, opipramol, and tianeptine) were associated with a greater odds ratio in comparison to acetaminophen. In contrast, there were mixed results regarding the relative risk of glaucoma with TeCAs. Specifically, the majority of TeCAs investigated were associated with a nonsignificant lower risk of glaucoma; however, mianserin and setiptiline were significantly associated with a greater risk of glaucoma than acetaminophen. Furthermore, there were also mixed results regarding the association between SMSs and NRIs with glaucoma since only 2 antidepressants of each drug class were FDA-approved. In contrast to the aforementioned trends, these findings indicate that all of the antidepressant drug classes investigated may be associated with a greater risk of ACG and OAG in comparison to acetaminophen.
Our results comport with previous findings from population-based studies regarding the association of antidepressants and glaucoma. Specifically, treatment with antidepressants may be associated with a greater risk of glaucoma diagnosis, including, but not limited to, ACG and OAG.Reference Chen, Ng and Chiu 14 , Reference Ciobanu, Dionisie and Neagu 25 In contrast, our findings using the FAERS pharmacovigilance database do not align with the results from previous population-based and meta-analyses with respect to the association between antidepressants and cataract.Reference Erie, Brue and Chamberlain 16 , Reference Etminan, Mikelberg and Brophy 17 , Reference Fu, Dai and Zhu 26 We found that not all antidepressant drug classes (ie, SSRIs, SNRIs SMSs, SARIs [trazodone], NDRIs [bupropion], TeCAs, and MAOIs) were associated with an increased risk of reported diagnoses of cataract. Rather only some TCAs, brexanolone, esketamine, and opopramol, were associated with a greater odds ratio of reported cataract cases.
Our findings regarding ROR should not be interpreted as indicative of causality. Furthermore, it is also important to note that the association between cataract formation and glaucoma with antidepressant use may also be confounded by other risk factors that were not controlled for including, but not limited to, smoking, family history, and use of additional therapeutic agents to treat other diseases or disorders.Reference Raju, George and Ve Ramesh 27 –Reference Constable, Al-Dasooqi and Bruce 31 It remains a testable hypothesis whether the use of antidepressant medication may directly induce the formation of cataracts and glaucoma or if other risk factors or comorbid conditions may be contributory (eg, diabetes mellitus).Reference Fang, Huang and Liao 32 , Reference Chen, Ng and Chiu 33 , Reference Costagliola, Parmeggiani and Semeraro 34
There are several methodological limitations to our study that should be considered. Firstly, the FAERS pharmacovigilance database consists of spontaneous reporting of adverse events; therefore, there is a possibility of duplicate reports when we searched both the generic and product/brand names of each antidepressant investigated. In addition, for the majority of adverse event cases, multiple suspected product names are listed; therefore, since cases are not routinely verified, there exists a possibility that the adverse event may be caused by another concomitant medication rather than the antidepressant of interest.
Furthermore, FAERS consists of many cases with missing patient demographic information, preventing the control for other factors, such as age, sex, weight, and so on. Lastly, for many antidepressants and the reference drug, acetaminophen, there were no reported cases; consequently, to calculate the ROR, we added 0.5 to the number of cases of cataract and glaucoma and total number of eye disorder cases. Consequently, this skews the proportion of positive and negative cases and ultimately affects the RORs and 95% CIs.
Conclusion
Herein, our results indicate that the majority of antidepressant classes are associated with a lower reporting risk of cataracts. In contrast, we did observe a higher reporting risk for select TCAs, brexanolone, esketamine, and opipramol. With regard to glaucoma (ie, glaucoma, ACG, and OAG), trend evidence existed associating most antidepressants with increased risk. Future research vistas should aim to determine mechanistic pathways linking antidepressants to the eye diseases studied herein. Clinicians should be vigilant of the potential of opthamologist-related disease possibly associated with some antidepressants as part of the safety surveillance of any persons prescribed with antidepressants. Although glaucoma is not a common safety concern with antidepressants, persons with anatomically narrow angles who do not have a patent iridectomy should be carefully monitored as they are at a higher risk for angle closure glaucoma with some antidepressants.
Author contributions
Writing – review & editing: B.C., R.B.M., M.V., J.D.R., A.K., G.H.L., K.T., R.H., S.W., S.L., T.G.R.; Conceptualization: G.H.L., R.S.M.; Formal analysis: G.H.L., S.W.; Investigation: G.H.L., R.S.M.; Methodology: G.H.L., R.S.M.; Visualization: G.H.L., S.W.; Writing – original draft: G.H.L., R.S.M.
Disclosures
Dr. Roger S. McIntyre has received research grant support from CIHR/GACD/National Natural Science Foundation of China (NSFC) and the Milken Institute; speaker/consultation fees from Lundbeck, Janssen, Alkermes, Neumora Therapeutics, Boehringer Ingelheim, Sage, Biogen, Mitsubishi Tanabe, Purdue, Pfizer, Otsuka, Takeda, Neurocrine, Sunovion, Bausch Health, Axsome, Novo Nordisk, Kris, Sanofi, Eisai, Intra-Cellular, NewBridge Pharmaceuticals,Viatris, Abbvie, Atai Life Sciences. Dr. Joshua D Rosenblat has received research grant support from the Canadian Institute of Health Research (CIHR), Physician Services Inc (PSI) Foundation, Labatt Brain Health Network, Brain and Cognition Discovery Foundation (BCDF), Canadian Cancer Society, Canadian Psychiatric Association, Academic Scholars Award, American Psychiatric Association, American Society of Psychopharmacology, University of Toronto, University Health Network Centre for Mental Health, Joseph M. West Family Memorial Fund and Timeposters Fellowship and industry funding for speaker/consultation/research fees from iGan, Boehringer Ingelheim, Janssen, Allergan, Lundbeck, Sunovion, and COMPASS. He is the Chief Medical and Scientific Officer of Braxia Scientific and the medical director of the Canadian Rapid Treatment Centre of Excellence (Braxia Health). Dr. Rodrigo B. Mansur has received research grant support from the Canadian Institute of Health Research; Physicians’ Services Incorporated Foundation; the Baszucki Brain Research Fund; and the Academic Scholar Awards, Department of Psychiatry, University of Toronto. Kayla M. Teopiz has received fees from Braxia Scientific Corp. Dr. Taeho Greg Rhee was supported in part by the National Institute on Aging (#R21AG070666; R21AG078972; R01AG088647), National Institute of Mental Health (#R01MH131528), and National Institute on Drug Abuse (#R21DA057540). Dr. Rhee serves as a review committee member for Patient-Centered Outcomes Research Institute (PCORI) and Substance Abuse and Mental Health Services Administration (SAMHSA) and has received honoraria payments from PCORI and SAMHSA. Dr. Rhee has also served as a stakeholder/consultant for PCORI and received consulting fees from PCORI. Dr. Rhee serves as an advisory committee member for International Alliance of Mental Health Research Funders (IAMHRF). Dr. Roger Ho received funding from National University of Singapore iHeathtech Other Operating Expenses (A-0001415-09-00). Dr Maj Vinberg has received consultancy fees from Lundbeck and Janssen Cilag the past 3 years. Gia Han Le, Sabrina Wong, Angela T.H. Kwan, Bing Cao, and Sonya Liao have no conflicts of interest to disclose.
Open access
