Systemic and Metabolic Profile of Sagging Eye Syndrome: A Comparative Analysis

Abstract

Purpose:

Sagging eye syndrome (SES) is an acquired strabismus characterized by distance esotropia (ET) and cyclo-vertical deviation, affecting older people. This study identified demographic, systemic, and ocular factors associated with SES and compared them to other forms of ET in older adults.

Design:

Retrospective cohort study.

Methods:

Demographic data, systemic comorbidities, ocular history, and strabismus were reviewed in esotropic patients at least 55 years old diagnosed with SES between 2013 and 2024 at a single institution, and compared with age-matched controls diagnosed with other forms of ET.

Results:

A total of 383 patients were studied, of whom 67 had SES and 316 had ET without SES ET. Male SES patients were significantly older at 81.5±10 years (standard deviation) than esotropic males without SES at 76±10 years (p = 0.04), and symptom onset was significantly older for SES in both genders (male: 72 ± 10 years vs. 68 ± 11 years, p = 0.02; female: 75 ± 9 years vs. 69 ± 9 years, p = 0.04). Women with SES had significantly lower mean body mass index (BMI) at 23 ± 4 than esotropic women without SES at 25±5 (p = 0.02). Logistic regression demonstrated that lower BMI was associated with higher odds of SES in women, with the odds of SES increasing by approximately 9.4% for each 1-unit decrease in BMI (OR = 0.91, 95% CI: 0.84-0.99). Osteopenia was more prevalent at 77% among esotropic women with SES than women without osteopenia at 51% (p=0.003), and remained an independent risk factor for SES after adjusting for age (OR=1.81, 95% CI: 1.02-3.23, p=0.04) and BMI (95% CI, 1.14-6.18; p=0.02). There were no significant differential associations with SES for gender, smoking, alcohol use, hypertension, hyperlipidemia, type 2 diabetes, cataract surgery, glaucoma, or use of progressive addition spectacles.

Conclusions:

SES is associated with advanced age at symptom onset in both genders, and with lower BMI and osteopenia in women.

Introduction:

Sagging eye syndrome (SES) is an increasingly recognized cause of acquired strabismus with diplopia in the elderly population. First described by Rutar and Demer in 2009,¹ it is characterized by a combination of distance esotropia (ET) and/or cyclo-vertical strabismus, with diplopia that often worsens in lateral gazes but is absent at near.² The condition results from elongation of the lateral rectus-superior rectus band ligament and subsequent inferior displacement and elongation of the lateral rectus path.³ While imaging can provide additional diagnostic specificity, SES is generally diagnosed clinically.

Several factors contribute to the risk of SES, with age being the most significant. The incidence of SES increases markedly with advancing age, particularly over age 70 years.^4-6 Characteristic external features often accompany this condition, including a deep superior sulcus, blepharoptosis, and baggy lower eyelids.^3,7 Additionally, affected individuals often have a history of blepharoplasty or similar cosmetic procedures.^1,7,8 Adnexal changes are likely the result of age-related alterations in connective tissue further reinforcing the connection between SES and aging.

There are notable geographic variations in gender distribution in SES. A Korean study of 23 SES patients demonstrated a balanced sex ratio (52% male vs 48% female), contrasting with Western studies that report female predominance (60-85% female) in Caucasian populations.^6,4 Body mass index (BMI) has also been identified as a significant factor in SES. A study by Iida et al. demonstrated that lower BMI is independently associated with a higher prevalence of SES, even after adjusting for age and sex.⁹

Beyond these considerations, the risk factors associated with SES are not yet fully understood. Our study aims to investigate additional variables potentially linked to SES development, including demographic, socioeconomic, behavioral, metabolic, and systemic factors.

Methods

This retrospective cohort study analyzed patients at least 55 years old who presented with ET at the Stein Eye Institute, University of California, Los Angeles (UCLA) between 2013 and 2024. We compared patients diagnosed with SES to age and sex matched patients diagnosed with other forms of ET who served as controls. The study was conducted in accordance with the Declaration of Helsinki and complied with HIPAA guidelines. Approval was obtained from the Institutional Review Board of UCLA.

Inclusion criteria for both groups were age ≥50 years and documentation of ET, along with complete clinical data including ocular history, systemic comorbidities, and current medications. Although patients aged 50 years and older were screened, the youngest individual meeting all inclusion criteria was 55 years old. SES was defined by the presence of ET greater at distance than at near, typically exceeding 5Δ, although in some cases the difference was minimal when the primary ET was small but increased in lateral gazes, along with characteristic age-related adnexal features such as deep superior sulcus and aponeurotic ptosis. Exclusion criteria for the SES group included: myopia exceeding −5.00 diopters, history of childhood strabismus, neurologic or restrictive strabismus, prior strabismus surgery, orbital trauma, Graves’ orbitopathy, myasthenia gravis, and syndromic connective tissue disorders. Pseudophakic patients were excluded if pre-lensectomy refractive data were unavailable or demonstrated myopia exceeding −5.00D. Pre-lensectomy refraction was required to confirm absence of high myopia; those with greater than −5.00 D before cataract surgery were reclassified into the non-SES group.

Controls were matched to SES patients in a 1:4 ratio using a ±5-year age tolerance within each sex. If fewer than four eligible controls were available for a given case, all qualifying matches within the specified range were included. The resulting matched cohorts showed no significant age difference between SES and control groups for either sex (p>0.05).

Data was extracted from electronic medical records. Demographic information included age, gender, and BMI. Systemic comorbidities were recorded, including hypothyroidism, arterial hypertension (HTN), hyperlipidemia (HLD), type 2 diabetes mellitus (T2DM), osteopenia/osteoporosis, and obstructive sleep apnea (OSA). Additional features included medication use, smoking status, and alcohol consumption habits. Ocular factors were assessed, including best-corrected visual acuity (BCVA), refractive error, horizontal and vertical strabismic angles in primary gaze at distance and near, type of spectacles used (single vision lenses, progressive lenses, bifocals, or reading glasses), history of glaucoma, and history of cataract surgery.

Statistical analyses were performed using GraphPad Prism software version 10 (La Jolla, CA, USA). Continuous variables were compared using independent t-tests or Mann-Whitney tests when data were not normally distributed, while categorical variables were analyzed using the chi-square test. Logistic regression analysis was conducted to evaluate the association between systemic and demographic variables and the diagnosis of SES, adjusting for potential confounders. Sex-specific multivariable logistic regression models were developed. Odds ratios (OR) with 95% confidence intervals (CI) were reported for all regression models. Statistical significance was set at p<0.05.

Results

A total of 383 consecutive patients were included, with 67 patients with SES and 316 patients with other forms of ET. Among the 316 patients in the non-SES group, 201 had a clearly identifiable alternative etiology for ET. The most common causes were prior strabismus surgery (n=82), sixth nerve palsy (n=27), high myopia (n=32), and thyroid-associated orbitopathy (n=20). Additional causes included ocular or orbital trauma (n=10), superior oblique palsy (n=6), fourth nerve palsy (n=6), third nerve palsy (n=2), myasthenia gravis (n=5), connective tissue disorders (n=4), and other neurologic or syndromic etiologies such as Duane or Parinaud syndromes, Parkinson’s disease, stroke, and meningioma (n=4). The remaining 115 patients had primary, comitant ET that did not meet criteria for SES. A detailed gender-based comparison of baseline characteristics, systemic conditions, treatments, and ocular and strabismic features is presented in Tables 1-3.

Table 1.

Patient Characteristics

	Female (n=228)			Male (n=153)
	SES Mean ± SD, range (n=43)	P value	Non-SES Mean ± SD, range (n=185)	SES Mean ± SD, range (n=24)	P value	Non-SES Mean ± SD, range (n=129)
Study Age (years)	78±11(58-99)	0.11	75±11(55-100)	81.5±10(65-98)	0.04	76±10(55-102)
Symptom Onset Age (years)	72±10(53-93)	0.02	68±11(43-92)	75±9(62-92)	0.04	69±9(45-93)
Body Mass Index	23±4(17-31.5)	0.02	25±5(16-47)	25.6±3(20-34)	0.26	27±5(16-42.5)
Ethnicity (Caucasian %)	34(79)	0.9	146 (79)	19 (79)	0.8	95(73.5)
Smoking (%)	17(40)	0.47	60(33.5)	9(37.5)	0.65	53(44)
Alcohol Consumption (%)	26 (60)	0.73	101 (57)	13(54)	0.35	79(66)

SES, Sagging eye syndrome

Table 3.

Ocular Characteristics and Strabismus Measurements

	Female (n=228)			Male (n=153)
	SES (n=43)	P value	Non-SES (n=185)	SES (n=24)	P value	Non-SES (n=129)
History of Cataract Surgery , n(%)	24(55)	0.12	78(42)	14(58)	0.27	57(45)
History of glaucoma , n(%)	16(38)	0.35	51(30)	7(30)	0.9	33(28)
Use of PAL, n (%)	17(39)	0.84	50(27)	6(25)	0.9	39(30)
Mean ET in primary gaze, distance (Δ ± SD, range)	10±7(0-35)	0.2	12±14(0-85)	9±6(2-25)	0.2	9±10(0-50)
Mean ET in primary gaze, near (Δ ± SD, range)	0±0.3(0-2)	0.23	1±7(0-60)	0	0.21	1±4(0-38)
Mean HT in primary gaze, distance (Δ ± SD, range)	1±1.5(0-8)	0.08	2±4(0-30)	1±1.5(0-4)	0.15	3.5±7(0-41)

Abbreviations: PAL, Progressive addition lenses; ET (Δ ± SD), Esotropia (Prism Diopters ± Standard Deviation); HT (Δ ± SD), Hypertropia (Δ ± Standard Deviation). P- values represent significance of differences between adjacent columns.

The SES cohort had demographic characteristics distinct from the control group. Among males, SES patients were significantly older than esotropic patients with mean age 81.5±10 years, compared to a mean 76±10 years in the control group (p=0.04). Among females, there was no significant age difference; SES patients had mean age 78±11 years, compared to 75±11 years in the control group (p=0.11). Symptom onset was significantly later in SES patients. In females, symptoms began at 72±10 years, significantly older than 68±11 years in the control group (p=0.02). In males, symptom onset was at 75±9 years in SES, significantly older than 69±9 years in the control group (p=0.04). There was no significant difference in gender distribution, with 64% of patients in the SES group and 58% in the ET group being female (p=0.45).

Females with SES had a significantly lower average BMI of 23±4 kg/m² compared to 25±5 kg/m² in those with other forms of ET (p=0.02, Fig 1). In the 208 women who had BMI data, logistic regression demonstrated that lower BMI was associated with higher odds of SES in women, with the odds of SES increasing by approximately 9.4% for each 1-unit decrease in BMI (OR = 0.91, 95% CI: 0.84-0.99). In contrast, the difference in BMI among males was not statistically significant, with an average of 25.6 in the SES group and 27 in the ET group (p=0.26). Smoking was reported in 38.8% of patients with SES and 37.5% of those with ET, not significantly differing between groups (p=0.89). Alcohol use was similarly comparable, occurring in 59% of the SES group and 61% of the ET group (p=0.78). Ethnic distribution was also similar, with 79% of the SES group and 76.8 percent of the ET group identifying as Caucasian (p=0.69).

Fig. 1:

Body mass index in sagging eye syndrome (SES) and control (non-SES) patients, stratified by gender. Each dot represents an individual patient. SD – standard deviation.

The systemic disease profiles of the two groups differed significantly. Osteopenia was significantly more prevalent among women with SES than among women with other forms of ET (77% vs 51%; p=0.003), whereas in men, osteopenia prevalence did not differ between groups (Fig 2). After adjustment for age, the odds of osteopenia were 1.8 times greater in females with SES than in those with other forms of ET (95% CI, 1.03-3.23; p=0.04). This relationship remained significant after additional adjustment for BMI, with females having SES demonstrating 2.7-fold higher odds of osteopenia than females without SES (95% CI, 1.14-6.18; p=0.02). In contrast, among men, SES, BMI, and age were not significantly associated with osteopenia (p>0.1). There were no significant differences in the prevalence of hypothyroidism, HTN, HLD, T2DM, depression, migraine, or OSA (p>0.05 for all).

Fig. 2:

Prevalence of osteopenia in sagging eye syndrome (SES) and control (non-SES) patients.

There were no significant differences in medication use between groups for statins, proton pump inhibitors, estrogens, systemic steroids, topical steroids, systemic immunosuppressants, topical immunosuppressants, prostaglandins, or other topical glaucoma eye medications.

Cataract surgery had been previously performed in 57% of patients with SES compared to 44% of patients with ET (p=0.06). Mean spherical equivalent refractive error was −0.25± 2.51 ranging from −3.9 to +5.9D in the SES group. In subjects without SES, mean spherical equivalent refraction was −0.97±3.9D, ranging from −17.0 to +9.0. This difference was not statistically significant (p=0.29). Progressive addition lenses were used by 34% (23/67) of patients with SES, compared to 28% (89/316) of patients without SES. Among patients with SES, 18% used single vision lenses and 15% used bifocals, while none used single vision reading glasses. Of cases without SES, 21% used single vision lenses, 10% used bifocals, and 2% used reading glasses. The distribution of lens types did not differ significantly among groups (p=0.35)

Discussion

While the present results confirm and extend the observation that strabismus symptoms develop later in life in SES than with other forms of ET,^2,4,10,5 we here report the novel finding that this age difference is driven primarily by males. The age at diagnosis in women with SES does differ significantly from women with other forms of ET. Previous studies have typically reported overall mean age and gender distribution in SES but have not provided sex-stratified age data.^4,6 It may be that SES results from progressive degeneration of orbital connective tissues that occurs earlier in women. Interestingly, although there are several reports of female predominance in SES,⁴ the current study found no significant difference in gender distribution between the SES and ET groups. In both groups, there was a slight predominance of females, comprising 64% of the SES group and 58% of the ET group. These findings align with those of Park et al., who reported a nearly equal gender distribution in their SES cohort, with 12 male patients (52%) and 11 female patients (48%).⁶ However, the small sample size in that study limits generalizability. This discrepancy may reflect variations in study populations or suggest that gender might not be as strong a risk factor for SES as previously thought. It is also possible that gender is less of a contributing factor in Asian populations, such as those studied by Park et al., than in predominantly Caucasian populations. The difference may also be due to exclusion in the current study of myopes of greater than −5D, a group highly prevalent in Asian populations.

In females, we found significantly lower BMI in patients with SES than with other forms of ET. This aligns with recent findings by Iida et al.,⁹ who reported an independent association between lower BMI and higher SES prevalence, even after adjusting for gender. However, we did not find this association in men, perhaps because of differences in baseline BMI across populations. In the Japanese cohort studied by Iida et al., BMI was lower overall for both genders: the mean BMI in the SES group was 21.2 ± 3.3 kg/m², compared to 23.1±3.0 kg/m² in the non-SES group (p=0.003). These lower baseline BMI values in the Japanese population may increase sensitivity to BMI-related anatomical or physiological factors that contribute to SES, making associations more readily detectable than in populations such as ours that have higher average BMI. Iida et al. suggested that because orbital fat volume is positively correlated with BMI, lower BMI and reduced orbital fat may exacerbate age-related weakening of the pulleys due to structural changes, including collagen and elastin depletion within the supportive pulley tissues.⁹

The markedly higher prevalence of osteopenia among females with SES, even after adjusting for age and BMI, represents a novel finding. This association probably reflects a common factor influencing involution of both bones and connective tissues. In contrast, no significant association between SES and osteopenia was observed among men, which can partially be explained by the absence of BMI differences between male groups. In ophthalmology, osteoporosis has been associated with an increased risk of dry eye disease.^11,12 However, there is no literature linking osteoporosis to ptosis or dermatochalasis. The observation that women with SES were diagnosed younger than men, had lower BMI, and a higher incidence of osteopenia, raises the possibility that a combination of fat distribution and hormonal status may contribute to the onset and progression of SES.

In the current study there were no significant differences between SES and other forms of ET in systemic comorbidities or medication use, suggesting that these factors are unlikely to account for the observed differences between groups. Prevalence of glaucoma was similar between SES and patients with other forms of ET. Interestingly, a higher proportion of patients with SES had undergone prior cataract surgery compared to those with other forms of ET, although this difference did not reach statistical significance in our sample. The trend toward significance (p=0.06) suggests that with a larger cohort, a true association between SES and prior cataract surgery may be revealed. In the literature, cataract surgery has been suggested as a potential trigger for the decompensation of preexisting esophoria.¹³

In the current study, use of progressive addition lenses (versus single vision, bifocals, or readers) did not significantly differ between SES and patients with other forms of ET (p=0.35). This finding contrasts with the observations reported by Chen et al., who suggested progressive addition lens use as an independent risk factor for divergence insufficiency (DI) esotropia in a large retrospective study over four decades, attributing this to chronically increased convergence tonus induced by underuse of the low-add segment of progressive lenses.¹⁴ One possible reason for the differing results between our study and Chen et al. is the difference in patient age and diagnosis. Chen et al. included a broader spectrum of patients with DI ET, encompassing younger individuals as well as older adults, whereas our cohort consisted exclusively of patients diagnosed over age 55 years.

This study has several limitations. Its retrospective nature and single-center design restrict the generalizability of our findings. Additionally, most patients in both groups were Caucasian, reflecting the predominant demographic of our facility, which may limit the applicability of our results to other populations. High myopes were also excluded from the current study.

Table 2.

Systemic Comorbidities and Medications

Comorbidities (%)	Female (n=228)			Male (n=153)
	SES Mean ± SD, range (n=43)	P value	Non-SES Mean ± SD, range (n=185)	SES Mean ± SD, range (n=24)	P value	Non-SES Mean ± SD, range (n=129)
Depression	14(35)	0.47	69(43)	2(9)	0.15	27(24)
Hyperlipidemia	27(67)	0.6	100 (61)	17(73)	0.99	84(73)
Hypertension	16(39)	0.11	88(53)	18 (78)	0.6	80 (70)
Hypothyroidism	17(42)	0.36	56(34)	4 (17)	0.5	13 (12)
Migraine	15(37)	0.17	42(26)	4(18)	0.48	13(12)
Obstructive Sleep Apnea	6(15)	0.22	13(8)	3(13)	0.28	27(24)
Osteopenia	31(77)	0.003	82(51)	6(27)	0.8	27(24)
Type 2 Diabetes Mellitus	5(12)	0.9	20(12)	3(13)	0.4	26 (23)
Drugs (%)
Statins	19(47)	0.9	80(48)	14(64)	0.9	73(64)
Proton Pump Inhibitors	12(30)	0.85	53(32)	10(45)	0.9	49(45)
Estrogens	17(42)	0.27	54(33)	0	0.9	1(0.9)
Systemic steroids	7(17)	0.62	24(14)	5(22)	0.6	32(29)
Topical steroids (eyedrops)	5(12)	0.9	24(14)	3(13)	0.76	20(18)
Systemic immunosuppressants (Adalimumab, methotrexate, tacrolimus)	0	0.34	7(4)	0	0.9	1(0.9)
Topical immunosuppressants (cyclosporine)	5(12)	0.8	26(15)	1(4)	0.9	8(7)
Prostaglandins (drops)	11(26)	0.17	28(16)	4(17)	0.74	16(14)
Glaucoma eyedrops	8(19)	0.64	27(16)	6(26)	0.38	20(17)

Abbreviations : SES, Sagging eye syndrome; PO/IV/Inj, Oral/Intravenous/Injection.

Table of Contents Statement.

In this retrospective cohort study of 383 patients aged ≥55 years with esotropia, 67 with sagging eye syndrome (SES) were compared with 316 controls. SES was associated with later symptom onset overall; however, women with SES were diagnosed earlier than men with SES and demonstrated lower body mass index and higher prevalence of osteopenia compared with all other esotropic patients, both remaining independent risk factors.

Data Availability

In conformity with requirements of the funding agency the U.S. Public Health Service, National Institutes of Health, the anonymized raw data for this study are provided for public access via Zenodo with the link 10.5281/zenodo.16879035.