Bidirectional association between vitiligo and melasma: A large-scale population-based study

Introduction

Vitiligo is a chronic autoimmune disease with a 0.5-2.0% prevalence.¹ It is characterised by depigmentation of the skin due to the destruction of melanocytes by autoreactive lymphocytes. Its pathogenesis potentially involves genetic predisposition,² oxidative stress,³ and melanocyte abnormalities.⁴ There is no long-term sustainable treatment.^5,6 Available options that aim to stabilise vitiligo lesions and restore skin pigmentation include corticosteroids, calcineurin inhibitors, phototherapy, and FDA-approved immunomodulatory drugs such as topical Janus Kinase (JAK) inhibitors.^7,8 Narrowband ultraviolet B (NBUVB) is the most common phototherapy, promoting melanocyte migration and increased melanin production.^9,10

Melasma is a chronic skin disorder characterised by symmetric hyperpigmented patches, mainly on the face.¹¹ It occurs predominantly in women (about 90%). Melasma is more common in darker skin types, affecting mainly individuals of Hispanic, African, and Middle Eastern origin.¹² Exacerbating factors include UV exposure, pregnancy, hormonal treatments, and genetics.¹³ The pathogenesis involves melanocyte hyperactivation and increased melanin production following sun exposure, with solar elastosis seen in most patients.¹⁴

Vitiligo and melasma represent opposite pigmentation disorders. Both carry a significant psychosocial burden and require long-term treatment because of their high recurrence rates.^5,11,15 Previous studies have examined treatments for both conditions,^16,17 but none have explored their possible association, especially given that treatment for one condition may induce the onset of the other.

We aimed to test the hypothesis of an independent bidirectional association between vitiligo and melasma through a population-based study, based on observations of patients affected by both conditions.

Methods

Results

Risk of melasma in patients with vitiligo

After excluding patients (n=209) and comparators (n=1,605) with a pre-existing diagnosis of melasma, the retrospective cohort design included 24,227 patients with vitiligo and 117,600 comparators, thus maintaining a 1:5 ratio. The total follow-up time was 208,835 person-years for patients with vitiligo and 993,813 person-years for comparators. During the follow-up period, new-onset melasma was diagnosed in 288 patients with vitiligo and 872 comparators, yielding an incidence rate of 1.38 per 1,000 person-years among patients with vitiligo (95% CI 1.22-1.54) and 0.88 per 1,000 person-years among comparators (95% CI 0.84-0.96).

The crude risk of developing new-onset melasma was 60% greater among patients with vitiligo than comparators [HR 1.62, 95% CI 1.40-1.88, P<0.001; Figure 1 and Table 2]. On Cox proportional hazards model, after controlling for confounders including demographic variables, comorbidities, and phototherapy [presented in Suppl. Table 1], vitiligo was found to be a significant independent risk factor of melasma [adjusted HR 1.61, 95% CI 1.38-1.87, P<0.001; Table 3]. Stratified analyses revealed that pre-existing vitiligo was a predictor of new-onset melasma in females (HR 1.62, 95% CI 1.39-1.9, P<0.001) and in patients aged 20 years or older (P<0.001), as presented in Table 2.

Close

Figure 1:

Cumulative incidence of melasma among patients with vitiligo and matched comparators. The shaded area surrounding the cumulative incidence curve represents the 95% confidence intervals. Bold values indicate statistically significant differences.

Export to PPT

Table 2: Risk of incident melasma in patients with pre-existing vitiligo, stratified by demographic and clinical parameters

Subgroup	Melasma in patients with vitiligo n (%) (n=24,227)a	Melasma in comparators n (%) (n=117,600)a	HR (95%CI)	P value
All	288 (1.2)	872 (0.7)	1.62 (1.40-1.88)	<0.001
Age (years)
<20	13 (0.2)	64 (0.2)	1.03 (0.53-1.99)	0.93
20-40	183 (2.6)	614 (1.8)	1.49 (1.24-1.8)	<0.001
>40	92 (1)	194 (0.4)	2.2 (1.65-2.93)	<0.001
Sex
Male	15 (0.1)	34 (0.06)	1.97 (0.99-3.9)	0.05
Female	273 (2.2)	838 (1.4)	1.62 (1.39-1.9)	<0.001
BMI (category)
Underweight	10 (0.4)	31 (0.3)	1.32 (0.65-2.7)	0.4
Normal	147 (1.2)	489 (0.8)	1.53 (1.2-1.95)	<0.001
Overweight	66 (1.2)	180 (0.8)	1.35 (0.83-2.18)	0.2
Obese	65 (1.4)	172 (0.8)	3.1 (1.84-5.17)	<0.001
Smoking
Yes	74 (1.2)	241 (0.7)	1.7 (1.1-2.6)	0.01
No	214 (1.2)	631 (0.8)	1.6 (1.3-1.9)	<0.001
Hormonal treatmentb
Yes	201 (3.7)	615 (2.6)	1.59 (1.31-1.92)	<0.001
No	72 (1.0)	273 (0.8)	1.71 (1.19-2.50)	0.004

^aExcluding a pre-existing diagnosis of melasma. ^bFemale participants only. Significant values appear in bold. BMI: Body mass index, CI: Confidence interval, HR: Hazard ratio. Bold values indicate statistically significant differences.

Table 3: Multivariate analyses assessing the bidirectional association between vitiligo and melasma

Model type	Incident melasmaa		Incident vitiligob
	HR (95% CI)	P value	OR (95% CI)	P value
Unadjusted	1.62 (1.40-1.88)	<.001	1.57 (1.34-1.85)	<.001
Adjusted (all)c	1.61 (1.38-1.87)	<.001	1.32 (1.12-1.55)	<.001
Adjusted (female only)d	1.58 (1.35-1.86)	<.001	None

^aRetrospective cohort design. ^bNested case-control design. ^cAdjustment includes demographics, body mass index, smoking status, thyroid disorders, and phototherapy exposure. ^dAdjustment includes pregnancy and hormonal treatment in addition to demographics, body mass index, smoking status, thyroid disorders, and phototherapy exposure. HR: Hazard Ratio, CI: Confidence Interval. Bold values indicate statistically significant differences.

In multivariate analysis limited tofemale patients hormones (birth control pills) and pregnancy relevant, including pregnancy and hormonal treatment as covariates, the HR remained significant [adjusted HR 1.58, 95% CI 1.35-1.86, P<0.001; Table 3]. The mean (SD) time from diagnosis of vitiligo, or the concordant matching date in comparators, to diagnosis of melasma was shorter in patients with vitiligo: 10.8 (4.3) years and 11.6 (4.1) years, respectively (P=0.01).

Discussion

The current population-based epidemiologic study shows a bidirectional association between vitiligo and melasma. Vitiligo conferred a 60% increased risk of developing incident melasma, regardless of phototherapy and hormonal treatment, while patients with pre-existing melasma had 30% increased odds of subsequently developing vitiligo, regardless of hydroquinone treatment. Pre-existing melasma was more prevalent in patients with vitiligo than controls (0.9% vs. 0.5%). Patients with a dual diagnosis were mostly female, were older at the diagnosis of vitiligo, and had an increased frequency of hypothyroidism.

A potential underlying factor explaining the association of vitiligo and melasma may be the treatment modalities used; that is, the treatment applied for one condition may provoke the development of the other. NBUVB is considered the standard of care for vitiligo due to its superior efficacy and fewer adverse effects compared to psoralen UVA (PUVA), in addition to its ability to alleviate oxidative stress-induced damage to melanocytes.²² Melasma, however, is known to be aggravated by light exposure, and its treatment mainly relies on strict photoprotection.²³

In the present study, the rate of utilisation of phototherapy was higher in patients with vitiligo (8.6%; 7.3% NB-UVB, 1.3% PUVA) than in comparators without vitiligo (0.6%; 0.5% NB-UVB, 0.1% PUVA). Using a retrospective cohort design, we found that phototherapy did not increase the risk of developing melasma [HR 1.2, 95% CI 0.80-1.85, P=0.40; Supplementary Table S1]. Sensitivity analysis excluding participants who did not receive phototherapy showed that the risk of developing melasma in patients with vitiligo was similar to the crude risk (HR 1.64, P<0.001). The lack of a contributory effect of UVB on the risk of development of melasma may be partly explained by accumulating evidence that visible light and UVA1 are the most significant triggers for melasma,^24,25 whereas most of the patients with vitiligo in our cohort were treated with NBUVB.

Melasma is a multifactorial disorder where sex and sex hormones play a key pathogenetic role. Hormonal imbalances due to pregnancy, hormonal replacement therapy, and hormonal contraceptives stimulate melanogenesis.¹¹ Consistent with the female predominance of melasma in the general population, in the present study, melasma was more prevalent in females than males among both patients (2.2% vs. 0.1%) and comparators (1.4% vs. 0.05%). However, on subgroup analysis, the risk of developing melasma was significant only in females with vitiligo compared to female comparators (HR 1.62, P<0.001).

Hormonal treatment was more prevalent in females with vitiligo than in female comparators (46.0% vs. 42.8%, P<0.001), while the number of pregnancies did not differ between the groups [Table 1]. On multivariate analysis, the risk for developing melasma was higher among female patients with vitiligo independently of their hormone treatment status [Table 3].

Hydroquinone, especially when used in a triple combination compound with a mild-potency steroid and a retinoid, is considered the gold standard of melasma treatment.^23,26 Hydroquinone is a phenolic compound with antioxidant properties that disrupts melanogenesis by inhibiting tyrosinase, which can also cause selective damage to melanocytes. The lightening effect typically takes months to become visible, necessitating long-term use.²³ There are only a few reports describing leukoderma following hydroquinone use,^27,28 in contrast to the expected depigmentation seen with hydroquinone’s structurally related chemical, monobenzyl ether of hydroquinone. Therefore, hydroquinone treatment is currently considered safe for patients with vitiligo, though its use should be evaluated on a case-by-case basis.²⁹

In the present cohort, hydroquinone was used significantly more frequently by patients than comparators (5% vs. 3.3%, P<0.001). It is possible that it was applied as an additive to “smoothen” the distinctive appearance of vitiligo lesions.²⁹ Nevertheless, although hydroquinone was significantly associated with an increased odds of development of vitiligo in patients with melasma (OR 1.54, P<0.001), on both sensitivity and multivariate analyses, melasma was associated with vitiligo independently of hydroquinone treatment status.

Autoimmune thyroid disease (both hypothyroidism and hyperthyroidism) is a well-recognised comorbidity of vitiligo.³⁰ It has also been identified as a provoking factor of melasma.^12,31 As expected, in the present cohort, both hypothyroidism and hyperthyroidism were more prevalent among patients with vitiligo [Table 1]. In our retrospective cohort design, both hypothyroidism and hyperthyroidism were associated with an increased risk of developing melasma [Supplementary Table S1], and in the nested case-control design, both were associated with increased odds of developing vitiligo in cases of pre-existing melasma [Supplementary Table S2]. These findings suggest that autoimmune thyroid diseases should be considered when investigating the association between these two chronic pigmentary disorders. Nonetheless, in our multivariate model, after adjusting for hypothyroidism and hyperthyroidism along with other relevant clinical variables, the independent risk of developing melasma in patients with vitiligo remained nearly unchanged (HR 1.6) in both the entire cohort and among female patients. In the nested case-control study, adjustment weakened the independent association between melasma and subsequent vitiligo, although it remained significant (adjusted OR 1.32, P<0.001).

Vitiligo and melasma are considered phenotypically opposite ends of the pigmentation spectrum, with a complete loss of melanocytes (vitiligo) at one pole and increased melanin deposition by hyper functional melanocytes (melasma) at the other pole. Nevertheless, the two relatively common acquired pigmentation disorders share several pathogenetic links, such as oxidative stress, which might explain their connection.¹¹ Specifically, both external and internal stimuli can lead to the generation of excess reactive oxygen species in the skin. External factors include UV radiation and air pollution, and intrinsic factors involve processes such as melanogenesis in melanocytes.^11,32-35

Given that both vitiligo and melasma are characterised by an imbalance in oxidative stress, triggering melanocyte destruction in vitiligo and melanogenesis in melasma may help explain the findings that having one condition increases susceptibility to the other. Antioxidant treatments could potentially treat or prevent both conditions. Polypodium leucotomas has shown benefits in both vitiligo and melasma due to its antioxidant and photoprotective properties.^36,37

This study directly investigates the association between vitiligo and melasma by evaluating the bidirectional relationship through both risk estimates (retrospective cohort) and odds ratios (nested case-control). To our knowledge, this bidirectional association has not been previously reported. Previous retrospective studies examined the effects of treatments for one condition on the other.^16,17 In one study, using tranexamic acid for melasma alongside phototherapy for vitiligo in 32 patients resulted in an 80% improvement in melasma without worsening vitiligo.¹⁶ Another study evaluated NBUVB phototherapy’s impact on vitiligo repigmentation in two cohorts: one with only vitiligo and one with a dual diagnosis. It found that vitiligo lesions improved more in patients with dual diagnosis for facial and limb lesions, while truncal lesions showed better repigmentation in the vitiligo-only group.¹⁷

Strengths

The study’s key strengths are a large cohort size and dermatologist-confirmed diagnoses, enhancing statistical power and validity.

Limitations

This study is limited by the lack of data on several important clinical parameters, including disease extent, severity (as assessed by validated instruments such as the Vitiligo Area Scoring Index [VASI] and the Melasma Area and Severity Index [MASI]), vitiligo activity status, lesion distribution, and co-localisation with facial melasma. These details were unavailable due to the exclusive use of computerised datasets, which did not capture comprehensive clinical evaluations. The study also did not account for skin phototype and natural sun exposure, both significant factors in melasma development, and the use of over-the-counter melasma treatments, which could influence vitiligo development. Nevertheless, sensitivity analysis and adjustments for putative covariates confirmed the independent bidirectional association between vitiligo and melasma, with primary-care clinic matching helping to control for differences in healthcare access and to ensure comparable environmental exposures. Future prospective studies could address these limitations and enhance the generalisability of the findings across diverse ethnic and population groups. They would also allow for better accounting of residual confounders that could not be adequately addressed due to the retrospective nature of the current study.

Conclusion

This large-scale study demonstrated a bidirectional association between vitiligo and melasma. Patients with vitiligo are at increased risk of developing melasma, and patients with melasma are at increased odds of developing vitiligo. Dermatologists should be aware of this association for the identification of subgroups at risk. Further prospective studies are needed to explore potential treatments for patients with a dual diagnosis.