To the Editor: Vitiligo is an inflammatory skin condition affecting 0.5% to 2%1 of the population. A recent meta-analysis of 28,325 patients with vitiligo identified association with diabetes mellitus and obesity.2 The direct effect of weight on vitiligo severity has yet to be elucidated. We hypothesized that with greater body mass index (BMI) there is greater skin friction, which could result in Koebner phenomenon on the central portion of the body.
Patients self-reported their sex, age, race, vitiligo regions of distribution, height, and weight. Only respondents providing both height and weight as well as proper units were included for analysis. BMI was calculated using the formula: weight (lb)/(height [in])2 × 703 (kg/m2). BMI classifications included underweight for patients with a BMI <18, normal for patients ≥18 but <25, overweight for >25 and <30, and obese for BMI ≥30. Analysis was performed with R (version 1.4).
The cohort (n = 1685) had a median age of 42 years, was predominantly female (72%), and White (64%), of which there were underweight (n = 51), normal weight (n = 820), overweight (n = 344), and obese (n = 344) individuals. As weight classification increased, vitiligo presentation (Table I) was more likely on the chest (P =.020), stomach (P <.001), axillae (P =.005), arms (P =.023), elbows (P =.048), wrists (P =.040), hands (P =.004), fingers (P <.001), genitals (P =.028), buttocks (P =.0015), ankles (P =.012), feet (P =.004), and toes (P =.011). Vitiligo presentation on the eyelids, lips, mouth, back, hips, legs, and knees did not have a relationship with one’s BMI classification.
Table I.
Characteristics of cohort stratified by weight classification
| Variable | N | Underweight, N = 51∗ | Normal, N = 820∗ | Overweight, N = 470∗ | Obese, N = 344∗ | P Value |
|---|---|---|---|---|---|---|
| Eyelids | 1420 | 20 (45) | 354 (52) | 201 (51) | 177 (58) | .2 |
| Lips | 1387 | 14 (33) | 258 (39) | 163 (43) | 121 (41) | .4 |
| Mouth | 1250 | 3 (7.0) | 51 (8.5) | 38 (11) | 31 (12) | .4 |
| Chest | 1429 | 27 (59) | 407 (59) | 227 (58) | 207 (68) | .020 |
| Stomach | 1385 | 15 (36) | 316 (48) | 187 (49) | 195 (65) | <.001 |
| Back | 1375 | 17 (40) | 391 (58) | 216 (58) | 154 (54) | .081 |
| Axillae | 1462 | 26 (58) | 477 (68) | 303 (74) | 238 (77) | .005 |
| Arms | 1461 | 23 (53) | 486 (69) | 298 (72) | 227 (75) | .023 |
| Elbows | 1467 | 28 (65) | 496 (70) | 285 (71) | 242 (78) | .048 |
| Wrists | 1469 | 27 (63) | 519 (74) | 320 (78) | 251 (80) | .040 |
| Hands | 1532 | 30 (70) | 584 (79) | 368 (85) | 281 (87) | .004 |
| Fingers | 1514 | 27 (64) | 564 (77) | 361 (85) | 273 (86) | <.001 |
| Hips | 1354 | 23 (53) | 371 (56) | 191 (52) | 177 (61) | .2 |
| Genitals | 1473 | 34 (74) | 502 (71) | 312 (77) | 245 (79) | .028 |
| Buttocks | 1316 | 22 (50) | 264 (42) | 160 (45) | 150 (53) | .015 |
| Legs | 1468 | 35 (76) | 499 (71) | 306 (74) | 236 (78) | .11 |
| Knees | 1405 | 24 (52) | 424 (63) | 248 (64) | 202 (68) | .2 |
| Ankles | 1408 | 22 (50) | 424 (63) | 262 (67) | 212 (71) | .012 |
| Feet | 1467 | 23 (53) | 491 (70) | 296 (72) | 240 (78) | .004 |
| Toes | 1372 | 20 (44) | 341 (52) | 213 (57) | 183 (62) | .011 |
n (%).
We found a significant stepwise relationship with the increasing presence of stomach and underarm lesions linked to increasing BMI quartile (Fig 1). For the first quartile, a 46% occurrence of vitiliginous lesions on the stomach was noted, but for the fourth quartile there was a 63% occurrence (P <.001). Similarly, for the underarms, the first quartile had 63% occurrence, and the fourth quartile had a 76% occurrence of vitiliginous lesions (P <.001).
Fig 1.
Quartile trends in localization and body mass index quartiles.
The effect of weight on vitiligo extent appears to be a promotional effect. Two recent studies examining patients with type 2 diabetic noted vitiligo in 12% vs 6% of nondiabetics in India, and 4.9% in diabetics3 vs 1.8% of nondiabetic patients in Iran.4 A picture is created of vitiligo aggravation by higher BMI in sites of dependence and friction, including folds, waistline, feet, and toes.
There are benefits and limitations identified within the survey process. Inherent in all survey processes is the possibility that patients may not provide accurate information. Specifically, it has been demonstrated that patient self-assessment of vitiligo is often not in agreement with physician estimates.5 To reduce this risk, we focused on self-identification by disease localization.
The relationship between vitiligo, the Koebner phenomenon, and inflammation in the metabolic syndrome requires prospective study. Our findings support prior data showing an association of vitiligo extent and metabolic syndrome. We demonstrated that increased BMI quartile confers increased risk of central disease (axillae, abdomen, chest, and arms). It remains to be seen if weight control can help control disease spread.
Conflicts of interest
Dr N. B. Silverberg has been an adviser or received honoraria from Amryt, Incyte, Lilly, Regeneron/Sanofi, and Verrica Pharmaceuticals. Dr J. I. Silverberg has received honoraria as a consultant and/or advisory board member for AbbVie, AOBiome, Arcutis, Alamar, Amgen, Arena, Arcutis, Asana, Aslan, BioMX, Biosion, Bodewell, Boehringer Ingelheim, Cara, Castle Biosciences, Celgene, Connect Biopharma, Dermavant, Dermira, DermTech, Eli Lilly, Galderma, GlaxoSmithKline, Incyte, Kiniksa, Leo Pharma, Menlo, Novartis, Optum, Pfizer, RAPT, Regeneron, Sanofi-Genzyme, Shaperon, and Union, speaker for AbbVie, Eli Lilly, Leo Pharma, Pfizer, Regeneron, and Sanofi-Genzyme, and institution received grants from Galderma and Pfizer. The other authors have no conflicts of interest to declare.
Footnotes
Funding sources: None.
Patient consent: Not applicable.
IRB approval status: An IRB-exempt study collected patient-reported data from support-group patients volunteers.
References
- 1.Krüger C., Schallreuter K.U. A review of the worldwide prevalence of vitiligo in children/adolescents and adults. Int J Dermatol. 2012;51(10):1206–1212. doi: 10.1111/j.1365-4632.2011.05377.x. [DOI] [PubMed] [Google Scholar]
- 2.Kang P., Zhang W.G., Ji Z.H., Shao Z.J., Li C.Y. Association between vitiligo and relevant components of metabolic syndrome: a systematic review and meta-analysis. J Dtsch Dermatol Ges. 2022;20(5):629–641. doi: 10.1111/ddg.14717. [DOI] [PubMed] [Google Scholar]
- 3.Raveendra L., Hemavathi R.N., Rajgopal S. A study of vitiligo in type 2 diabetic patients. Indian J Dermatol. 2017;62(2):168–170. doi: 10.4103/ijd.IJD_360_16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Afkhami-Ardekani M., Ghadiri-Anari A., Ebrahimzadeh-Ardakani M., Zaji N. Prevalence of vitiligo among type 2 diabetic patients in an Iranian population. Int J Dermatol. 2014;53(8):956–958. doi: 10.1111/ijd.12148. [DOI] [PubMed] [Google Scholar]
- 5.Coias J., Hynan L.S., Pandya A.G. Lack of correlation of the patient-derived Vitiligo Disease Activity Index with the clinician-derived Vitiligo Area Scoring Index. J Am Acad Dermatol. 2018;78(5):1015–1016. doi: 10.1016/j.jaad.2017.11.034. [DOI] [PubMed] [Google Scholar]