Does Weight Loss Reduce the Severity and Incidence of Psoriasis or Psoriatic Arthritis?

A Critically Appraised Topic

S.K. Mahil; S.M. McSweeney; E. Kloczko; B. McGowan; J.N. Barker; C.H. Smith


The British Journal of Dermatology. 2019;181(5):946-953. 

In This Article


Results of the Systematic Literature Search

Of 176 full-text articles reviewed, 14 met the inclusion criteria: nine randomized controlled trials (RCTs), three cohort studies and two case series (Table S3; see Supporting Information).[27–40] Nine studies examined lifestyle weight loss interventions in the treatment of psoriasis (n = 8) or PsA (n = 1) and three assessed pharmacological weight loss treatments in psoriasis. Two studies investigated the impact of bariatric surgery on the incidence of psoriasis or PsA.

Sources of Study Bias

Sources of bias were assessed using tools appropriate to each study design (Table S4 and Table S5; see Supporting Information). Nine RCTs and three cohort studies were assessed using the Cochrane Risk of Bias Assessment Tool (Figure S1; see Supporting Information).[26] Five of these 12 studies were at high or unclear risk of bias.

Figure S1.

Summary of risk-of-bias assessments.
Each coloured disc represents the review authors' (SMM and EW) consensus regarding each risk of bias item for the included studies.

The case series by Ahern et al.[37] and Buysschaert et al.[36] were assessed using the PROCESS guidelines. Both studies met the relevant reporting criteria stipulated by these guidelines but the provided level of evidence was considered low based on the study design.[25]

Lifestyle (Diet or Physical Activity) Weight Loss Interventions in Psoriasis

Seven RCTs and one cohort study examining the impact of lifestyle (diet or physical activity) weight loss interventions in obese patients with psoriasis were identified (Table S3; see Supporting Information). In four RCTs, participants were established on other psoriasis treatments prior to randomization. Two RCTs examined lifestyle interventions alongside newly initiated low-dose ciclosporin (2·5 mg kg−1; Gisondi et al.)[27] or narrowband ultraviolet B treatment (Kimball et al.).[32] Del Giglio et al.[33] randomized patients if they had achieved PASI 75 for at least 12 weeks on methotrexate, which was then stopped prior to the lifestyle intervention.

Six RCTs examined low-calorie diets alone. Naldi et al.[30] combined the diet with a physical activity programme, while Guida et al.[29] supplemented their diet with n-3 polyunsaturated fatty acids and minimized n-6 polyunsaturated fatty acid intake due to its potential anti-inflammatory effect.[41] The duration of interventions ranged from 12 to 24 weeks and the mean weight change across intervention groups was −3·0% to −14·8%.

One RCT did not achieve 5% weight loss using diet or exercise (Naldi et al.)[30] and one RCT demonstrated no improvement in psoriasis after diet-induced weight loss (Kimball et al.);[32] however, the study size was small (n = 30) and follow-up was limited to 12 weeks. Three RCTs demonstrated a statistically significant improvement in psoriasis severity following dietary interventions that achieved at least 5% weight loss (Al-Mutairi et al., Gisondi et al. and Guida et al.)[27–29] and one RCT identified a positive trend in the data (Jensen et al.)[34] (Table S3; see Supporting Information). Further analysis of the outcomes at each time point in two studies highlighted a dose–response relationship between weight loss and the reduction in psoriasis severity (Gisondi et al.[27] and Jensen et al.).[31]

Meta-analysis of six RCTs was undertaken after exclusion of the RCT by Del Giglio et al.,[33] as it initiated the lifestyle intervention after successful treatment with methotrexate. Our meta-analysis demonstrated that lifestyle weight loss interventions resulted in a significant improvement in the severity of psoriasis compared with control, as quantified by the change in relative PASI (i.e. PASI 75 response rates) or absolute PASI (ΔPASI) (Figure 2).

Figure 2.

Meta-analysis. (a) Forest plot summarizing the risk of achieving ≥ 75% improvement in Psoriasis Area and Severity Index (PASI) in individuals with psoriasis and obesity receiving lifestyle weight loss interventions vs. control. (b) Forest plot summarizing the mean change in PASI in individuals with psoriasis and obesity receiving lifestyle weight loss interventions vs. control. CI, confidence interval; M-H, Mantel–Haenszel; IV, inverse variance.

Sensitivity analyses demonstrated that exclusion of RCTs at high risk of bias (Al-Mutairi et al.[28] and Kimball et al.)[32] abrogated the effect of lifestyle-induced weight loss on PASI 75 response rates, although exclusion of high-risk studies (Guida et al.)[29] had minimal effect on pooled mean ΔPASI (Table S2; see Supporting Information). Subgroup analyses of heterogeneity could not be performed due to lack of sufficient data.

Lifestyle Weight Loss Interventions in Psoriatic Arthritis

An RCT by DiMinno et al.[35] showed that MDA was achieved by 8% more participants receiving a low-calorie diet compared with controls (Table S3; see Supporting Information). Diet-induced weight loss (> 5%) predicted a higher odds of achieving MDA regardless of treatment assignment (adjusted OR 4·20, 95% confidence interval 1·82–9·66; P < 0·001). A dose–response relationship was observed, as the odds of achieving MDA increased to 6·67 following greater (> 10%) weight loss (95% confidence interval 2·41–18·41; P < 0·001).

Pharmacological Weight Loss Treatments in Psoriasis

Studies assessing pharmacological weight loss treatments in psoriasis were very limited. Two small case series (Ahern et al.[37] and Buysschaert et al.)[36] and one RCT (Faurschou et al.)[38] examined the effect of GLP-1 agonists (liraglutide and exenatide) in obese patients with psoriasis (Table S3; see Supporting Information). The case series examined patients with type 2 diabetes and the RCT assessed normoglycaemic individuals. The duration of interventions ranged from 8 to 18 weeks, and the mean weight change among those receiving the intervention was −4·1% to −7·3%.

Both case series demonstrated a significant reduction in PASI in obese diabetic patients treated with liraglutide or exenatide (combined mean ΔPASI −3·3 ± 4·1) compared with baseline.[36,37] However, this effect was not observed in the RCT of obese nondiabetic patients treated with liraglutide vs. placebo,[38] suggesting that the beneficial effects of GLP-1 agonists in psoriasis are dependent on the presence of diabetes.

Bariatric Surgery in the Prevention of Psoriasis and Psoriatic Arthritis

Two large population-based cohort studies examined the impact of bariatric surgery on the onset of de novo psoriasis or PsA (Table S3; see Supporting Information). Egeberg et al. examined Danish health registry data spanning 16 years.[39] By comparing the incidence of psoriasis and PsA among participants before and after bariatric surgery, they demonstrated that gastric bypass but not gastric banding reduced the risk of new-onset psoriasis and PsA (Table S3).

Maglio et al. used Swedish health registry data to identify incident psoriasis and PsA in obese patients undergoing bariatric surgery over a median follow-up of 18 years.[40] Surgical interventions included gastric bypass (13%), vertical banded gastroplasty (68%) and gastric banding (19%). Controls were obese individuals who declined bariatric surgery and were group matched across potentially confounding variables including exact BMI, duration of obesity and presence of type 2 diabetes. The mean changes in BMI at 10 years were −17% and +1·7% in the surgery and control arms, respectively. Bariatric surgery reduced the risk of new-onset psoriasis; however, no significant difference in risk of psoriasis between the surgical interventions was observed. In contrast to Egeberg et al., no reduction in risk of de novo PsA following bariatric surgery was found.