Psoriatic arthritis and environment

What the research found.

• One study followed 90189 people with newly recorded psoriasis, and 1409 later had a record of psoriatic arthritis. Against a body mass index below 25, the adjusted odds were 1.79 from 1.46 to 2.19 at 25.0 to 29.9, then 2.10 from 1.67 to 2.63 at 30.0 to 34.9. At 35.0 and above they reached 2.68, from 2.09 to 3.43.

  Green and colleagues, British Journal of Dermatology, 2020

• One study followed 75395 people with psoriasis for a mean of 5 years, and 976 developed psoriatic arthritis. That works out at 26.5 cases per 10000 person-years. Against a body mass index below 25, the relative risks were 1.09 from 0.93 to 1.28, then 1.22 from 1.02 to 1.47, then 1.48 from 1.20 to 1.81.

  Love and colleagues, Annals of the Rheumatic Diseases, 2012

• One study of 21429 people with psoriatic arthritis looked at where they lived. High social deprivation reached 25.7 percent among Black patients, against 2.3 percent Asian, 12.5 percent White, and 17.3 percent other. High disease activity followed that order too, at 40.2 percent against 25.8, 29.6, and 33.5 percent.

  Dowell and colleagues, Arthritis and Rheumatology, 2026

• One study of 960 people with psoriatic arthritis found 166 who had failed four or more advanced drugs. Four things were associated with that after adjustment, which are female sex at an odds ratio of 2.3 from 1.4 to 3.8, and axial disease at 2.1 from 1.2 to 3.6. The other two are depression at 2.0 from 1.1 to 3.7, and obesity at 1.7 from 1.0 to 2.8.

  Haberman and colleagues, Arthritis Research and Therapy, 2025

• One genetic study tested vitamin D against a list of several autoimmune diseases. Higher genetically predicted vitamin D was associated with lower odds of psoriasis, at 0.91 from 0.85 to 0.97 per 10 ng/ml. It found nothing for the autoimmunity group as a whole, at 0.99 from 0.95 to 1.03.

  Zhao and colleagues, Seminars in Arthritis and Rheumatism, 2023

Body weight, in two large studies

Two studies followed people who had psoriasis and watched which of them went on to develop psoriatic arthritis. Between them they covered more than 165000 people, and both asked about body weight. The first followed 90189 people with newly recorded psoriasis, and over the study 1409 of them had a record of psoriatic arthritis. It sorted them into weight bands and compared each band against a body mass index below 25.

The odds rose with each band it went through, reaching 1.79 from 1.46 to 2.19 at 25.0 to 29.9. At 30.0 to 34.9 it was 2.10, from 1.67 to 2.63. At 35.0 and above it reached 2.68, from 2.09 to 3.43, which is the figure everybody quotes.

The second study followed 75395 people with psoriasis for a mean of 5 years, and 976 of them developed psoriatic arthritis. That works out at 26.5 cases per 10000 person-years, and it used the weight bands the first study used. Its numbers were a good deal smaller, with relative risks of 1.09 from 0.93 to 1.28, then 1.22 from 1.02 to 1.47, then 1.48 from 1.20 to 1.81. The lowest band crosses 1, so it isn't significant.

What to make of two answers that far apart

A reader could take the larger set and say the risk nearly triples, or take the smaller set and say it rises by about half. Both of those would be quoting a real study. That is the awkward part, and it's why this page gives both sets rather than picking one.

The size itself stays unsettled, because the two studies used different populations, different follow-up, and different adjustments. Nothing here settles which is closer to the truth. What they agree on is the form of it, because in both the risk rose in step as weight rose, across three bands, with no reversal anywhere.

That rising line is one of the better signs that a link is real rather than an accident of one dataset. So the direction is worth taking seriously, and the multiplier itself isn't worth memorizing. Anybody quoting you one figure without the other is quoting half the evidence.

What neither study tested

Neither study assigned anybody a weight, because both measured what people already weighed and then followed them. That single fact decides what the figures can mean. The larger study did report one thing pointing forward, which is what happened when weight changed over time.

A body mass index that fell year on year over a 10-year period was associated with lower risk than one staying constant. That's a useful observation and it isn't a trial. People whose weight falls over ten years differ from people whose weight doesn't, in health, in circumstances, and in much else the study can't fully adjust for.

So nothing on this page shows that losing weight prevents psoriatic arthritis. It says that the evidence points that way and that it has never been put to a trial. Your own team is the right place to take a question like that.

Weight after the diagnosis

One study asked a different question, and it looked at 960 people who already had psoriatic arthritis. Of those, 166 had failed four or more advanced drugs. Four things were associated with that failure once the numbers had been adjusted for everything else.

Female sex came out at an odds ratio of 2.3, from 1.4 to 3.8, and axial disease at 2.1, from 1.2 to 3.6. Depression came out at 2.0, from 1.1 to 3.7. Obesity came out at 1.7, from 1.0 to 2.8, which is the weakest of the four.

Read that last interval carefully, because its lower bound falls right on 1.0, which makes it the weakest of the four. The finding is real enough to report and fragile enough to hold loosely. What's useful about it is the company it keeps, since depression appears in that list right beside the obesity. Both are things a treatment plan can address rather than things a patient simply has.

Where you live

One study of 21429 people with psoriatic arthritis asked about their neighborhoods. It's the only source on this page about the environment in that wider sense, and it's also the largest study here. High social deprivation was unevenly spread across the group.

It reached 25.7 percent among Black patients, 17.3 percent among other groups, 12.5 percent among White patients, and 2.3 percent among Asian patients. High disease activity followed that order too, reaching 40.2 percent among Black patients and 33.5 percent in other groups. It was 29.6 percent among White patients and 25.8 percent among Asian patients.

About 7 percent of the patients lived in districts holding a historical grading from the last century, applied long before any of them were ill. Smoking, obesity, high social deprivation, federal insurance, and high disease activity were all commoner in districts historically graded hazardous. This page reports those figures and offers no explanation for them. What they do show is that the environment in this disease isn't only a question of chemicals and dust.

Vitamin D

One genetic study tested vitamin D against a list of autoimmune diseases, using genetic variants that predict blood levels. No level was measured directly anywhere in it. Higher predicted vitamin D was associated with lower odds of psoriasis, at an odds ratio of 0.91 per 10 ng/ml, from 0.85 to 0.97, and that result replicated in independent data.

It found nothing for the autoimmunity group as a whole, which came out at 0.99, from 0.95 to 1.03, on an interval falling squarely across 1. So there's a signal in the skin disease and no signal in the wider group. Psoriatic arthritis wasn't reported separately anywhere in the study. This page gives no vitamin D target and no dose, and nothing here is a reason to start taking one.

What these trials didn't test

Every study on this page measured something that people already had. Not one assigned a weight, a neighborhood, or a vitamin, and not one removed anything and watched what happened. So each of these findings is a link rather than a cause of anything.

A link is a reason to look further and a reason to raise something with your team. It isn't proof that changing the thing changes the disease itself. The one partial exception is the falling body mass index finding, and even that is still an observation.

It describes what happened to people who lost weight rather than what happens when somebody sets out to lose it. Tell your own team what applies to you. Ask what it means for your own plan, because that's what these figures are genuinely good for, and it's more than most environmental evidence on this site supports.

Common misconceptions.

Cautions specific to this condition.

• Ask your team what your own weight means for your treatment. Don't take a figure from this page. One study found obesity associated with failure of four or more advanced drugs, at 1.7 from 1.0 to 2.8.
• Read a rising series of odds ratios as a stronger signal than a single number. Both weight studies here rose in step across three bands.
• Don't read either weight study as showing that losing weight prevents arthritis. Neither assigned anybody a weight or removed one.
• Mention smoking and where you live if they apply. One study found both travelling with higher disease activity across 21429 people.
• This page gives no diet, no weight target, and no vitamin D dose. Nothing here has been tested by changing what anybody did.
• Treat the difference between 2.68 and 1.48 as real uncertainty. Two large studies of the same question got answers that far apart.

Questions patients ask.