Do poultices and compresses (ginger, onion, cabbage, clay, charcoal) actually work?

What a poultice or compress does

A poultice is a moist mass of plant material (or clay) held on the skin; a compress is a cloth soaked in hot, cold or herbal liquid; a pack wraps a body part. The most dependable ingredient in all of them is heat or cold plus moisture. In two trials of 258 people with low back pain, heat wraps reduced pain after five days more than an oral placebo, and a heated blanket eased acute back pain at once. The reviewers rated this moderate evidence of a small, short-term effect, with too little evidence on cold⁶.

Cabbage leaves

In an 81-person trial of knee osteoarthritis, wraps of cabbage leaves worn at least two hours a day for four weeks lowered pain more than usual care, but no more than diclofenac gel¹. For breast engorgement, a Cochrane review found that cold cabbage leaves may reduce pain compared with routine care, but the certainty was very low².

Ginger and mustard compresses

Ginger compresses have only uncontrolled evidence. In a before-and-after study, 20 adults with osteoarthritis were randomized to a nurse-applied ginger compress or a ginger patch for a week, and pain scores fell 48%, with no placebo group⁴. In 18 healthy adults, ginger and mustard chest compresses made people feel warmer than warm water alone, and mustard raised skin temperature; the trial tested warmth, not illness⁵. In a 500-woman comparison, an herbal compress eased breast pain more than a hot compress, with low certainty; the review's herbal compresses included ginger, cactus and aloe, and hollyhock².

Clay and mud

A meta-analysis of 11 trials with 1,106 people found that mud therapy reduced short-term knee osteoarthritis pain and improved function⁸. These were trials of pelotherapy, the clinical use of therapeutic mud, mostly run in Italy and Turkey, not of a kitchen clay paste; we found no trial of homemade clay poultices. See bentonite clay.

Charcoal, onion and mustard

Charcoal trials involve sterile, manufactured wound pads containing activated carbon cloth, such as one that shrank wounds faster than a silver foam dressing; they say nothing about a charcoal poultice¹⁰. We found no trial of onion poultices. A case report describes a second-degree burn from a mustard seed mixture⁹.

Staying safe

Hot water bottles caused 80 burns in 15 months at one UK burns unit¹¹. For burn first aid, see minor burns and wounds; for castor oil packs, see castor oil packs. The recipe library has a Home Remedies collection.

Cabbage leaves

Lauche and colleagues randomized 81 adults (mean age 65.9) with stage II to III knee osteoarthritis to cabbage leaf wraps (at least 2 hours daily), diclofenac gel or usual care for four weeks. After four weeks, pain on a visual analog scale was lower with cabbage than with usual care (difference -12.1; 95% CI -23.1 to -1.0; P=0.033) but not significantly lower than with diclofenac gel (-8.6; 95% CI -21.5 to 4.4; P=0.19). Function, quality of life and chair-stand scores also improved against usual care. Only two adverse events were reported across both active groups¹. A cabbage wrap cannot be blinded, and one comparison group received usual care. The engorgement Cochrane update (21 studies, 2,170 women) could not pool results. Cold cabbage leaves may lower breast pain versus routine care (mean difference -1.03 on a 0 to 10 scale; 95% CI -1.53 to -0.53; 152 women; very low certainty) and versus cold gel packs (-0.63; very low certainty), and women were more often satisfied (RR 1.42; low certainty). Room-temperature and cold leaves could not be told apart, and the evidence for cabbage extract cream over placebo cream was uncertain². The eight-study meta-analysis of cold cabbage leaves found a significant fall in breast pain but not in engorgement³.

Ginger, mustard and herbal compresses

Therkleson randomly assigned 20 adults with chronic osteoarthritis to seven days of nurse-applied ginger treatment, either a hand-prepared compress or a standardized patch, and followed them by weekly questionnaires. Pain, fatigue, global effect and function scores fell by 48%, 49%, 40% and 31%, and further improvement continued over 24 weeks of optional self-treatment with the patch. With no placebo or untreated group, expectation and natural fluctuation cannot be separated from ginger⁴. Vagedes and colleagues ran a crossover trial in 18 healthy adults who received warm-water, mustard and ginger chest compresses in random order for up to 20 minutes. Back skin temperature was higher after mustard than after water or ginger (p<0.001), while self-perceived warmth was higher with both mustard and ginger than water⁵. In the Cochrane engorgement review, herbal compresses beat hot compresses for breast pain (mean difference -1.80 on a 0 to 10 scale; 95% CI -2.07 to -1.53; 500 women; low certainty), and skin irritation occurred in 2 of 250 women using herbal compresses versus none with hot compresses².

Heat

French and colleagues reviewed nine trials with 1,117 people with low back pain. In two trials of 258 people, heat wraps lowered pain after five days (weighted mean difference 1.06 on a 0 to 5 scale; 95% CI 0.68 to 1.45) compared with oral placebo. A heated blanket reduced pain at once in 90 people (difference -32.2 on a 100-point scale), and adding exercise to heat wraps reduced pain further in 100 people. There was too little evidence on cold, and the evidence comparing heat with cold conflicted⁶. For an acute internal hordeolum (an eyelid gland infection), a Cochrane review found no trials at all of warm compresses, lid scrubs, antibiotics or steroids, so any benefit is unproven⁷.

Mud, charcoal and mustard

Hou and colleagues pooled 11 randomized trials (1,106 patients) of mud therapy for knee osteoarthritis and found significant improvements in pain scores and in WOMAC pain, stiffness and function scores, with low heterogeneity. The 11 trials came from Italy, Turkey, France, Hungary, Iran and Spain and assessed short-term effects; therapeutic mud is hard to blind⁸. Probst and colleagues randomized 77 wound-care patients to a sterile pad with activated carbon cloth (n=38) or silver foam (n=39). Wound area shrank by 0.45 cm2 a day versus 0.2 cm2, though the carbon pad stayed on longer (9.5 versus 8.1 days); odor, pain and infection were similar¹⁰. A case report described a second-degree knee burn from a mustard seed mixture used for osteoarthritis; the authors note plasters should stay on no more than 15 minutes and that prolonged application can burn skin and damage nerves⁹.

Burns and infection

Begum and colleagues reviewed 80 hot-water-bottle burns at one London burns unit over 15 months: 41 were contact burns and 38 scalds, half were on the lower limb, 30 were full thickness and 37 patients needed an operation¹¹. The IDSA guideline covers skin infections from minor to necrotizing fasciitis and stresses prompt diagnosis and timely treatment¹².

Where the evidence conflicts

Cabbage leaves beat usual care for knee pain but not diclofenac gel, and cold and room-temperature leaves performed alike for engorgement, which hints that moisture and temperature, not cabbage itself, may be the active factors^1,2. Mud trials are positive but are for clinical mud therapy, not household clay⁸. Ginger's reported benefit comes from uncontrolled studies, while the controlled ginger and mustard trial measured only warmth^4,5. Most trials were open-label, small and short.

Bottom line

Confidence is low (2 of 5). Warmth is the best-supported part of a compress⁶. Cabbage leaves have small, weak evidence for knee osteoarthritis and engorgement^1,2,3. Ginger, onion, charcoal and homemade clay poultices are traditional practices without controlled evidence of benefit, and burns are the main documented harm^4,5,9,11.