Gringras and colleagues ran a Phase III randomized, placebo-controlled crossover trial with 73 children aged 5-16 who had confirmed autism spectrum disorder and severe sleep problems. Each child used a weighted blanket and a look-alike control blanket for two weeks each, with sleep measured by wrist-worn activity monitors (actigraphy) and parent diaries. The weighted blanket did not increase total sleep time, did not reduce the time it took to fall asleep, and did not reduce nighttime awakenings compared to the control blanket on any measure - objective or subjective. However, most parents and children said they preferred the weighted blanket, suggesting the perceived benefit may be driven by expectations rather than actual sleep improvement. A commentary in Nature Reviews Neurology concluded there was "no evidence for efficiency of weighted blankets in improving sleep in children with ASD." Published in Pediatrics (Vol. 134, pp. 298-306).

Meth and colleagues at Uppsala University found that while weighted blankets increased melatonin release by 32% compared to light blankets, none of the expected sleep benefits materialized. Subjective sleepiness before bed was the same with both blankets. Total sleep duration was the same. Cortisol (a stress hormone) was unchanged. Oxytocin (sometimes called the "comfort hormone") was unchanged. Sympathetic nervous system activity (fight-or-flight response) was unchanged. This means the one physiological mechanism identified for weighted blankets - increased melatonin - did not actually produce better sleep or relaxation in this controlled laboratory study of 26 healthy adults. Published in the Journal of Sleep Research (Vol. 32, e13743, 2023).

Eron, Kohnert, and colleagues conducted a systematic review of weighted blanket research and could find only 8 studies that met basic quality standards - 4 at the highest evidence level (randomized controlled trials) and 4 at lower levels. After reviewing all available evidence, they concluded that while weighted blankets may help reduce anxiety, there is not enough evidence to recommend them for insomnia. The extremely small number of quality studies, combined with small sample sizes and inconsistent findings across studies, meant no firm conclusions could be drawn about sleep benefits. Published in the American Journal of Occupational Therapy (Vol. 74, 2020).

A 2024 meta-analysis assessed the quality of all available weighted blanket studies and found that half were rated as having high risk of bias. The main problem is fundamental and likely unsolvable: participants can always feel whether a blanket is heavy or light, so true blinding (where neither the participant nor researcher knows which treatment they received) is practically impossible. Since most outcome measures are self-reported questionnaires about how well someone slept or how anxious they felt, knowing you have the "real" treatment can inflate the results through the placebo effect. Even studies that used a lighter "sham" blanket could not fully solve this problem, because the weight difference is obvious. This means much of the positive evidence for weighted blankets may be driven by expectations rather than the blanket itself.

When looking at the weighted blanket research as a whole, a troubling pattern appears. Self-report measures like the Insomnia Severity Index and Pittsburgh Sleep Quality Index often show improvement, but when researchers use objective tools like actigraphy (wrist-worn motion sensors) or polysomnography (brain-wave sleep recordings), total sleep time frequently does not change significantly. For example, in the Ekholm 2020 trial, subjective insomnia scores improved dramatically, but the objective actigraphy findings were less clear-cut. In the Gringras 2014 trial, objective measures showed zero benefit despite parents preferring the weighted blanket. This gap between what people report and what instruments measure is a hallmark of placebo effects and is one of the biggest challenges to the claim that weighted blankets genuinely improve sleep.

The Ekholm 2020 trial - the most frequently cited evidence in favor of weighted blankets - used a control blanket weighing only 1.5 kg compared to the weighted blanket at 6 or 8 kg. Critics have pointed out that a difference this large (4 to 5 times heavier) makes it very easy for participants to figure out which group they are in, especially since participants knew the study was about whether weight helps sleep. If participants realized they had the "real" treatment, their expectations alone could explain the dramatic results. The study did not report whether participants were asked to guess which blanket they received (a standard check for successful blinding), so there is no way to know how many figured it out. This design weakness affects the credibility of the study's headline-grabbing 59.4% response rate.

The weighted blanket evidence base is remarkably thin. Key studies had 32 participants (Mullen 2008), 30 participants (Champagne 2015), 26 participants (Meth 2023), and 31 participants (Ackerley 2015). Even the largest randomized controlled trial (Ekholm 2020) had only 120 participants. A 2024 meta-analysis could pool only 553 total participants across 9 studies. A 2025 scoping review found just 12 randomized controlled trials in the entire history of weighted blanket research. By comparison, pharmaceutical sleep aids are typically tested on thousands of participants before approval. Small studies are known to overestimate effect sizes because a few unusual responses can strongly influence the results. The field simply does not have enough data to make confident claims about whether weighted blankets work.

Multiple clinical sources have identified safety concerns with weighted blankets that are rarely discussed in the marketing around these products. The weight can restrict chest expansion, posing a breathing risk for people with obstructive sleep apnea, asthma, or chronic obstructive pulmonary disease. People who are elderly, have limited mobility, or have neuromuscular conditions (such as muscular dystrophy or ALS) may be unable to remove the blanket if it becomes uncomfortable or causes distress, creating a suffocation risk. Several deaths have been linked to weighted blankets in care facilities. They are also contraindicated for people with coronary heart disease, very low blood pressure, or diabetes-related nerve damage. The weight can worsen circulation problems. For these populations, the potential risks outweigh any possible sleep benefits.

Losinski and colleagues tested weighted blankets with 2 four-year-old children who had autism and sleep disturbances, using a structured single-subject design with an 8-day baseline, 14-day intervention, and 7-day withdrawal phase. While there were some minor improvements in morning mood and a slight decrease in the time to fall asleep, the overall changes in sleep patterns were minimal. The researchers concluded that the findings were not strong enough to recommend weighted blankets for clinical use. Though the sample was very small, this study is notable because it used a rigorous withdrawal design (removing the blanket to see if any improvements disappeared) and found the blanket made little difference. Published in Children (Vol. 8, p. 10, 2021).