Honey

What it is

Honey is a sweet syrup made by bees, used as a food and a traditional cough remedy [1][2]. It is chemically a mix of sugars with small amounts of proteins, enzymes, organic acids, vitamins, minerals and plant compounds [1]. In cough trials, children received single bedtime doses (10 g in a 300-child blinded trial [3]), and adults took honey up to three times a day for up to 14 days [4].

What the research says

A 2018 Cochrane review pooled six randomised trials in 899 children aged 12 months to 18 years [2]. On 7-point scales where lower is better, honey reduced cough frequency by 1.05 points versus no treatment (95% CI -1.48 to -0.62; 2 studies, 154 children; moderate certainty) and by 1.62 points versus placebo (95% CI -3.02 to -0.22; 2 studies, 402 children; moderate certainty) [2]. Honey was similar to dextromethorphan (mean difference -0.07; 95% CI -1.07 to 0.94; low certainty) and better than diphenhydramine (-0.57; 95% CI -0.90 to -0.24; 1 study, 80 children; low certainty) [2]. The advantage held for the first three days, and most children were treated for a single night [2].

Individual trials show the same direction: three honey types beat a date-extract placebo in 300 children aged 1 to 5 [3]; buckwheat honey beat no treatment in 105 children while dextromethorphan did not [5]; two Iranian honeys improved cough and sleep in 87 children [6]; and a Nigerian trial of 84 children found honey reduced cough frequency and severity more than diphenhydramine [7]. But a 2022 double-blind multicentre trial in 161 Japanese children found acacia honey no better than a honey-flavoured syrup placebo [8], and a 2026 review of seven trials (576 children) found no significant differences from cough medicines except for children's sleep in its main analysis, with benefits appearing only in post-hoc sensitivity analyses that the authors call hypothesis-generating [9]. In a subgroup analysis, honey beat dextromethorphan on two endpoints [9].

In adults, a 2021 review of 14 studies in upper respiratory infections found honey beat usual care on combined symptom scores (mean difference -3.96; 95% CI -5.42 to -2.51; 3 studies), cough frequency (standardised mean difference -0.36; 95% CI -0.50 to -0.21; 8 studies) and cough severity (-0.44; 95% CI -0.64 to -0.25; 5 studies), but not versus placebo (standardised mean difference -0.63; 95% CI -1.44 to 0.18; 2 studies) [10]. A pragmatic trial of 194 adults with acute bronchitis found no difference from usual care: median days with moderate-to-severe cough were 5 with usual care (IQR 4 to 8.75) and 6 with honey (IQR 3.5 to 7) [4].

Doses used in studies

Children's trials used single bedtime doses: 10 g of eucalyptus, citrus or labiatae honey in 300 children aged 1 to 5 [3], and buckwheat honey in 105 children [5]. Some trials gave three consecutive bedtime doses [7] or two consecutive nights of acacia honey [8]. The Cochrane authors noted that most children received treatment for a single night [2]. In the adult trial, honey was taken three times a day for up to 14 days [4]. These are study doses, not recommendations.

Safety and interactions

Never give honey to an infant under 12 months: health authorities uniformly advise against it because honey is the one consistently confirmed food source of infant botulism [11]. Reviews of that evidence led the CDC to recommend that honey not be given to infants younger than 12 months [12]; in 1979, Clostridium botulinum spores were found in honey fed to three California infants and in 10% (9 of 90) of honey samples tested [13]. Infant botulism is rare but serious: California recorded 1,345 cases from 1976 to 2016 (6.5 per 100,000 live births per year), 88% in babies six months or younger, and more than 99% of cases were hospitalised [14]; deaths are under 1% with intensive care [11]. Worldwide, 2,943 cases were reported from 2007 to 2021, with honey consumed in 19.7% of cases outside the US and 3.8% in the US [15].

In trials in older children, side effects were mainly gut symptoms, reported by 12% on honey versus 11% on placebo (risk ratio 1.91; 95% CI 1.12 to 3.24), and nervousness, insomnia or hyperactivity in 9.3% on honey versus 2.7% on dextromethorphan [2]. Honey allergy is uncommon: among 46 pollen-allergic people given 30 g of honey, 26% reported minor, mostly subjective symptoms versus 41% given a placebo syrup, with no serious reactions [16]. The cough trials reported no medicine interactions and no safety data for pregnancy or breastfeeding [2][4]. No upper limit or toxic dose was identified for honey; trial doses were small, from a single 10 g dose in children to three daily doses for up to 14 days in adults [3][4].

Caveats and open questions

Blinding is impossible because honey tastes like honey, and two of the six Cochrane trials were at high risk of performance and detection bias [2]. The best-blinded trials disagree: honey beat a date-extract placebo in one [3] and matched a honey-flavoured syrup placebo in another [8]. The 2026 review found benefits only after post-hoc sensitivity analyses, which its authors call hypothesis-generating [9]. Most children were treated for a single night, so longer-term benefit is unknown [2]. The main adult trial found no benefit over usual care [4].