Bedding materials and early infant wheezing: A randomised controlled trial

Abstract

Background

Various observational studies have suggested that infants and young children who regularly sleep in synthetic bedding materials are more likely to experience wheezing and asthma, while children who use feather duvets and/or feather pillows are less likely to wheeze.

Methods

In Wellington, New Zealand, we conducted a three‐armed, parallel, randomised trial of 460 infants who were assigned to use different bedding materials: synthetic, wool or feather bedding in the form of sleepsacks from 3 months of age to 2 years of age to test the hypothesis that children exposed to feather materials are less likely to develop wheezing. Pregnant women were recruited before birth. Parents were unaware of the primary research hypothesis and were told this was a study of child warmth and wheezing. We have reported wheezing (parental and GP), a variety of respiratory health parameters and atopic status at 2 years.

Results

One hundred and forty‐seven infants received a synthetic sleepsack, 150 wool and 144 feather. We have found no significant differences in reported or doctor‐diagnosed wheezing or other respiratory health measures by bedding material used. For frequency of wheezing presentation at GP surgery, there was a significant increased rate for children using feather materials compared to synthetic, relative rate 2.00 (95% CI: 1.14, 3.52).

Conclusion

This study does not support earlier observational studies that suggest higher rates of wheezing for children using synthetic bedding or lower rates for feather materials, at least for early childhood wheezing. Our study suggests that the explanation for the observational study findings may lie in selection bias, where the parents of at‐risk children avoid feather bedding materials.

Key message.

Observational studies have suggested that feather bedding use in childhood may be protective for wheezing. We have undertaken a randomised trial of three bedding materials on the respiratory health of children during the first 2 years of life. We find no significant difference between synthetic, wool and feather bedding materials for the primary respiratory outcome. Observational study findings may be the result of incomplete control of selection bias.

1. INTRODUCTION

In 1995, Strachan and Carey, in a UK case–control study of severe asthma amongst 11–16‐yearyear‐old children, showed that non‐feather bedding and furry pet ownership were strongly associated with severe asthma. ¹ Feather pillows were protective, being associated with an adjusted 60% reduction in the risk of severe wheezing, while feather duvets showed a 22% reduction. Furry pet ownership was associated with a 70% increased risk of severe asthma. These estimates were derived from children whose parents denied making any changes to their child's bedding or avoiding a pet because of allergy. The authors concluded that either their questionnaire did not capture the avoidance of feather bedding or synthetic bedding might pose an unknown increased risk, possibly from volatile organic compounds. Following this, we showed that feather bedding material had some eightfold less house dust mite (HDM) allergen by weight compared with synthetic bedding. ² We subsequently attributed this difference to the enclosing fabric envelope being more closely woven in feather bedding compared with synthetic bedding, thereby reducing the ingress of mites. ³ In a follow‐up letter, Strachan and Carey showed that a difference in HDM allergen was an unlikely explanation for the lower asthma risk amongst those with feather bedding, as the protective effect against severe asthma was of similar magnitude amongst those with and without HDM sensitisation. ⁴ However, Custovic et al. ⁵ have shown that synthetic bedding materials do contain higher levels of cat and dog allergens. Other observational studies have shown protective effects of feather bedding materials on childhood wheezing ⁶ and others have shown increased wheezing and asthma associated with synthetic bedding. ⁷ In 2000, a systematic review was attempted but found no trials to review. ⁸

In 2011, we reviewed many of these observational studies ⁹ , ¹⁰ and concluded that there was considerable observational evidence supporting both contentions regarding wheezing (feathers protective, synthetic increased risk). We suggested that this might be related to increased allergens and fungi or volatile organic compounds in synthetic bedding or reduced allergen in feather bedding and suggested that clinical trials were required to explore the issue further. Our aim in this study was to explore the development of wheezing and atopy in a randomised trial of three different bedding materials from birth to 2 years. We wished to observe whether there were any important differences between bedding materials for early wheezing, with a particular interest in possible protective effects of feather bedding.

2. METHODS

2.1. Participants

Pregnant women (n = 460) were recruited into the study via parent support programmes (Naku Enei Tamariki, ‘My Children’), local healthy housing support programmes, social media and word of mouth from the wider Wellington region. Ethics approval was obtained from the Central Health and Disability Ethics Committee (HDEC 15/CEN/181), and informed written consent was obtained from a parent/guardian. The trial was registered with the Australian and New Zealand Clinical Trials Register ACTRN12615001039572.

2.2. Study design

Pregnant women, who were not planning to move regions during the 2‐year study period and were willing to be randomly assigned to use any one of the bedding types for their infants, were eligible to take part. The research team enrolled families, and they were randomised (by a team member not assessing outcomes) to receive one of three bedding types, each made from different materials made into a sleeping bag garment (sleepsack), into which infants were secured. The materials were feather, wool or synthetic insulation materials. The feather and synthetic materials were encased in a cotton cover and the wool in a woolen cover; see Table 1 for specifications.

TABLE 1.

Population demography by bedding materials.

	Synthetic	Wool	Feather	Total
	N = 147 (%)	N = 150 (%)	N = 144 (%)	N = 441 (%)
	n (%)	n (%)	n (%)	n (%)
Male	65 (44.5)	81 (54.3)	76 (52.7)	222 (50.6)
Average gestation (weeks)	39.6	39.4	39.3	39.3
Caesarean birth	37 (27)	42 (29.2)	39 (28.5)	118 (28.2)
First born	62 (42.2)	63 (42.0)	60 (41.7)	185 (42.0)
Breastfed	132 (96.4)	142 (98.6)	134 (97.1)	408 (97.4)
Exclusively breastfed	81	95	80	256
≥12 weeks	(59.1)	(66.9)	(58.0)	(62.7)
Family history of any asthma, eczema or hayfever	115 (78.2)	119 (79.3)	117 (81.3)	351 (79.6)
Household size (average number of people)	4.0	4.2	4.1	4.1
Daycare attendance	69 (59.5)	69 (58.0)	72 (57.1)	212 (58.7)
Ethnicity*
NZ European	115 (55.8)	108 (52.2)	110 (58.5)	333 (55.4)
NZ Māori	44 (21.4)	44 (21.3)	38 (20.2)	126 (21.0)
Pacific Island	8 (3.9)	8 (3.9)	7 (3.7)	23 (3.8)
Other	39 (18.9)	47 (22.7)	33 (17.6)	119 (19.8)
Mother completed high school	88 (64.7)	101 (71.6)	100 (73.0)	289 (69.8)
Mother smoked in pregnancy	10 (7.4)	12 (8.6)	7 (5.1)	29 (7.1)
Median annual household income before tax**	$100,001–$120,000	$80,001–$100,000	$80,001–$100,000	$80,001–$100,000

^*Some participants responded with more than one ethnicity; 42 participants did not respond.

^**49 participants either did not know their household income, did not wish to reveal their household income, or did not respond.

In order to minimise the health effects of poor‐quality housing on participants, all were provided with a check of underfloor and ceiling insulation, and low income participants were offered subsidised insulation. All families were also provided with a NZ$200 electricity credit and, if they did not have one, a thermostatically controlled bedroom heater and a temperature monitor. They were asked to keep the child's bedroom between 18 and 21 degrees Celsius during the winter months. ¹¹

Parents were asked to start using the sleepsack when the infants were 3 months old and to use their discretion regarding particularly hot nights during summer, when they would not have used any bedding covers. Some children (91, 20%) outgrew their original sleepsack during the study and were given a larger sleepsack of the same type.

We arranged for three types of sleepsack to be constructed. First, a synthetic polyester fill with a cotton shell; second, a feather and down fill again with a more closely woven cotton shell; and third, a double layer merino wool sleepsack.

2.3. Outcomes

The primary outcome measure was a parental report of ever wheezing during the first 2 years of life. The question asked, ‘has you child ever had wheezing or whistling in the chest’ Questionnaires were asked at enrolment, 3–5 months after birth, and at 2 years of age, over the phone or in person. Secondary outcomes included General Practitioner (GP) diagnosed wheeze ever, GP reported inhaled asthma medication, otitis media, croup and any antibiotic prescriptions during the first 2 years and the frequency of GP visits for wheeze by bedding type. Parents were also asked about their child's history of eczema, ‘Has your child ever had eczema?’.

Questionnaires were used to collect covariate information on prioritised ethnicity (Prioritised ethnicity means where more than one ethnicity is reported they are prioritised Māori > Pacific > Others), environmental tobacco smoke exposure (ETS) in the home, first born children and family history (parents or siblings) with any allergic disease, asthma, hay fever or eczema. Questionnaire information was also collected on compliance with sleepsack use, both the overall duration of use and frequency per week. Atopic status was determined at 2 years by skin prick tests.

2.4. Clinical visits

To determine atopic sensitisation, a skin prick test (SPT) was performed at 2 years of age for those children with parental consent. Allergens tested included Dermatophagoides pteronyssinus, cat dander, mixed grasses, egg white, peanut, cow's milk (Dome/Hollister‐Stier, Spokane, WA, USA), together with a positive (histamine 10%) and negative solute control, using a standard method of administration and measurement. ¹² Mean wheal diameter was measured using both the longest and orthogonal diameters, recorded to the nearest 0.5 mm. A positive reaction was defined as a mean wheal diameter of 3 mm or greater to any allergen, with atopy by SPT defined as one or more positive reactions. Training ensured that the coefficients of variation in histamine wheal size for each operator were <.2.

2.5. General practitioner records

Secondary outcomes included doctor diagnosed wheeze and were collected by electronic transfer of each child's complete general practitioner medical record for the first 2 years of life. Child records were sent as pdf files and initially coded by researchers for key terms and all asthma medications. Key terms included all variants of the spelling of wheeze, wheezing, rhonchi and asthma. If there was correspondence from a hospital or after hours clinic where the child was reported to have wheezing or was treated with asthma medications then these children were included. Any issues of interpretation of a GP record that was unclear were reviewed by JC who was blinded to the bedding allocation. For a child to be included as ‘doctor diagnosed wheeze’ they must also have also had evidence in the record that some asthma medication had been prescribed on at least one occasion. The children's GPs were also unaware of the bedding allocation. The other health terms were also searched with their various possible spellings.

2.6. Analysis

Study power: The study was powered to detect differences in the primary outcome for the two bedding types, wool and feathers, versus the control bedding type, synthetic. From previous New Zealand studies, we estimated parental reported wheezing by 2 years at 35%. We aimed to recruit ~150 children per group with a 12% drop out (leaving 132 per group) which would give 88% power to detect an absolute reduction in wheezing from 35% to 20% by any bedding type (alpha .05).

All analyses were undertaken using SAS, version 9.4. Chi‐squared tests were used to test for trend. PROC LOGISTIC was used to estimate odds ratios for binary outcomes (patient reported wheeze: yes/no; doctor diagnosed wheeze: yes/no; inhaled medications prescribed: yes/no) for bedding type feather versus synthetic and for wool versus synthetic. PROC GENMOD was used to estimate relative rates for counts of GP visits for wheeze symptoms, applying a Poisson model with an overdispersion factor to allow for extra‐Poisson variation in these counts. Relative rates were estimated for bedding type feather versus synthetic and for wool versus synthetic. For the binary outcomes, Wald chi‐square statistics tested the significance of the effect of the bedding type on the outcome. For the counts of events, an F statistic was used to test the effect of the bedding type on the counts of outcomes.

3. RESULTS

Figure 1 shows the flow of subjects through the study. Five hundred and eighty‐eight families were initially contacted, with 460 enrolled and randomised to one of the three bedding groups shortly after birth. Infants were randomised using random number tables and by staff not involved directly in data collection or analysis. Four hundred and eleven (89%) infants completed the study, and their parents completed the outcome questionnaire at 2 years. General practitioner records were obtained for 324 infants (70%), and skin prick tests were performed on 237 children (52%).

FIGURE 1.

Participant flow through the study.

Table 1. shows the population demographics and family history of allergic disease for the child participants and that these are evenly distributed across the study groups. 21% identified as Māori, similar to the national average for the age group. A family history of allergic disease was reported for 80% of children, and this is similar to other birth population surveys in New Zealand. ¹³

Table 2 shows the primary outcome of ever parental reported wheeze and GP diagnosed wheeze by bedding type. Overall, parents reported that 38% of infants had wheezed during the first 2 years of life, with a 28% non‐significant increase amongst those sleeping in feather materials compared to synthetic and a 12% reduction in those sleeping in wool. Synthetic bedding materials are the most common in New Zealand. A similar pattern is seen for GP diagnosed wheeze, with a non‐significant increased wheeze amongst feather bedding users and a similar pattern for wool, although again not significant.

TABLE 2.

Parental and GP reported wheezing ever by bedding material.

Bedding type	Parent reported wheeze					GP diagnosed wheeze
	Yes	Total	%	OR	95% CI	Yes	Total	%	OR	95% CI
Synthetic	51	136	37.5	1*		27	106	25.5	1*
Feather	59	136	43.4	1.28	0.79–2.10	35	104	33.7	1.48	0.82–2.70
Wool	48	139	34.5	0.88	0.54–1.44	27	114	23.7	0.91	0.49–1.68
Total	158	411	38.4			89	324	27.5

Note: Effect of bedding material on parental reported wheezing p = .31. Effect of bedding type on GP diagnosed wheeze p = .22.

^*Reference.

For the frequency of GP visits for wheezing, there is a two‐fold significant increase in the rate amongst those children sleeping in feather sleepsacks compared to synthetic sleepsacks, Table 3.

TABLE 3.

Frequency of visits for GP reported wheezing by bedding material.

Frequency of GP visits for wheeze	Synthetic	Feather	Wool	Total
0	76	67	87	230
1	13	13	14	40
2	9	8	6	23
3+	8	16	7	31
Total	106	104	114	324
Relative rate (95% CI)	1 (reference)	2.00* (1.14, 3.52)	0.91 (0.48, 1.76)

Note: Effect of bedding materials on rates.

^* p = .01.

We have compared the whole study population to those who self‐reported greater compliance with using the sleepsacks Table 4. There appear to be no differences between the whole group and the more compliant group; formal comparisons have not been made.

TABLE 4.

Bedding material by parental reported wheezing ever, intention to treat versus most compliant families.

Bedding type	Parent reported wheeze			Parent reported wheeze
	All subjects			*Most compliant
	Yes	Total	%	Yes	Total	%
Synthetic	51	136	37.5	16	47	34.0
Feather	59	136	43.4	25	58	43.1
Wool	48	139	34.5	29	79	36.7
Total	158	411	38.4	70	184	38.0

Note: Compliance defined as regular use ≥5 nights per week and for ≥1 year. (no formal comparisons were made)

We have recorded various GP‐recorded health parameters by bedding material, including any inhaled asthma treatments, otitis media, croup and antibiotic prescriptions, with no differences between material types, Table 5. There was no difference in the frequency of atopy at 2 years between bedding materials, and overall, the rate of atopy at 2 years was 8% (19 children) Table 6. There was no difference between bedding materials for parentally reported eczema. Overall, 40% of parents reported eczema at some time in the first 2 years. For children in synthetic bedding, 43.8%; Feather, 37.5%; Wool, 38.4% (n = 410).

TABLE 5.

Various GP reported health parameters.

Bedding type	GP any prescribed inhaled asthma Rx	GP any reported otitis media	GP any reported croup	GP any antibiotic prescribed
	% (N = 324)
Synthetic	20.8	49.1	13.2	61.3
Feather	25.0	43.3	14.4	63.5
Wool	20.2	45.6	15.8	68.4
Total	21.9	46.0	14.5	64.5
	p = .28	p = .33	p = .89	p = .16

TABLE 6.

Frequency of atopy at 2 years by bedding material.

Bedding type	Atopic	Non atopic	Total	Percent
Synthetic	5	70	75	6.7
Feather	6	78	84	7.1
Wool	8	70	78	10.3
Total	19	218	237	8.0

Note: Effect of bedding materials on atopy at 2 years p = .67.

4. DISCUSSION

In this study, we find no evidence of a protective effect of feather bedding on early childhood wheezing, or any evidence that synthetic bedding materials were associated with increased wheezing (compared to wool or feather bedding) as has been suggested in various observational studies. On the contrary, there is a suggestion that wheezing may be more common and severe amongst those children sleeping in feather bedding compared to synthetic materials or wool. Observational studies have associated children's bedding materials with wheezing, both increasing risk from synthetic materials and increased protection from feather materials, and as mentioned, we have reviewed the literature in some detail previously. ⁹

Using data from a large birth cohort, Nafstad et al. explored bedding/asthma associations. ⁶ They found a 40% reduction in the risk of bronchial obstruction by 2 years and a 60% reduction in the risk for asthma at 4 years for children using a feather quilt at age 6 months. They controlled for environmental, socioeconomic, and disease propensity factors using multiple regression and then stratified the analysis based on these factors, but the feather bedding associations remained significant. However, the authors acknowledged that their control for these factors may not fully eliminate pre‐birth selection bias. ⁶ Ponsonby et al. have used a cross‐sectional study from a birth cohort to explore bedding and wheezing and also showed a protective effect of feather bedding. ¹⁴ The same authors have also shown an increased risk of wheezing at age 7 with synthetic bedding materials either used in infancy or currently at 7 years. They did not consider that the risks were associated with original bedding choices. ⁷

In 2011, a randomised control trial (RCT) of feather bedding (feather pillow and quilt versus standard washing advice) amongst children with asthma was reported by Glasgow et al. ¹⁵ The primary endpoints were the proportion of children reporting four or more episodes of wheezing, or an episode of speech‐limiting wheeze, or one or more episodes of sleep disturbed by wheezing in a week. There were no significant differences in wheeze frequency or severity at 12 months by bedding details. Despite this finding, there is overwhelming evidence that house mite allergen exposure is strongly associated with asthma and effective allergen avoidance can result in clinical asthma improvement. ¹⁶ Studies of severe asthma in childhood have also shown reduced severity when mite impermeable covers are added to bedding. ¹⁷

A variety of explanations have been put forward relating to synthetic materials producing VOCs and or isocyanates ¹ both sensitisers of the airway and associated with asthma. ¹⁸ Reduced wheezing associated with feather bedding might, as we have previously suggested, be related to reduced house dust mite contamination of feather bedding because feather bedding is usually encased in a closer woven fabric. ³ However, feather materials have often been regarded by the public as causing asthma ⁶ rather than preventing it, and the associations between house dust mites and feathers have long been recognised. ¹⁹ Thus, the inverse associations between wheezing and feather exposure found in observational studies are somewhat surprising. But here lies the potential explanation of selection bias as has been discussed in many of the studies. This would arise if parents consciously or unconsciously avoided feather bedding material because of a belief in the possibility that it might cause allergy and asthma in their children, already at higher risk of developing asthma. In a number of the observational studies, attempts have been made to reduce this by excluding any children whose families had altered their bedding preferences based on knowledge of possible increased risks from feather materials. Authors often conceded that their questions may not be capturing this avoidance completely. Our study, which found little effect of bedding on early wheezing, would suggest that selection bias might be the explanation for a protective effect of feather bedding on early wheezing. It is notable that the only other RCT of feather bedding, though of established asthma, also found no effect of bedding on wheezing episodes. ¹⁵

Our study found no effect of bedding on parentally reported eczema or on atopy as measured by skin prick test at 2 years. Interestingly, the rate of atopy, just 8% in this population at 2 years, is considerably lower than we would expect at this age. In a New Zealand population‐based birth cohort study, the rate of atopy by SPT at 15 months was 27%. ²⁰ This raises the possibility that having new bedding materials reduces allergen exposures in early childhood and hence sensitisation. We have collected dust from the bedding materials from 60 children with a view to analysing the dust for HDM allergens if we had found significant differences in wheezing by bedding materials. We hope to follow these children in the future to observe possible differences in wheezing and asthma in later childhood by their early bedding experience. If we found differences, we would analyse their bedding for HDM allergens.

One of the limitations of our study is that we have studied newborn children and followed their respiratory history for just 2 years. Such early wheezing is often associated with viral infections. However, as has been shown, early viral childhood wheezing is strongly associated with later asthma. ²¹ As mentioned, we hope to follow up these children in older childhood by survey or from prescription databases. Another limitation may be that we used sleepsacks instead of traditional bedding, and this may be associated with less inhalation of material or allergens from bedding. Lastly, while we asked parents about their use of the bedding, as a measure of compliance, we had no objective measure of compliance. However, we found no evidence that increased compliance by parental report had any effect on outcomes.

In this randomised trial of three bedding materials, we find no support either for synthetic materials causing more wheezing or feather materials causing less, by 2 years of age. Our findings would support the hypothesis that the positive associations between synthetic and the negative associations with feather bedding in observational studies may be the result of uncontrolled selection bias, whereby parents of children predisposed to allergic disease avoid feather bedding.

AUTHOR CONTRIBUTIONS

Caroline Halley: Conceptualization; methodology; software; data curation; investigation; validation; formal analysis; supervision; visualization; project administration; resources; writing – original draft; writing – review and editing. Janice Kang: Software; data curation; investigation; formal analysis; supervision; project administration; resources; writing – original draft; writing – review and editing. Phillipa Barnes: Data curation; software; writing – review and editing. Michael Keall: Formal analysis; writing – review and editing; writing – original draft. Robert Siebers: Conceptualization; methodology; writing – review and editing. Cheryl Davies: Writing – review and editing; data curation; investigation; supervision. Philippa Howden‐Chapman: Writing – review and editing; funding acquisition; conceptualization; methodology. Julian Crane: Writing – review and editing; writing – original draft; conceptualization; methodology; supervision; visualization; formal analysis; validation.

FUNDING INFORMATION

The study was supported by The Prime Minister's Science Prize and the Health Research Council of New Zealand.

CONFLICT OF INTEREST STATEMENT

None of the authors have any conflicts of interest to declare with regard to this manuscript.

Halley C, Kang J, Barnes P, et al. Bedding materials and early infant wheezing: A randomised controlled trial. Pediatr Allergy Immunol. 2025;36:e70073. doi: 10.1111/pai.70073

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy and ethical restrictions.