BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 28, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

Study protocol: “Nutritional impact on Immunological maturation during Childhood in relation to the Environment (NICE)”, a prospective birth-cohort in northern

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2018-022013

Article Type: Protocol

Date Submitted by the Author: 29-Jan-2018

Complete List of Authors: Barman, Malin; Chalmers University of Technology, Department of Biology and Biological Engineering; Sahlgrenska universitetssjukhuset, Department of Obstetrics and Gynecology Murray, Fiona; Region Norrbotten, Sunderby Research Unit Bernardi, Angelina; University of , Institute of Biomedicin, Department of Infectious Diseases Broberg, Karin; Karolinska Institutet, Institute of Environmental Medicine Bölte, Sven; Karolinska Institutet, Department of Women’s and Children’s Health Hesselmar, Bill; , Institute of Clinical Sciences, Department of Paediatrics Jacobsson, Bo; University of Gothenburg, Institute of Clinical Sciences, Department of Obstetrics and Gynecology Jonsson, Karin; Chalmers University of Technology, Department of Biology http://bmjopen.bmj.com/ and Biological Engineering Kippler, Maria; Karolinska Institutet, Institute of Environmental Medicine Rabe, Hardis; Gothenburg University, Institute of Biomedicine, Department of Infectious Diseases Ross, Alastair; Chalmers University of Technology, Department of Biology and Biological Engineering Sjöberg, Fei; University of Gothenburg, Institute of Biomedicine, Department of Infectious Diseases

Strömberg, Nicklas; Umeå University, Department of Odontology/Cariology on September 28, 2021 by guest. Protected copyright. Vahter, Marie; Karolinska Institutet, Institute of Environmental Medicine Wold, Agnes; University of Gothenburg, Institute of Biomedicine, Department of Infectious Diseases Sandberg, Ann-Sofie; Chalmers University of Technology, Department of Biology and Biological Engineering Sandin, Anna; Umeå University, Department of Clinical Sciences

birth cohort, Allergy < THORACIC MEDICINE, Nutrition < TROPICAL Keywords: MEDICINE, Microbiology < PATHOLOGY, toxicants, Immunology < BASIC SCIENCES

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 1 Study protocol: “Nutritional impact on Immunological maturation 4 2 5 during Childhood in relation to the Environment (NICE)”, a 6 3 prospective birthcohort in northern Sweden 7 4 8 5 Barman, M.1+2, Murray, F.3, Bernardi, AI.4, Broberg, K.5, Bölte, S.,6 Hesselmar, B.7, 9 6 Jacobsson, B.8, Jonsson, K.1, Kippler, M.5, Rabe, H.4 Ross, A.1, Sjöberg, F.4, Strömberg, N.9, 10 5 4 1 3+10 11 7 Vahter, M. , Wold, A. E. , Sandberg, AS. , and Sandin, A.

12 8 1 13 9 Food and Nutrition Science, Department of Biology and Biological engineering, Chalmers 14 10 University of Technology, Göteborg, Sweden. 2 15 11 Department of Obstetrics and Gynecology, Sahlgrenska University Hospital, Gothenburg, 16 12 Sweden. For peer review only 17 13 3Sunderby Research Unit, Region Norrbotten, Sweden. 18 14 4Department of Infectious Diseases, Institute of Biomedicine, the Sahlgrenska Academy, 19 15 University of Gothenburg, Göteborg, Sweden. 20 16 5Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden. 21 17 6Center of Neurodevelopmental Disorders (KIND), Division of Neuropsychiatry, Department 22 18 of Women’s and Children’s Health, Karolinska Institutet, Stockholm, Sweden. 23 7 24 19 Department of Paediatrics, Institute of Clinical Sciences, University of Gothenburg, 20 Göteborg, Sweden. 25 8 26 21 Department of Obstetrics and Gynecology, Institute of Clinical Sciences, the Sahlgrenska 27 22 Academy, University of Gothenburg, Göteborg, Sweden. 9 28 23 Department of Odontology/Cariology, Umeå University, Umeå, Sweden. 29 24 10Department of Clinical Sciences, Unit of Pediatrics, Umeå University, Umeå, Sweden. 30 25 31 26 Corresponding author: Malin Barman, Chalmers University of Technology, Department of 32 27 Biology and Biological Engineering, Food and Nutrition Science, SE412 96 Gothenburg, 33 28 Sweden. Telephone: +46(0)317723811; Fax: +46(0)317723830; e mail: http://bmjopen.bmj.com/ 34 29 [email protected] 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 30 Abstract 4 5 31 Introduction 6 32 Pre and neonatal environmental factors, such as nutrition, microbes and toxicants, may affect 7 33 health throughout life. Many diseases, such as allergy, and impaired child development may 8 34 be programmed already in utero or during early infancy. Birthcohorts are important tools to 9 35 study associations between early life exposure and disease risk. Here, we describe the study 10 11 36 protocol of the prospective birthcohort, “Nutritional impact on Immunological maturation 12 37 during Childhood in relation to the Environment” (NICE). The primary aim of the NICE 13 38 cohort is to clarify the effect of key environmental exposures – diet, microbes and 14 39 environmental toxicants during pregnancy and early childhood, on the maturation of the 15 40 infant’s , including initiation of sensitisation and allergy. As the studied 16 41 exposures may alsoFor affect peerother outcomes, review the NICEcohort only will investigate some secondary 17 42 outcomes: infant growth, obesity, neurological development and oral health. Banking of 18 43 samples will enable additional investigations into future biomarkers. 19 20 44 Methods and analysis 21 45 The NICEcohort will recruit about 650 families during midpregnancy. The principal 22 46 inclusion criterion is planned birth at the Sunderby Hospital in the northern region of Sweden, 23 47 during 2015 – 2018. Questionnaires and biological samples are collected at ten timepoints, 24 48 from pregnancy until the children reach four yearsofage. Samples are collected primarily 25 26 49 from mothers and children, but also from fathers. Biological samples include blood, urine, 27 50 placenta, breast milk, meconium, faeces, saliva and hair. Information regarding allergic 28 51 heredity, diet, socioeconomic status, lifestyle including smoking, siblings, pet ownership etc. 29 52 is collected using questionnaires. Sensitization to common allergens is assessed by skinprick 30 53 testing and allergic disease is diagnosed by a paediatrician at one and four years of age. At 31 54 four yearsofage the children are also examined regarding growth, neurobehavioral and 32 55 neurophysiological status and oral health. 33 http://bmjopen.bmj.com/ 34 56 Ethics and dissemination 35 57 The NICEcohort study protocol has been approved by the Regional Ethical Review Board in 36 58 Umeå, Sweden (2013/1831M). 37 38 39 59 Key words (6 maximum) 40 60 Birthcohort, allergy, immunology, nutrition, microbiology, toxicants. 41 on September 28, 2021 by guest. Protected copyright. 42 43 61 Strengths and limitations of this study 44 62 • Prospective study design covering a period from midpregnancy to the age of four 45 63 years with biological samples from mother, father and child for evaluation of 46 64 microbiology, nutrition, immunology, environmental toxicants, genetics and 47 48 65 epigenetics. 49 66 • Advanced analytical methods, including flow cytometry on fresh blood samples, 50 67 multiomics methodologies and advanced exposure assessment methodologies. 51 68 • Interdiciplinary, translational approach with expertise in allergology, immunology, 52 69 nutrition, microbiology, toxicology, obstetrics, childhood growth, neurophysiological 53 70 development and oral health. 54 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 71 • The primary outcome, allergic disease (food allergy, atopic eczema, asthma and 4 72 allergic rhinoconjunctivitis) is diagnosed by a paediatric allergology specialist 5 73 according to strictlydefined protocols. 6 74 • Although the study sample will allow in depth analyses on multiple exposures, genetic 7 75 factors and outcomes, both frequences of potential factors and etiological 8 76 heterogeneity of outcomes may require follow upstudies. 9 10 11 77 Introduction 12 78 The foetal and neonatal periods could be regarded as a window of opportunity for influencing 13 79 health during childhood and throughout life. Environmental exposures, such as nutrition, 14 80 microbiota and toxicants, and genetic or epigenetic factors may synergise to impact growth, 15 81 neurodevelopment and immune maturation. However, which factors are most important and 16 For peer review only 17 82 the mechanisms by which they operate are still poorly understood. The Nutritional impact on 18 83 Immunological maturation during Childhood in relation to the Environment (NICE) cohort 19 84 focuses primarily on early life exposures, immune maturation and risk of immunemediated 20 85 diseases. 21 86 22 87 Allergy, dental caries and overweight are the most common chronic conditions in childhood, 23 88 affecting the health and wellbeing of around 25%, 20% and 13% of children respectively [1]. 24 89 Chronic diseases place a substantial pressure on both the individual and on the health care 25 90 system. Hence, it is important to explore how lifestyle, environmental and genetic/epigenetic 26 91 factors during pregnancy and the first years of life protect against or facilitate the 27 92 development of chronic diseases. 28 29 93 30 94 Allergy is characterized by dysregulation of the immune system leading to an incapacity to 31 95 develop tolerance to normally harmless proteins in food and air. Decreased exposure to 32 96 microorganisms and reduced complexity of the in early life have been 33 97 associated with an increased risk for the development of allergy [2, 3]. Also, the diet of http://bmjopen.bmj.com/ 34 98 pregnant and lactating women [4] and their exposure to environmental toxicants [5, 6] have 35 99 been suggested to play a role in immune maturation and allergy development. In spite of 36 100 extensive research in this area, the specific causes and pathways leading to immune 37 101 maturation, and their subsequent protective or detrimental influences on allergy development, 38 102 remain largely unknown 39 103 40

104 In order to identify the key factors leading to the development of healthy immune regulation on September 28, 2021 by guest. Protected copyright. 41 105 or allergic disease, it is necessary to study immune events at a very early age [7], and account 42 43 106 for external factors that could influence immune development. Allergy may present as early 44 107 as a few months of age, usually manifested as atopic eczema or food allergy, while asthma 45 108 and rhinoconjunctivitis generally appear at school age. A major source of imprecision in 46 109 many studies on allergy development is the diagnostic methodology. While questionnaires are 47 110 convenient and costeffective, the diagnosis of childhood allergy by a skilled clinician, aided 48 111 by tests of sensitization, is the gold standard. For food allergy, a careful investigation of 49 112 medical history by an experienced clinician, aided by eliminationprovocation is the only 50 113 method to adequately identify this type of allergic manifestation. 51 114 52 115 Multiple environmental factors, including diet and exposure to toxicants, may affect both the 53 116 immune system and neurophysiological development [811]. Nutritional factors considered to 54 117 be immunomodulatory are unsaturated fatty acids, vitamins A and D, iron and zinc. Some 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 118 nutrients as well as some metals pass the placenta to the foetus. Some metals accumulate in 4 119 the placenta, thus elemental metals may cause both impaired placental function and foetal 5 120 toxicity. Even lowlevel exposure to certain metals, e.g. cadmium (Cd), arsenic (As), 6 121 manganese (Mn) and lead (Pb), has been found to impair foetal and childhood immune 7 122 function, growth and neurodevelopment [1215]. 8 123 9 124 Overweight and obesity are increasing worldwide. Overweight at a young age is an important 10 125 risk factor for obesity later in life. In Sweden obesity has increased during the last 40 years 11 126 from 2,3 to 4,7% amongst girls and from 2,6 to 8,5% amongst boys [16]. Socioeconomic 12 127 status and physical activity are important determinants, but there are other factors to explore 13 14 128 such as the impact of different dietary components (e.g. fatty acids) and environmental 15 129 toxicants, metabolic processes and epigenetic factors, to name just a few. 16 130 For peer review only 17 131 In Western countries, one fifth of children are socalled highcaries cases (children with a 18 132 high burden of dental caries) that do not respond to traditional prevention [17]. These high 19 133 caries cases derive from an immune deficiency or carriage of high virulence strains of 20 134 Streptococcus mutans, consistent with their nonresponder behaviour [18, 19]. By contrast, 21 135 the lowtomoderate risk children develop caries as expected from eating and oral hygiene 22 136 habits [18]. Therefor the NICE study aims to evaluate the oral health of all participating 23 137 children in order to better understand the link between impaired immune maturation and the 24 138 risk for caries development. 25 26 27 139 Aims and objectives 28 29 30 140 Primary aim 31 32 141 The primary aim of the NICE birthcohort is to evaluate how multiple environmental 33 142 exposures such as lifestyle, diet, microbes and toxicants influence maturation of the immune http://bmjopen.bmj.com/ 34 143 system and affect allergy development. 35 36 144 Secondary aims 37 38 145 The secondary aim of NICE is to examine if these same environmental exposures influence 39 146 pregnancy outcomes, childhood growth, obesity, neuropsychological development, and caries 40 147 development. 41 on September 28, 2021 by guest. Protected copyright. 42 43 148 Research questions 44 149 The NICE cohort has been designed to answer the following research questions: 45 150 i. Which nutrients affect the maturation of the infant’s immune system, allergy 46 151 development, oral health, neurophysiological development, growth and obesity? 47 152 ii. Can a metabolic fingerprint be identified that predicts diseaserisk during childhood? 48 49 153 iii. What role do the oral and gut microbiota play in immunological maturation, allergy 50 154 development, oral health, neurophysiological development, growth and obesity? 51 155 iv. What is the impact of early life exposure to toxicants, particularly toxic metals, on 52 156 immune maturation, allergy development, caries diseases, neurophysiological 53 157 development, growth and obesity? 54 158 v. Does maturation of the immune system during foetal life and infancy affect the risk of 55 159 allergy development, oral health or neurophysiological development? 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 160 vi. Can certain immune signatures be identified that are associated with immune 4 161 dysregulation, caries development or, conversely, with health? 5 162 vii. Is early life exposure to toxic metals and essential elements reflected in the dental 6 163 enamel? 7 164 viii. Which nutritional, environmental and lifestyle profiles are protective against caries in 8 165 the high versus the lowtomoderate caries risk groups and types? 9 166 ix. What is the impact of nutrition and toxic exposures on the epigenetic programming of 10 167 children in relation to immune maturation, allergy development, oral health, 11 168 neurophysiological development, growth and obesity? 12 169 x. What are the effects of perinatal exposures (diet, lifestyle and environmental 13 14 170 toxicants) on pregnancy outcomes such as gestational diabetes, preeclampsia, 15 171 pregnancy length, intraeuterine growth and birth weight? 16 For peer review only 17 172 Methods and Analysis 18 19 20 173 Study design and setting 21 22 174 The study design is a prospective longitudinal birthcohort study. All expectant parents 23 175 planning to give birth at the Sunderby Hospital in Luleå, during 2015 – 2018, are invited to 24 176 participate. The Sunderby Hospital, with approximately 2,000 deliveries per year, is located in 25 177 Region Norrbotten in the very north of Sweden makes up almost a quarter of Sweden 26 178 geographically, but has a population of only approximately 250,000 inhabitants. Expectant 27 179 mothers are cared for at 22 maternity clinics across the region. Routine ultrasound is 28 180 performed at gestational week 1819 at either of three sites: at the Sunderby, Kalix or Piteå 29 181 hospitals. These three hospitals have combined catchment area of approximately 200,000 30 182 inhabitants. 31 32 33 183 Recruitment and inclusion and exclusion criteria http://bmjopen.bmj.com/ 34 35 184 All expectant parents living in the catchment area of the Sunderby Hospital receive a concise 36 185 information leaflet about the study at their first visit to their local maternity clinic in gestation 37 186 week 1012. Recruitment takes place in connection with the routine ultrasound at gestational 38 187 week 1819. All expectant parents are given detailed information about the study. The parents 39 188 who wish to participate are requested to provide informed consent. They subsequently 40 189 received an acknowledgement letter and are assigned a study ID number. 41 190 on September 28, 2021 by guest. Protected copyright. 42 191 The criteria for participation in the study are as follows: the intention to give birth at the 43 192 Sunderby Hospital, the ability to communicate in written and spoken Swedish and current 44 193 habitation in Region Norrbotten. Families are excluded from the study if they move from 45 194 Norrbotten before the child is born. Families who move outside of Norrbotten after birth are 46 47 195 allowed to continue participation in the study if so wished and followup appointments are 48 196 arranged where possible. 49 50 197 Estimated sample size 51 52 198 The NICEcohort is based on the conservative inclusion of children to be delivered at the 53 199 Sunderby Hospital with runin during the period 20150115 – 20180228 (Table 1). 54 200 Assuming equal samples size in the exposed and unexposed groups and two sided confidence 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 201 level of 95 %, the planned sample size of 650 families will enable us to detect a difference in 4 202 allergy rate of 10 % in unexposed subjects and 18 % in exposed subjects by a power of 80%. 5 6 203 Study outcomes 7 8 204 Primary outcome 9 10 205 The primary outcome in NICE is immune maturation in the neonatal period and its effect on 11 206 development of sensitization and allergy over a followup period of four years. 12 13 207 Secondary outcomes 14 208 The secondary outcomes are: 15 209 pregnancy outcomes (preeclampsia, gestational diabetes, gestational week at birth, 16 210 intraeuterineFor growth peer and birth weight). review only 17 211 child growth (including overweight and obesity). 18 212 neurophysiological development (test of motor, cognitive and behavioural function). 19 213 caries and oral health 20 21 22 214 Biological sample collection 23 215 Mother, fathers and infants participating in the NICEcohort are repeatedly followedup from 24 216 pregnancy until 4 years of age with both questionnaires and biological sample collection 25 217 (Figure 1). The flow chart of the study is shown in figure 1 and a detailed overview of the 26 27 218 collection of biological samples and questionnaires are shown in tables 2 and 3. 28 29 219 Subcohort 30 220 Approximately one in four recruited families are offered the possibility to participate in a 31 221 subcohort with more extensive sampling of mother and infant (see Table 2). The assignment 32 222 of families to the subcohort is performed at delivery, based on the feasibility of extended 33 223 sampling at that timepoint http://bmjopen.bmj.com/ 34 35 224 Blood samples 36 225 Blood samples collected from the parents include: 10 mL/5 mL SST II Advance serum tube, 37 226 10mL/3mL EDTA tube (all fromBecton Dickinson, Plymouth, UK), 6 mL Naheparin trace 38 227 elementfree tube (Greiner bioone, Kremsmünster, Austria), and 10 mL RNAstabilised 39 228 blood collection Tempus™ tubes (Applied Biosystems, Foster City, CA), the latter from 40 229 mothers only. The mothers are asked to be fasting for 8 h before study visits. on September 28, 2021 by guest. Protected copyright. 41 230 42 231 At birth, venous and arterial cord blood are collected from the umbilical cord into small 500 43 44 232 µL microtainer EDTA tubes (Becton Dickinson) using a syringe. Mixed umbilical cord blood 45 233 is collected into another 3 mL and 6 mL EDTA tube (the 3 mL tube is stored at room 46 234 temperature for subsequent flow cytrometric analysis) and a 6 mL Naheparin trace element 47 235 free tube (Greiner bioone). Midwives follow WHO guidelines with regards to umbilical cord 48 236 clamping, that is, cord clamping is carried out more than one minute after the birth or when 49 237 cord pulsation has ceased. 50 238 51 239 Venous blood is collected from the infants by the study nurses into 3 mL EDTA tubes 52 240 (Becton Dickinson) using a Microneedle (Sarstedt, Numbrecht, Germany) or Eclipse Signal 53 241 Blood Collection needle (Becton Dickinson). Blood is drawn from the dorsal side of the hand 54 242 at 48 h, 1 month, 4 months and from the hand or arm at 12 months and 48 months of age. 55 56 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 243 Urine samples 4 244 Infant’s urine at 4 months of age is collected in a urine collection bag placed around the penis 5 245 or vagina under the infant’s nappy at the start of a study visit, with the bag being collected at 6 246 the end of the study visit. Urine samples at 4 yearsofage are collected in a potty lined with a 7 247 plastic bag. Midstream urine samples, taken at a random time point during the day, are 8 248 collected from adults. 9 10 249 Placenta 11 250 After delivery, the placenta is examined according to local guidelines by the delivery personal 12 251 and placed in a plastic bag and stored at 4°C. The placenta is collected by the study nurses 13 252 and processed at the laboratory according to the outline in figure 2. 14 15 253 Breast milk 16 254 Mothers are askedFor to express peer 30 mL review of breast milk in only the morning before the second 17 255 breastfeeding occasion of the day, at 1 and 4 months postpartum. Samples are collected in a 18 19 256 clean bowl or pump and poured into plastic centrifuge tubes for immediate storage at 20°C at 20 257 home. Mothers are requested to bring the frozen tubes with them at the 4 months followup 21 258 and reminded to keep the samples in a cooling bag or wrapped in paper with a cooling block 22 259 to prevent thawing of the samples during transport to the hospital. The samples are transferred 23 260 directly to a 80°C freezer on arrival to the research laboratory. If thawing appears to have 24 261 occurred during transport, this is noted in the sample preparation protocol. 25 26 262 Meconium 27 263 Meconium is sampled when voided using a sterile spatula and put in a sterile container that is 28 264 immediately stored at 4°C until aliquoting at the research laboratory. 29 30 265 Faeces 31 266 Faecal samples are collected from the infant’s nappy by the parents at home using a sterile 32 267 spatula. Faeces is put in sterile tubes prelabelled with the child’s study ID. Parents are http://bmjopen.bmj.com/ 33 268 requested to write the date of sampling on the tubes and then place the samples in the freezer 34 269 as soon as possible and to bring it to the hospital at the next followup. Participants are 35 36 270 requested to keep the samples in a cooling bag or wrapped in newspaper to prevent thawing of 37 271 the samples during transport to the hospital. The stool samples are transferred directly to a 38 272 80°C freezer on arrival to the research laboratory. If thawing appears to have occurred during 39 273 transport, this is noted in the sample preparation protocol. 40 41 274 Saliva on September 28, 2021 by guest. Protected copyright. 42 275 Approximately 1.52 mL saliva is collected from mothers, by spitting into a plastic tube. 43 276 Infant saliva (approximately 0.5 mL) is collected from the buccal cavity using a plastic 44 277 Pasteur pipette. 45 46 278 Buccal swabs 47 279 Buccal swabs (eSwab, Copan, Bresia, Italy) are collected by swiping the swab tip pass 48 280 beneath the upper lip, and the inside of the cheeks, taking care to avoid touching the teeth. 49 281 The swabs are stored at 4°C until centrifugation at 12,000 g for 10 min. Excess fluid is 50 282 removed from the bacterial pellet before storage. 51 52 283 Hair specimens 53 284 At 48 months hair specimens are cut by stainless steel scissors from the occipital part of the 54 285 head, beneath some covering hair, as close to the scalp as possible. 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 286 Handling, transport and storage of biological samples 4 287 At gestational week 28, samples are collected at the maternity clinics. The RNA Tempus™ 5 6 288 blood samples and urine samples are stored immediately at 4°C, whilst blood samples are 7 289 kept at room temperature for 30 min before centrifugation. All samples are stored at 4°C until 8 290 transport and are subsequently transported at 410°C to the hospital laboratory, either the 9 291 same day or the following workday. 10 292 11 293 Samples taken at delivery and at 48 h are stored at 4°C and processed within 212 h if 12 294 sampled on weekdays and within 1265 h if sampled on weekends. Samples taken at the 13 295 followup visits at 1, 4, 12 and 48 months of age are collected by the NICEcohort study 14 296 nurses and handled the same work day. All biological samples stand for 1530 min in room 15 297 temperature until centrifugation of the blood. After centrifugation of the blood, all samples 16 298 are stored at 4°CFor until aliquoting peer (performed review on a cooling block) only and interim storage at 20°C 17 299 (for a maximum of 12 weeks). Once filled, storage boxes are transferred to longterm storage 18 19 300 at 80°C. Hair samples are collected in paper envelopes and stored at room temperature. A 20 301 laboratory protocol is filled in for all samples where all transport and preparation times are 21 302 noted in order to be able to account for the time from sampling to freezer in all downstream 22 303 analyses. 23 24 304 Collection of participant data 25 26 305 All study participants will be closely followedup during pregnancy from inclusion until the 27 306 delivery from physician’s notes in their medical records. Data from the medical pregnancy 28 307 records will be analysed and collected in the database. 29 308 30 309 At inclusion, both parents are asked to fill out one online questionnaire each regarding 31 310 allergic heredity and environment. When the infant has reached 18 months of age participants 32

311 are sent the second online questionnaire regarding the infant’s current living environment and http://bmjopen.bmj.com/ 33 34 312 lifestyle. The third online questionnaire is sent out to the participants when the infant has 35 313 reached 48 months of age and concerns the infant’s environment and lifestyle (Table 3). 36 37 314 Assessment of dietary intake in mothers and children 38 315 The mothers receive a webbased, food frequency questionnaire (MealQ) by email during 39 316 pregnancy and at 1 and 4 months postpartum, consisting of questions regarding their diet 40 317 during the previous month [20, 21] (Table 4). Reminders are sent out by email one and two 41 318 weeks after the initial email and, alongside the last reminder, a text message is sent to their on September 28, 2021 by guest. Protected copyright. 42 319 cell phones. The questionnaires have a mealbased and interactive format, i.e. only those who 43 320 report a relevant frequency of consumption of a certain food or food group will receive 44 321 followup questions, to promote ease of use and minimize the time required to answer the 45 322 questionnaire. Depending on the number of followup questions, the MealQ questionnaire 46 323 contains 102174 questions. 47 324 48 325 When the child reaches 12 months and 48 months of age, the parents fill out a food frequency 49 50 326 questionnaire about the infant’s food habits (TodMealQ and KidMealQ, respectively), 51 327 which is similar to the MealQ but adopted for young children [22]. Parents from families 52 328 where both parents have attended the followup visits will be asked to participate in a 53 329 subcohort with extended sampling, including a MealQ questionnaire for each of the parents 54 330 at 12 months of age of the child. 55 331 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 332 To measure total energy expenditure, all dietary questionnaires are combined with the activity 4 333 questionnaires ActiveQ [23]or KidActiveQ [24]. 5 334 6 335 The questionnaires MealQ and KidMealQ have been validated in their original form [2022] 7 336 but all dietary questionnaires were modified to better fit the specific needs of the study of 8 337 evaluation of fat quality, consumption of fish from polluted areas, sugar and probiotic use. 9 338 Modifications are detailed in Supplemental file 1. Nutritional composition of the diet is 10 339 calculated from the dietary questionnaires using the database of nutrient content published by 11 340 the Swedish National Food Agency. The nutrient conversion for the questionnaires will be 12 341 calculated by computer programs developed specifically for those questionnaires. 13 14 342 Diary for recoding introduction of complementary foods to the infant diet: 15 343 Practices of introduction of complementary foods and formulas to the infant’s diet are 16 For peer review only 17 344 assessed each month during the first 12 months of life by a webbased questionnaire. The 18 345 questions include the infant’s intake of breast milk from breast or bottle, and whether certain 19 346 foods have been introduced, including formula (percentages of total feeding and brand name), 20 347 gruel (regular, with/without gluten or lactose, soya based), porridge (regular, without gluten 21 348 or lactose) and different food groups (potatoes/roots, fruit/juice, berries, nuts/almonds, 22 349 peanuts/peanut oil, bread/crackers/cookies, butter, margarine, oil, milk, icecream, sour milk, 23 350 yoghurt, meat, fish and egg), as well as if there are any food the child has started to eat more 24 351 than twice per week that is not included above. Additional questions are asked about 25 352 probiotics, vitamin D, omega3 or other supplements and frequency of canned versus home 26 353 made food. 27 354 28 29 355 Parents are asked to estimate the proportion of commercial versus homemade baby food. In 30 356 order to be able to adjust the statistical analyses for reverse causation, the parents are also 31 357 asked the questions: ‘during the past month, have you actively avoided to give any type of 32 358 food due to concern of allergy?’ and ‘if so what food items have been excluded from the 33 359 child’s diet?’. http://bmjopen.bmj.com/ 34 35 360 Outcome measures 36 37 361 Infections during year one 38 39 362 Infectious episodes are recorded monthly by the parents during the first 12 months using the 40 363 abovementioned webbased questionnaire. If an infectious episode is recorded, followup 41 364 questions are posed concerning duration of the illness, body temperature, breathing problems, on September 28, 2021 by guest. Protected copyright. 42 365 antibiotic treatment, medications, hospitalisation etc. 43 44 366 Lung- function test 45 367 Lungfunction is assessed at 4 and 12 months, using respiratory resistance with interrupter 46 368 technique (MicroRint, MicroMedical Ltd, UK). The method estimates alveolar pressure 47 369 (Palv) by performing a rapid occlusion of the airways during the respiratory circle at rest, 48 370 assuming that Palv will equal mouth pressure (Pmo). Respiratory resistance with interrupter 49 371 technique (RInt) is a noninvasive method that requires minimal cooperation and can be used 50 372 in evaluating airway obstruction in uncooperative and small children [25, 26]. At 48 months 51 373 lung function will be measured by standardized Impulse Oscillometry. 52 53 54 55 56 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 374 Skin prick test 4 375 Sensitization to common allergens is assessed by skin prick testing at 12 and 48 monthsof 5 376 age. Testing is performed on the volar side of the lower arm and is carried out according to 6 377 the standards of the Subcommittee on Skin Tests of the European Academy of Allergy and 7 378 Clinical Immunology [27]. Six different standardized allergens (egg, milk, birch, timothy, cat 8 379 and dog) are applied with a lancet; histamine dihydrocloride (10 mg/mL) is used as positive 9 380 control (Soluoprick, ALK Nordic A/S, Denmark). The test is considered positive for a 10 381 particular allergen if the mean wheal diameter is at least 3 mm after 15 min. 11 12 382 Allergy diagnosis 13 383 Allergy is diagnosed at 12 and 48 months by a specialist in paediatric allergology. The 14 384 diagnosis is based on a clinical history and a clinical examination using the same 15 385 methodology as in the BAS cohort [28], enabling direct comparisons. The clinical protocol is 16 For peer review only 17 386 carefully standardized and complies with the diagnostic protocols of our previous 18 387 ALLERGYFLORA study [29] and FARMFLORA [4] study. 19 388 20 389 The following diagnostic criteria are used for clinical allergic diagnosis at 12 and 48 months 21 390 of age: 22 23 391 Atopic eczema, atopic dermatitis (AD): 24 392 Various diagnostic criteria for AD have been proposed and validated. For the NICE cohort 25 393 followup at 12 and 48 months we choose to employ the simplified criteria proposed by 26 394 Williams et al. with high sensitivity and specificity [3032]. 27 395  major criteria (=must have): an itchy condition or parental report of 28 396 scratching/rubbing by the child 29 397  plus three or more of the following criteria: 1) personal history of itchiness in skin 30 398 creases such as cheeks, folds of the elbows, behind knees, front of ankles or around 31 32 399 the neck 2) personal history of asthma or hayfever or other atopic disease or a history 33 400 of atopic disease among a firstdegree relatives 3) a personal history of generally dry http://bmjopen.bmj.com/ 34 401 skin in the previous year 4) visible eczema on the cheeks, folds of the elbows, behind 35 402 knees, front of ankles or around neck 5) a debut before 24 months of age. 36 37 403 Food allergy 38 404 A diagnosis of food allergy requires: a medical history of allergic reactions, on at least two 39 405 consecutive occasions when eating a specific food allergen. The allergic reaction disappears 40 406 on elimination of the allergen. For early food reaction, allergic symptoms will appear <2h 41 407 with rush, itching, urticaria, angioedema, gastrointestinal pain, vomiting, oral symptoms, on September 28, 2021 by guest. Protected copyright. 42 408 asthma or anaphylactic reaction. Late food reaction (>2h) will present eczema, 43 409 gastrointestinal reactions with pain, diarrhoea, obstipation and weight loss. For all suspected 44 410 cases of food allergy, the diagnosis is confirmed by eliminationprovocation, i.e. the child is 45 411 given the suspected food allergen and symptoms are recorded. The eliminationprovocation 46 47 412 procedure is carried out at home following local guidelines if the symptoms are mild. If this is 48 413 not feasible, or if the case is unclear, the child will be admitted to the paediatric ward at the 49 414 Sunderby Hospital for repeated open food provocation until the diagnosis is clarified (food 50 415 allergic, or healthy). Open provocation will start with a dose determined by the individual 51 416 child’s food allergen reaction and will be followed by increasing doses of specific food 52 417 allergen given with 30 min interval in 35 steps. The child will stay at the ward for 2 hours 53 418 after the last dose and parents will thereafter report any symptoms during the following days. 54 55 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 419 Allergy to animals 4 420 A diagnosis of allergy to animal dander requires: a medical history of symptoms from the 5 421 eyes and/or nose and/or asthma on at least two consecutive occasions when exposed to a 6 422 specific animal dander (e.g. dog, cat, rabbit, horse). The reaction disappears when the animal 7 423 allergen is removed. 8 9 424 Pollen allergy 10 425 A diagnosis of pollen allergy requires: a medical history of symptoms from eyes and/or nose 11 426 and/or asthma on two consecutive occasions when exposed during pollen season. The reaction 12 427 disappears in areas with lower pollen content or following medication. 13 14 428 Asthma 15 429 At 12 months: An asthma diagnosis requires at least one of the following: 1) The presence of 16 430 wheeze betweenFor infections peer 2) Persistent reviewwheeze for at least only4 weeks 3) At least one period of 17 431 wheeze during an infectious disease with concomitant allergic disease (AD, food allergy or 18 19 432 allergic rhinoconjunctivitis) or 4) Three episodes of wheeze during infectious disease, with 20 433 no concomitant allergic disease. 21 434 At 48 months: An asthma diagnosis requires at least one of the following: The presence of 1) 22 435 wheeze within the last 12 months (during or between infections) or 2) ongoing controller 23 436 medication. 24 437 If there is any doubt concerning the asthma diagnosis at 12 or 48 months, the child will be 25 438 admitted to the children’s clinic at the Sunderby Hospital for a second clinical followup for 26 439 doctor´s diagnosed asthma. 27 28 440 Neuropsychological and neurobehavioral examination 29 441 At 48 months of age the parents will be asked to answer questionnaires and the child will 30 442 undergo examination with focus on neuropsychological behaviour. The questionnaires will 31 443 consist of parentreported Child Behaviour Checklist (CBCL), Adaptive Behaviour 32

444 Assessment System (ABAS) and Social Responsiveness Scale (SRS). Cognitive development http://bmjopen.bmj.com/ 33 34 445 will be tested facetoface with Wechsler Preschool and Primary Scale of Intelligence 35 446 (WPPSI) and a physical assessment of motor skills with Peabody Development Motor Scales 36 447 (PDMS; by a physiotherapist). 37 38 448 Oral health examination. 39 449 Caries will be recorded in children at 48 months of age by one to three experienced and 40 450 calibrated dentists (intra and interexaminer kappa ≥ 0.98) using a mirror and probe as the 41 451 number of decayed (enamel caries included), filled surfaces in the primary dentition (defs) on September 28, 2021 by guest. Protected copyright. 42 452 [1719]. At the same visit, caries will also be recorded for the parents using also two bitewing 43 453 radiographs as the number of decayed (enamel caries included), filled surfaces in the 44 454 permanent dentition (DeFS). Whole saliva and buccal swabs will be collected from the 45 455 parents. Complementary questionnaires on oral behavior data, such as oral hygiene, intake 46 456 frequency of sweets and sugary drinks, use of fluorides and oral hygiene, will be taken for 47 457 children and parents. Enamel discolorations and irregularities on visual inspection and 48 49 458 probing will registered and exfoliated primary teeth will be collected by parents in prepared 50 459 tubes. 51 52 53 54 55 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 460 Analyses 4 5 461 Immune analysis 6 462 Blood samples will be analysed by flow cytometry regarding numbers and percentages of T 7 463 and B cells and their subsets and their markers of naivety, activation and memory [33, 34]. 8 9 464 Genetic analyses 10 465 Genomic DNA will be extracted from blood collected in EDTA tubes from both mothers, 11 466 fathers and infants for analysis of genetic variation. Blood is also collected from the mothers 12 467 in Tempus™ Blood RNA Tubes for stabilization and isolation of total RNA from whole 13 468 blood for gene expression analysis. 14 15 469 Metabolomics 16 470 To study metabolicFor alterations peer in the participants, review nontargeted only metabolomics profiling will be 17 18 471 employed on plasma samples by liquid chromatography mass spectrometry (LCMS) and gas 19 472 chromatography mass spectrometry (GSMS) analysis [35]. 20 21 473 Transcriptomics 22 474 Whole transcriptome analysis will be performed on placenta tissue from those placentas 23 475 sampled within 4 h from birth and on maternal blood collected in Tempus™ RNA stabiliser 24 476 tubes at gestation, at birth and during lactation. 25 477 Total RNA will be analysed and its quality will be checked using a bioanalyzer. RNA 26 478 microarray analysis will be performed by next generation sequencing and the raw data will be 27 479 analysed using R biostatistical computing platform RStudio GUI. Gene set enrichment 28 480 analysis (GSA) will be performed to visualize regulated biological processes [36]. 29 30 481 Epigenetics 31 482 Wholeepigenome genespecific methylation analysis (Illumina EPIC) will be performed on 32 483 selected samples of the placenta, cord blood, and blood of children at 48 months of age to http://bmjopen.bmj.com/ 33 484 identify persisting genespecific changes in DNA methylation. Significant DNA methylation 34 485 markers will be validated by pyrosequencing in all children [37, 38]. 35 36 486 Fatty acid and vitamin A and D composition in blood cell membranes and plasma 37 38 487 Blood cell fatty acid composition will be analysed in unwashed red and white blood cells. 39 488 Fatty acids from blood cells will be extracted [39] and analysed using gas chromatography 40 489 flame ionisiation (GCFID) [40]. Concentrations of vitamin A and D will be determined in 41 490 plasma with LCMS/MS [41, 42]. on September 28, 2021 by guest. Protected copyright. 42 43 491 Determination of toxic metals, other elements and micronutrients 44 492 Concentration of toxic and essential elements will be measured in serum (Ca, Mg, Zn, Se), 45 493 erythrocytes (Cd, Mn, MeHg, Pb) and urine (As, Cd, U and I) using Inductively Coupled 46 494 Plasma Mass Spectrometry (ICPMS), developed for highthroughput multiple element 47 495 analyses [4345]. Speciation of urinary As (metabolites of inorganic arsenic) is performed by 48 496 HPLCICPMS. Urinary F is measured using an ionselective electrode (ISE). 49 50 497 Secretory IgA 51 498 A saliva sample will be collected from the mother close to the time of birth and from the 52 499 infant at 4 months for the analysis of secretory IgA and salivary microbiota [46]. 53 54 55 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 500 Oral and gut microbiota analyses 4 501 Infant faeces, oral swab and saliva will be collected for later analysis of microbiota 5 502 composition using DNA based sequencing methods [47]. 6 7 503 Data analysis 8 9 504 Most of the questionnaires are webbased and data are downloaded directly form the surveys 10 505 to SPSS files. Data will be cleaned and entered into a computer database securely stored 11 506 within the hospital and handled and protected with the same confidentiality as patient journal 12 507 data. One statistician, at the Unit for Research and Innovation, Department of Development, 13 508 Region Norrbotten, is responsible for entering data into the database and for handling raw 14 509 data files. A requirement for all requests for analysis of data from this cohort is that a detailed 15 16 510 statistical analysisFor plan ispeer submitted forreview approval by the only steering committee (explained 17 511 below). All data transferred to the Universities is blinded so that researchers cannot link 18 512 individual data to any specific individual. 19 20 513 Statistical analyses 21 514 SPSS, STATA, R and SIMCA are among statistical software’s used for data analysis. 22 515 Univariate and multivariate analysis and data mining, including cluster and hierarchical 23 516 models, will be used depending on data and outcome measures. Descriptive analysis will be 24 517 performed first to examine general maternal, paternal and infant demographic data and 25 518 distribution of the outcomes. Statistical methods will depend on the outcome measures. Chi2 26 519 test for proportions will be used for comparisons between categorical data, Ttests and 27 520 ANOVA (MannWhitney Utest and KruskalWallis test for nonparametric data) for 28 521 categorical versus continuous data and Pearson’s Correlation (Spearman’s Correlation for 29 522 nonparametric data) for continuous data. Regression analysis will be performed to identify 30 31 523 independent outcome predictors while controlling for possible confounders. 32 524 33 525 Statistical analyses will be performed in two steps: pilot analyses and final analyses. Pilot http://bmjopen.bmj.com/ 34 526 analyses refer to analysis of a subset of samples collected during 20152016, and final 35 527 analyses refer to data from samples collected from the whole study (20152018). 36 37 38 528 NICE consortium 39 40

529 Steering committee on September 28, 2021 by guest. Protected copyright. 41 42 530 The Steering Committee, consisting of MD, PhD, PI Anna Sandin, Professor AnnSofie 43 531 Sandberg and Professor Agnes Wold, is responsible for administrative and financial 44 532 coordination and monitoring the progress of the cohort. The Steering Committee will meet 45 533 twice a year to handle the allocation of resources to the research projects, the recruitment of 46 534 personnel and the scientific direction. 47 48 49 535 Research groups 50 51 536 Research groups involved in the NICEcohort are shown in table 5. 52 53 54 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 537 Collaborators 4 538 Within the NICEcohort there are close collaborators with a range of expertise. For 5 6 539 transcriptomics analysis and systems biology of the NICE samples we have a close 7 540 collaboration with Professor Jens Nilsen at Systems Biology at the Department of Biology 8 541 and Biological Engineering at Chalmers University of Technology in Gothenburg, Sweden. 9 542 Robert Lundqvist, statistician at the Research and Innovation Unit at Region Norrbotten, will 10 543 advise on statistical analyses and is also responsible for the management of the NICE cohort 11 544 database. 12 13 14 545 Ethics and dissemination 15 546 The study is conducted in accordance with the Helsinki Declaration and has been approved by 16 547 the Regional EthicalFor Review peer Board in Umeå,review Sweden. The only participants are informed about 17 548 the aims and the requirements of the NICEcohort, and are provided both written and verbal 18 549 information. They are requested to complete and sign a consent form before being enrolled in 19 550 the study. Once the infant is born, the parents are again requested to give consent for their 20 551 child’s participation in the study. All participants have the right to withdraw from the study at 21 552 any point and have their data removed from all study documentation. 22 23 553 24 554 All data are securely stored according to European laws and password protected and 25 555 accessible by the research team only. All personal data is blinded so that researchers cannot 26 556 link individual data to any specific individual. Information resulting from the study will be 27 557 disseminated through conferences and scientific publications, as well as to nonscientific 28 558 audiences in the region where the study has been carried out to keep health professionals and 29 559 interested members of the public abreast of the findings 30 31 32 560 Authors’ contributions 33 561 Anna Sandin, AnnSofie Sandberg and Agnes Wold initiated the study and are responsible for http://bmjopen.bmj.com/ 34 562 the conception and design of the work. Fiona Murray and Anna Sandin are responsible for 35 563 data collection. All authors have been involved in the design of the study and is involved in 36 564 data analysis and interpretation. Malin Barman and Alastair Ross drafted the first version of 37 565 the manuscript. All authors have been involved in writing different parts of the manuscript. 38 566 All authors contributed to the critical revision of the manuscript and have read and approved 39 40 567 the manuscript before submission. 41 on September 28, 2021 by guest. Protected copyright. 42 568 Acknowledgements 43 44 569 We would like to thank study nurses Marjut Larsson, Ulrika Börlin and study assistant nurse 45 570 Emma LundinRåberger, study midwifes Louise Lindgren and Lisa Sundén and study 46 571 administrator Elvira Sandin for their fantastic work with recruitment, followup assessments, 47 572 sample preparation and data collection. We also thank all personnel at the maternity clinics, 48 573 antenatal clinics and delivery ward for informing and recruiting patients and collecting 49 574 samples. We deeply thank all the families who participate in this study. We gratefully 50 575 acknowledge our collaborators Professor Jens Nilsen, Statistician Robert Lundqvist, PhD 51 576 Susanne Lager, MD PhD Linda EnglundÖgge, PhD Karin Larsson and Midwife Maria 52 577 Hallingström. 53 54 55 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 578 Funding 4 579 Funding has been received from the Swedish Research Council (VR) project no 5212013 5 580 3154, Swedish Research Council for Health, Working Life and Welfare (FORTE) project no 6 581 20140923, the Västra Götaland Region RUN 612061815, Åke Wibergs Stiftelse, Jane and 7 8 582 Dan Olsson Foundation, the Research and Innovation Unit at Region Norrbotten and through 9 583 a regional agreement between Umeå University and Västerbotten County Council in the field 10 584 of Medicine, Odontology, and Health (711251). 11 12 585 Competing interests 13 14 586 No competing interests are declared. 15 16 587 Patient consentFor peer review only 17 18 588 Patient consent is obtained for participation in this study, and in the case of minors, consent is 19 589 given by the parents 20 21 22 590 Ethics approval 23 591 The NICEcohort is approved by the Regional Ethical Review board in Umeå, Dept. for 24 592 Medical Research, Sweden: 2013/1831M, supplementary applications: 201428132M, 2015 25 26 593 7132M, 201530432M and 201623232M. 27 28 594 Provenance and peer review 29 30 595 This work has not been commissioned, and has been externally peerreviewed. 31 32 596 Open access 33 http://bmjopen.bmj.com/ 34 597 This is an Open Access article distributed in accordance with the terms of the Creative 35 598 Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt 36 599 and build upon this work, for commercial use, provided the original work is properly cited. 37 600 See: http://creativecommons.org/licenses/by/4.0/ 38 39 40 601 References 41 602 1. Berggren, S., et al., Parents with overweight children two and five years of age did not on September 28, 2021 by guest. Protected copyright. 42 603 perceive them as weighing too much. Acta Paediatr, 2017. apa.14174. 43 604 2. Wang, M., et al., Reduced diversity in the early fecal microbiota of infants with atopic 44 605 eczema. J Allergy Clin Immunol, 2008. 121(1): p. 12934. 45 46 606 3. Sandin, A., et al., Faecal short chain fatty acid pattern and allergy in early childhood. 47 607 Acta Paediatr, 2009. 98(5): p. 8237. 48 608 4. Jonsson, K., et al., Fat intake and breast milk fatty acid composition in farming and 49 609 nonfarming women and allergy development in the offspring. Pediatr Res, 2016. 79(1 50 610 1): p. 11423. 51 611 5. Dietert, R.R., Developmental Immunotoxicity, Perinatal Programming, and 52 612 Noncommunicable Diseases: Focus on Human Studies. Adv Med, 2014. 2014: p. 53 613 867805. 54 614 6. Ahmed, S., et al., Arsenic exposure and cell-mediated immunity in pre-school children 55 615 in rural Bangladesh. Toxicol Sci, 2014. 141(1): p. 16675. 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 616 7. Barman, M., et al., High levels of both n-3 and n-6 long-chain polyunsaturated fatty 4 617 acids in cord serum phospholipids predict allergy development. PLoS One, 2013. 5 618 8(7): p. e67920. 6 619 8. Gilman, S.E., et al., Socioeconomic disadvantage, gestational immune activity, and 7 620 neurodevelopment in early childhood. Proc Natl Acad Sci U S A, 2017. 114(26): p. 8 621 67286733. 9 622 9. Grether, J.K., et al., Prenatal and Newborn Immunoglobulin Levels from Mother- 10 623 Child Pairs and Risk of Autism Spectrum Disorders. Front Neurosci, 2016. 10(1662 11 624 4548): p. 218. 12 625 10. Meltzer, A. and J. Van de Water, The Role of the Immune System in Autism Spectrum 13 14 626 Disorder. Neuropsychopharmacology, 2017. 42(1): p. 284298. 15 627 11. Vuong, H.E. and E.Y. Hsiao, Emerging Roles for the Gut Microbiome in Autism 16 628 SpectrumFor Disorder. peer Biol Psychiatry, review 2017. 81(5): p. 411423.only 17 629 12. Gardner, R.M., et al., Environmental exposure to metals and children's growth to age 18 630 5 years: a prospective cohort study. Am J Epidemiol, 2013. 177(12): p. 135667. 19 631 13. Kippler, M., et al., Early-life cadmium exposure and child development in 5-year-old 20 632 girls and boys: a cohort study in rural Bangladesh. Environ Health Perspect, 2012. 21 633 120(10): p. 14628. 22 634 14. Hamadani, J.D., et al., Critical windows of exposure for arsenic-associated 23 635 impairment of cognitive function in pre-school girls and boys: a population-based 24 636 cohort study. Int J Epidemiol, 2011. 40(6): p. 1593604. 25 637 15. Skroder, H., et al., Early-Life Selenium Status and Cognitive Function at 5 and 10 26 27 638 Years of Age in Bangladeshi Children. Environ Health Perspect, 2017. 125(1552 28 639 9924). 29 640 16. Collaboration, N.C.D.R.F., Worldwide trends in body-mass index, underweight, 30 641 overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population- 31 642 based measurement studies in 128.9 million children, adolescents, and adults. Lancet, 32 643 2017. 390(10113): p. 26272642. 33 644 17. Kallestal, C., The effect of five years' implementation of caries-preventive methods in http://bmjopen.bmj.com/ 34 645 Swedish high-risk adolescents. Caries Res, 2005. 39(1): p. 206. 35 646 18. Stromberg, N., et al., Genetic- and Lifestyle-dependent Dental Caries Defined by the 36 647 Acidic Proline-rich Protein Genes PRH1 and PRH2. EBioMedicine, 2017. 37 648 26(Supplement C): p. 3846. 38 649 19. Esberg, A., et al., Streptococcus Mutans Adhesin Biotypes that Match and Predict 39 650 Individual Caries Development. EBioMedicine, 2017. 24: p. 205215. 40 41 651 20. Christensen, S.E., et al., Two new meal- and web-based interactive food frequency on September 28, 2021 by guest. Protected copyright. 42 652 questionnaires: validation of energy and macronutrient intake. J Med Internet Res, 43 653 2013. 15(6): p. e109. 44 654 21. Christensen, S.E., et al., Relative validity of micronutrient and fiber intake assessed 45 655 with two new interactive meal- and Web-based food frequency questionnaires. J Med 46 656 Internet Res, 2014. 16(2): p. e59. 47 657 22. Delisle Nystrom, C., et al., Validation of an Online Food Frequency Questionnaire 48 658 against Doubly Labelled Water and 24 h Dietary Recalls in Pre-School Children. 49 659 Nutrients, 2017. 9(1). 50 660 23. Bonn, S.E., et al., Active-Q: validation of the web-based physical activity 51 661 questionnaire using doubly labeled water. J Med Internet Res, 2012. 14(1): p. e29. 52 662 24. Bonn, S.E., et al., Feasibility of a novel web-based physical activity questionnaire for 53 54 663 young children. Pediatr Rep, 2012. 4(4): p. e37. 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 664 25. Zuriarrain, R., et al., Assessing the measurement of airway resistance by the 4 665 interrupter technique. Arch Argent Pediatr, 2013. 111(6): p. 495501. 5 666 26. Rech, V.V., et al., Airway resistance in children measured using the interrupter 6 667 technique: reference values. J Bras Pneumol, 2008. 34(18063756). 7 668 27. Skin tests used in type I allergy testing Position paper. Sub-Committee on Skin Tests of 8 669 the European Academy of Allergology and Clinical Immunology. Allergy, 1989. 9 670 44(Suppl 10:159.). 10 671 28. Sandin, A., B. Bjorksten, and L. Braback, Development of atopy and wheezing 11 672 symptoms in relation to heredity and early pet keeping in a Swedish birth cohort. 12 673 Pediatr Allergy Immunol, 2004. 15(4): p. 31622. 13 14 674 29. Hesselmar, B., et al., Early fish introduction is associated with less eczema, but not 15 675 sensitization, in infants. Acta Paediatr, 2010. 99(12): p. 18617. 16 676 30. Williams,For H.C., et al.,peer The U.K. Workingreview Party's Diagnostic only Criteria for Atopic 17 677 Dermatitis. I. Derivation of a minimum set of discriminators for atopic dermatitis. Br J 18 678 Dermatol, 1994. 131(3): p. 38396. 19 679 31. Williams, H.C., et al., The U.K. Working Party's Diagnostic Criteria for Atopic 20 680 Dermatitis. III. Independent hospital validation. Br J Dermatol, 1994. 131(3): p. 406 21 681 16. 22 682 32. Williams, H.C., et al., The U.K. Working Party's Diagnostic Criteria for Atopic 23 683 Dermatitis. II. Observer variation of clinical diagnosis and signs of atopic dermatitis. 24 684 Br J Dermatol, 1994. 131(3): p. 397405. 25 685 33. Lundell, A.C., et al., High proportion of CD5+ B cells in infants predicts development 26 27 686 of allergic disease. J Immunol, 2014. 193(2): p. 5108. 28 687 34. Strombeck, A., et al., High proportions of FOXP3(+) CD25(high) T cells in neonates 29 688 are positively associated with allergic sensitization later in childhood. Clin Exp 30 689 Allergy, 2014. 44(7): p. 94052. 31 690 35. Savolainen, O.I., A.S. Sandberg, and A.B. Ross, A Simultaneous Metabolic Profiling 32 691 and Quantitative Multimetabolite Metabolomic Method for Human Plasma Using 33 692 Gas-Chromatography Tandem Mass Spectrometry. J Proteome Res, 2016. 15(1): p. http://bmjopen.bmj.com/ 34 693 25965. 35 694 36. Soni, N.K., et al., Splenic Immune Response Is Down-Regulated in C57BL/6J Mice 36 695 Fed Eicosapentaenoic Acid and Docosahexaenoic Acid Enriched High Fat Diet. 37 696 Nutrients, 2017. 9(1). 38 697 37. Broberg, K., et al., Arsenic exposure in early pregnancy alters genome-wide DNA 39 698 methylation in cord blood, particularly in boys. J Dev Orig Health Dis, 2014. 5(4): p. 40 41 699 28898. on September 28, 2021 by guest. Protected copyright. 42 700 38. Kippler, M., et al., Sex-specific effects of early life cadmium exposure on DNA 43 701 methylation and implications for birth weight. Epigenetics, 2013. 8(5): p. 494503. 44 702 39. Masood, A., K.D. Stark, and N. Salem, Jr., A simplified and efficient method for the 45 703 analysis of fatty acid methyl esters suitable for large clinical studies. J Lipid Res, 46 704 2005. 46(10): p. 2299305. 47 705 40. Barman, M., et al., Serum fatty acid profile does not reflect seafood intake in 48 706 adolescents with atopic eczema. Acta Paediatr, 2014. 103(9): p. 96876. 49 707 41. Barman M, et al., Serum levels of Vitamin A and Atopic Rhinoconjunctivitis in 50 708 Swedish adolescents. Food Sci Nutr The, 2017. 3(1): p. 014019. 51 709 42. Barman, M., et al., No association between allergy and current 25-hydroxy vitamin D 52 710 in serum or vitamin D intake. Acta Paediatr, 2015. 104(4): p. 40513. 53 54 711 43. Lu, Y., et al., Alkali dilution of blood samples for high throughput ICP-MS analysis- 55 712 comparison with acid digestion. Clin Biochem, 2015. 48(3): p. 1407. 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 713 44. Rydbeck, F., et al., Urinary iodine concentrations of pregnant women in rural 4 714 Bangladesh: a longitudinal study. J Expo Sci Environ Epidemiol, 2014. 24(5): p. 504 5 715 9. 6 716 45. FrenckMestre, C., M. Besson, and J. Pynte, Finding the locus of semantic satiation: 7 717 an electrophysiological attempt. Brain Lang, 1997. 57(3): p. 40622. 8 718 46. Sandin, A., et al., High salivary secretory IgA antibody levels are associated with less 9 719 late-onset wheezing in IgE-sensitized infants. Pediatr Allergy Immunol, 2011. 22(5): 10 720 p. 47781. 11 721 47. Sjoberg, F., et al., Low-complexity microbiota in the duodenum of children with newly 12 722 diagnosed ulcerative colitis. PLoS One, 2017. 12(10): p. e0186178. 13 14 723 15 724 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from Page 20 of 28 19 19

low low up l

BMJ Open http://bmjopen.bmj.com/

on September 28, 2021 by guest. Protected copyright. 014 014 2015 2016 2017 2018 2019 2020 2021 2022

2 For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

Table 1: Overview of the NICE the Overview of 1: studyTable recruitment andfo 726 726 Data collection collection Data 48 months follow follow up 48 months 12 month follow up follow up 12 month Children born born Children Recruitment of patients of patients Recruitment Project start, planning planning start, Project

725 725 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from 20 20

C C

1 2 C

C 1 C

C C C C M M C C C C C M M C C

1 1 1 1

4 mo 12 mo 48 mo

1 M C M C C 1 1 1

BMJ Open

http://bmjopen.bmj.com/

on September 28, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

28 Birth 48 h 1 mo M For3 peer review only M C M C M C C C C C M C M M P C C M

M M P

GW C C P C C C 4

Table 2: sample Overview of collection2: Tablein NICE.

Oral health Growth and obesity Growth obesity and Doctor’s diagnosis of allergy Doctor’s of diagnosis C C SPT SPT RiNT Clinical examination Clinical examination

Breast Breast milk Placenta Mouth swabs Mouth swabs Faeces Faeces Saliva Saliva Hair Urine Blood Biological samples Biological samples

Infections and and Infections feeding Dietary Assessment Dietary Assessment Allergy and environment environment and Allergy M P C Questionnaires Neurologic Neurologic test 727 727 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 21 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from Page 22 of 28 21 21 BMJ Open http://bmjopen.bmj.com/ on September 28, 2021 by guest. Protected copyright. For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Only subcohort subcohort Only 18 months At 34 week gestational At of age to months from 12 month birth Every

development, anthropometric measurements as well as an oral health examination. examination. as health an oral as well measurements anthropometric development, the the Sunderby Hospital. Midwives at the maternity ward are responsible for all sampling at the birth timepoint and assist with collection of Abbreviations: Abbreviations: Gw = gestational week, M = maternal samples, P = paternal samples, C = child samples, SPT = skin prick test, RiNT = samples samples from the neonates at 48h. For all other time points post48 h the same clinical team, consisting of two specialized neonatal nurses, respiratory resistance with interruption technique. technique. with interruption resistance respiratory will will meet all participating families. The same paediatric allergy specialist will examine all infants at 12 months and 48 months of age to diagnose diagnose any allergy and/or allergen sensitisation. At 48 months, the children will also undergo an assessment of neuropsychological All All studyrelated visits, with the exception of the first sampling of mothers at gestational week 28 at the maternity clinics, will take place at 1 2 3 4 736 736 732 732 737 737 733 733 738 738 734 734 739 739 735 735 728 728 729 730 731 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 23 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 740 Table 3: Overview of questionnaires distibuted in the NICE-cohort 5 Questionnaire Time Answered by Questions about 6 7 Q1: “Allergic GW 2025 Mothers and Education, profession, nationality, 8 heredity and partners answer allergic heredity, smoking habits, 9 environment” one questionnaire alcohol habits, residence, heating, 10 each ventilation, “mould”, siblings, 11 pets. 12 Q2: “Infant 18 months Parents answers Questions to both parents 13 environment” about the family regarding: profession, education, 14 ethnicity, civil status, smoking 15 habits, hunting habits, residence. 16 For peer reviewQuestions only about the infant: 17 feeding, vaccinations, travels, 18 daycare, behaviour, siblings, pets. 19 20 Q3: “Child 48 months Parents answers Questions to both parents 21 environment” about the family regarding: profession, education, 22 civil status, smoking habits, 23 hunting habits, residence 24 Questions about the infant: 25 feeding, vaccinations, travels, 26 daycare, behaviour , siblings, pets, 27 741 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 742 Table 4: Overview of the dietary assessments in NICE 5 Questionnaire Time Answered by 6 7 MealQ1+ ActiveQ1 GW 34 Mothers 8 MealQ2+ ActiveQ2 1 month Mothers 9 MealQ3 + ActiveQ3 4 months Mothers 10 TodMealQ+ TodActiveQ 12 months Parents answer for the infant 11 KidMealQ+ KidActiveQ 48 month Parents answer for the infant 12 743 GW: Gestational week 13 744 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 745 Table 5: Overview of research sites involved in the NICE-consortium 5 Institution, Department, Group Principal investigator Main research interests in the 6 7 of the site NICE-cohort 8 Umeå University, Clinical Anna Sandin, MD, PhD, Environmental exposures in 9 Sciences, Paediatrics. Principal Investigator for relation to immune maturation 10 the NICEcohort. and allergy outcomes. 11 Chalmers University of AnnSofie Sandberg, Nutritional factors in relation to 12 Technology, Biology and Professor allergy, growth and neurological 13 Biological Engineering, Food and function. Metabolomics to 14 Nutrition Science. predict risk of disease. Systems 15 biology. 16 Gothenburg University,For Clinical peer Agnes review Wold, Professor onlyImmune maturation, microbiota 17 Microbiology. and allergy development. 18 19 Karolinska Institutet, Institute of Marie Vahter, Professor Metal exposure and epigenetics 20 Environmental Medicine. in relation to growth and 21 neurological function. 22 Gothenburg University, Bo Jacobsson, Professor Pregnancy exposures and 23 Sahlgrenska Academy, Obstetrics outcome with special interest in 24 and Gynecology. placenta. 25 Umeå University, Department of Nicklas Strömberg, Nutritional, immune, microbiota 26 Odontology/Cariology. Professor and metal factors in caries 27 development. 28 Karolinska Institute, Department Sven Bölte, Professor Neuropsychological and 29 of Women’s and Children’s neuropsychiatric assessments 30 31 Health, Division of 32 Neuropsychiatry, Center of 33 Neurodevelopmental Disorders http://bmjopen.bmj.com/ 34 (KIND). 35 746 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 28 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 747 Figure legends 5 748 Figure 1: Flow chart of the NICE-cohort 6 749 7 8 750 Figure 2: Schematic overview of placenta collection at delivery. 751 Samples from the placenta are collected for analysis of histology, DNA, RNA, proteins, 9 3 10 752 metabolomics, fatty acids, toxicants and essential elements. First 4 pieces, 4 cm of villous 11 753 tissue including decidua is collected from the basal plate from 4 separate places of the 12 754 placenta. The decidua is removed stored in a separate cryotube. The remaining villous tissue 13 755 is divided into 20 smaller pieces, 4 pieces of 3 mm Ø that are put in RNAlater® and 16 pieces 14 756 that are put 4 and 4 in separate cryotubes. Two different cross sections are then collected: one 15 757 crosssection measuring approximately 2x2 cm taken within 4 cm from the cord insert is 16 758 placed in formalin.For Another peer crosssection review formed as a V taken only from the cord insert to the end 17 759 of the placenta is sampled with specific metal free ceramic knife and put in to plastic bags. 18 760 The crosssections are taken with care to avoid selecting a tissue section that has previously 19 761 been sampled for villous tissues. At one cm from the cord insert a 2 cm piece of the umbilical 20 21 762 cord is sampled and stored in formalin. From the same end of the cord 2 pieces, 1 cm each in 22 763 length, are sampled and placed in a cryotube. Also, samples from the membrane are 23 764 collected: pieces of the amnion and chorion membranes are placed in separate cryotubes, 24 765 while one piece of membrane with both the membranes still attached to eachother (“double 25 766 membrane”) is placed in formalin. 26 767 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 5 6 768 Figure 1: 7 Information was sent 8 home to families that Accepted information 9 accepted 10 information. 11 Declined participation 12 13 Mothers and fathers 14 signed a consent Included families 15 form 16 For peer reviewDrop outsonly 17 18 19 Mothers sampled at 28 GW follow-up 20 the maternity clinics. 21 22 Drop outs 23 24 Mothers and children 25 sampled at the Children born 26 delivery ward. 27 Drop outs 28 29 Children and fathers 30 sampled at the after 48 h follow-up 31 birth ward. 32 33 Drop outs http://bmjopen.bmj.com/ 34 Mothers and children 35 sampled by the study 1 months follow-up 36 nurses at the NICE Only sub-cohort 37 laboratory. 38 Drop outs 39 40 Mothers and children 41 sampled by the study on September 28, 2021 by guest. Protected copyright. 4 month follow-up 42 nurses at the NICE 43 laboratory. 44 Drop outs 45 46 Clinical examination 47 of the children at the 1 year follow-up 48 paediatric ward. 49 Drop outs 50 51 Clinical examination 52 of the children at the 4 years follow-up 53 paediatric ward. 54 769 55 56 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from Page 28 of 27 27

Proteins

RNA &

acids

Fatty metabolomics tissue

Decidua

Villous DNA

& Proteins

cm

pieces

2x2 4 BMJ Open http://bmjopen.bmj.com/ 2. 4.

cord

plate)

on September 28, 2021 by guest. Protected copyright. Umbilical PLACENTA (Basal For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 3. For peer1. review only

Foetal

membranes Cross Toxicants sections Histology

Figure 2: Figure 2:

1 1 7 770 770 7 772 772 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from 28 28 BMJ Open http://bmjopen.bmj.com/ on September 28, 2021 by guest. Protected copyright. For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

783 783 they eat that report who those respondents ato followup question 3. of Addition flaxseeds in the adult questionnaire, plus corresponding modification of the of the modification corresponding plus adult questionnaire, flaxseeds in the nutrition calculation program (i.e. different nutritional content of crushed crushed content of nutritional (i.e. different calculation program nutrition 791 791 and sweet distinguish between program to calculation the nutrition of 5. Modification 788 788 flaxseeds, respectively). whole and products with pattern of consumption the participants’ regarding questions 4. of Addition and "light" drinks. The question of how often you drink "soda, cider, cider, "soda, you often drink of how "light"The question drinks. and adult questionnaire. These questions were not included in the nutrition calculation. calculation. nutrition in the included were questions not These adult questionnaire. fruit syrups, juice" on page 4 was coupled with the presence of light products light products presence of the coupled with juice"4 was on page fruit syrups, and the alternative "Soft drinks, or juice" on page 14. juice" page "Soft drinks, or on the alternative and Supplementary file 1: Modifications of the dietary questionnaires Meal-Q, Todmeal-Q and Kidmeal-Q Kidmeal-Q and Todmeal-Q Meal-Q, questionnaires dietary of the 1: Modifications file Supplementary 774 774 different between distinguish program to calculation the nutrition of 1. Modification 796 796 of the nutritional modifications and children subsequent small of "Infant to Formula"types different regarding questions 6. of Addition fat contents of "Milk, sour milk or yogurt" and "Other milk products, e.g. cream cream e.g. milk "Other products, and or yogurt" milk "Milk, sour of fat contents 799 799 calculation program. or crème fraiche", when consumption patterns of these food items are asked about, it is followed by a question if you usually choose light light choose usually you if question a by followed about, it is asked are items food of these consumption patterns fraiche", when or crème 779 779 during and before fish and species game regarding lamb, meat questions 2. of Addition probiotics in the adult survey (does not affect the nutrition calculation program). program). calculation affect the nutrition (does adult not survey in probiotics the pregnancy, organic food purchases, and gluten and lactose in in the and lactose and purchases, gluten food organic pregnancy, products. products. 782 782 784 784 785 785 790 790 786 786 787 792 792 781 781 789 789 793 793 780 780 794 794 773 773 795 795 775 775 797 797 798 776 776 777 777 778 Page 29 of 28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

Study protocol: “Nutritional impact on Immunological maturation during Childhood in relation to the Environment (NICE)”, a prospective birth cohort in northern Sweden

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2018-022013.R1

Article Type: Protocol

Date Submitted by the Author: 10-Jul-2018

Complete List of Authors: Barman, Malin; Chalmers University of Technology, Department of Biology and Biological Engineering; University of Gothenburg, Institute of Clinical Sciences, Department of Obstetrics and Gynaecology Murray, Fiona; Region Norrbotten, Sunderby Research Unit Bernardi, Angelina; University of Gothenburg, Institute of Biomedicin, Department of Infectious Diseases Broberg, Karin; Karolinska Institutet, Institute of Environmental Medicine Bölte, Sven; Karolinska Institutet, Department of Women’s and Children’s Health Hesselmar, Bill; University of Gothenburg, Institute of Clinical Sciences, Department of Paediatrics Jacobsson, Bo; University of Gothenburg, Institute of Clinical Sciences, Department of Obstetrics and Gynecology Jonsson, Karin; Chalmers University of Technology, Department of Biology http://bmjopen.bmj.com/ and Biological Engineering Kippler, Maria; Karolinska Institutet, Institute of Environmental Medicine Rabe, Hardis; Gothenburg University, Institute of Biomedicine, Department of Infectious Diseases Ross, Alastair; Chalmers University of Technology, Department of Biology and Biological Engineering Sjöberg, Fei; University of Gothenburg, Institute of Biomedicine, Department of Infectious Diseases

Strömberg, Nicklas; Umeå University, Department of Odontology/Cariology on September 28, 2021 by guest. Protected copyright. Vahter, Marie; Karolinska Institutet, Institute of Environmental Medicine Wold, Agnes; University of Gothenburg, Institute of Biomedicine, Department of Infectious Diseases Sandberg, Ann-Sofie; Chalmers University of Technology, Department of Biology and Biological Engineering Sandin, Anna; Umeå University, Department of Clinical Sciences

Primary Subject Paediatrics Heading:

Immunology (including allergy), Nutrition and metabolism, Dentistry and Secondary Subject Heading: oral medicine, Neurology, Genetics and genomics

birth cohort, Allergy < THORACIC MEDICINE, Nutrition < TROPICAL Keywords: MEDICINE, Microbiology < PATHOLOGY, toxicants, Immunology < BASIC SCIENCES

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 1 Study protocol: “Nutritional impact on Immunological maturation 4 2 5 during Childhood in relation to the Environment (NICE)”, a 6 3 prospective birth cohort in northern Sweden 7 4 8 5 Barman, M.1+2, Murray, F.3, Bernardi, AI.4, Broberg, K.5, Bölte, S.,6 Hesselmar, B.7, 9 6 Jacobsson, B.2, Jonsson, K.1, Kippler, M.5, Rabe, H.4 Ross, A.1, Sjöberg, F.4, Strömberg, N.8, 10 5 4 1 3+9 11 7 Vahter, M. , Wold, A. E. , Sandberg, AS. , and Sandin, A.

12 8 1 13 9 Food and Nutrition Science, Department of Biology and Biological engineering, Chalmers 14 10 University of Technology, Göteborg, Sweden. 2 15 11 Department of Obstetrics and Gynecology, Institute of Clinical Sciences, the Sahlgrenska 16 12 Academy, UniversityFor of Gothenburg, peer Göteborg, review Sweden. only 17 13 3Sunderby Research Unit, Region Norrbotten, Sweden. 18 14 4Department of Infectious Diseases, Institute of Biomedicine, the Sahlgrenska Academy, 19 15 University of Gothenburg, Göteborg, Sweden. 20 16 5Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden. 21 17 6Center of Neurodevelopmental Disorders (KIND), Division of Neuropsychiatry, Department 22 18 of Women’s and Children’s Health, Karolinska Institutet, Stockholm, Sweden. 23 7 24 19 Department of Paediatrics, Institute of Clinical Sciences, University of Gothenburg, 20 Göteborg, Sweden. 25 9 26 21 Department of Odontology/Cariology, Umeå University, Umeå, Sweden. 10 27 22 Department of Clinical Sciences, Unit of Pediatrics, Umeå University, Umeå, Sweden. 28 23 29 24 Corresponding author: Malin Barman, Chalmers University of Technology, Department of 30 25 Biology and Biological Engineering, Food and Nutrition Science, SE412 96 Gothenburg, 31 26 Sweden. Telephone: +46(0)317723811; Fax: +46(0)317723830; e mail: 32 27 [email protected] 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 28 Abstract 4 5 29 Introduction 6 30 Pre and neonatal environmental factors, such as nutrition, microbes and toxicants, may affect 7 31 health throughout life. Many diseases, such as allergy and impaired child development, may 8 32 be programmed already in utero or during early infancy. Birth cohorts are important tools to 9 33 study associations between early life exposure and disease risk. Here, we describe the study 10 34 protocol of the prospective birth cohort, “Nutritional impact on Immunological maturation 11 12 35 during Childhood in relation to the Environment” (NICE). The primary aim of the NICE 13 36 cohort is to clarify the effect of key environmental exposures – diet, microbes and 14 37 environmental toxicants during pregnancy and early childhood, on the maturation of the 15 38 infant’s immune system, including initiation of sensitisation and allergy as well as some 16 39 secondary outcomes:For infant peer growth, obesity, review neurological development only and oral health. 17 18 40 Methods and analysis 19 41 The NICEcohort will recruit about 650 families during midpregnancy. The principal 20 42 inclusion criterion will be planned birth at the Sunderby Hospital in the north of Sweden, 21 43 during 2015 – 2018. Questionnaires data and biological samples will be collected at ten time 22 44 points, from pregnancy until the children reach four yearsofage. Samples will be collected 23 45 primarily from mothers and children, but also from fathers. Biological samples include blood, 24 46 urine, placenta, breast milk, meconium, faeces, saliva and hair. Information regarding allergic 25 47 heredity, diet, socioeconomic status, lifestyle including smoking, siblings, pet ownership etc. 26 48 will be collected using questionnaires. Sensitization to common allergens will be assessed by 27 49 skinprick testing and allergic disease will be diagnosed by a paediatrician at one and four 28 29 50 years of age. At four yearsofage the children will also be examined regarding growth, 30 51 neurobehavioral and neurophysiological status and oral health. 31 32 52 Ethics and dissemination 33 53 The NICEcohort has been approved by the Regional Ethical Review Board in Umeå, Sweden http://bmjopen.bmj.com/ 34 54 (2013/1831M). Results will be disseminated through peer reviewed journals and 35 55 communicated on scientific conferences. 36 56 37 38 39 57 Key words (6 maximum) 40 58 Birth cohort, allergy, immunology, nutrition, microbiology, toxicants. 41 on September 28, 2021 by guest. Protected copyright. 42 43 59 Strengths and limitations of this study 44 60 • Prospective study design covering a period from midpregnancy to the age of four 45 61 years with biological samples from both the mother, father and child for evaluation of 46 62 microbiology, nutrition, immunology, environmental toxicants, genetics and 47 63 epigenetics. 48 64 • Interdisciplinary, translational approach and advanced analytical methods within the 49 65 field of allergology, immunology, nutrition, microbiology, toxicology, obstetrics, 50 66 childhood growth, neurophysiological development and oral health. 51 52 67 • The primary outcome, allergic disease (food allergy, atopic eczema, asthma and 53 68 allergic rhinoconjunctivitis) is diagnosed by a paediatric allergology specialist 54 69 according to strictlydefined protocols. 55 70 • The large participant burden may also result in selection bias and loss to followup. 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 71 • Although the study sample will allow indepth analyses of the effect of multiple 4 72 exposures and genetic factors on several outcomes, these analyses may largely be 5 73 exploratory and may require confirmation in follow upstudies 6 7 8 74 Introduction 9 75 The foetal and neonatal periods could be regarded as a window of opportunity for influencing 10 76 health during childhood and throughout life. Environmental exposures, such as nutrition, 11 77 microbiota and toxicants, and genetic or epigenetic factors may synergise to impact growth, 12 78 neurodevelopment and immune maturation. However, which factors are most important and 13 79 the mechanisms by which they operate are still poorly understood. The Nutritional impact on 14 80 Immunological maturation during Childhood in relation to the Environment (NICE) cohort 15 81 focuses primarily on early life exposures, immune maturation and risk of immunemediated 16 82 diseases. For peer review only 17 18 83 19 84 Allergy, dental caries and overweight are the most common chronic conditions in childhood, 20 85 affecting the health and wellbeing of around 25%, 20% and 13% of children respectively [1]. 21 86 Chronic diseases place a substantial pressure on both the individual and on the health care 22 87 system. Hence, it is important to explore how lifestyle, environmental and genetic/epigenetic 23 88 factors during pregnancy and the first years of life protect against or facilitate the 24 89 development of chronic diseases. 25 90 26 91 Allergy is characterized by dysregulation of the immune system leading to an incapacity to 27 92 develop tolerance to normally harmless proteins in food and air. Decreased exposure to 28 93 microorganisms and reduced complexity of the gut microbiota in early life have been 29 94 associated with an increased risk for the development of allergy [2, 3]. Also, the diet of 30 31 95 pregnant and lactating women [4] and their exposure to environmental toxicants [5, 6] have 32 96 been suggested to play a role in immune maturation and allergy development. In spite of 33 97 extensive research in this area, the specific causes and pathways leading to immune http://bmjopen.bmj.com/ 34 98 maturation, and their subsequent protective or detrimental influences on allergy development, 35 99 remain largely unknown 36 100 37 101 In order to identify the key factors leading to the development of healthy immune regulation 38 102 or allergic disease, it is necessary to study immune events at a very early age [7], and account 39 103 for external factors that could influence immune development. Allergy may present as early 40 104 as a few months of age, usually manifested as atopic eczema or food allergy, while asthma on September 28, 2021 by guest. Protected copyright. 41 105 and rhinoconjunctivitis generally appear at school age. A major source of imprecision in 42 106 many studies on allergy development is the diagnostic methodology. While questionnaires are 43 107 convenient and costeffective, the diagnosis of childhood allergy by a skilled clinician, aided 44 45 108 by tests of sensitization, is the gold standard. For food allergy, a careful investigation of 46 109 medical history by an experienced clinician, aided by eliminationprovocation is the only 47 110 method to adequately identify this type of allergic manifestation. 48 111 49 112 Multiple environmental factors, including diet and exposure to toxicants, may affect both the 50 113 immune system and neurophysiological development [811]. Nutritional factors considered to 51 114 be immunomodulatory are unsaturated fatty acids, vitamins A and D, iron and zinc. Some 52 115 nutrients as well as some metals pass the placenta to the foetus. Some metals accumulate in 53 116 the placenta, thus elemental metals may cause both impaired placental function and foetal 54 117 toxicity. Even lowlevel exposure to certain metals, e.g. cadmium (Cd), arsenic (As), 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 118 manganese (Mn) and lead (Pb), has been found to impair foetal and childhood immune 4 119 function, growth and neurodevelopment [1215]. 5 120 6 121 Overweight and obesity are increasing worldwide. Overweight at a young age is an important 7 122 risk factor for obesity later in life. In Sweden obesity has increased during the last 40 years 8 123 from 2.3 to 4.7% amongst girls and from 2.6 to 8.5% amongst boys [16]. Socioeconomic 9 124 status and physical activity are important determinants, but there are other factors to explore 10 125 such as the impact of different dietary components (e.g. fatty acids) and environmental 11 126 toxicants, metabolic processes and epigenetic factors, to name just a few. 12 127 13 14 128 In Western countries, one fifth of children are socalled highcaries cases (children with a 15 129 high burden of dental caries) that do not respond to traditional prevention [17]. These high 16 130 caries cases deriveFor from peer an immune deficiencyreview or carriage only of high virulence strains of 17 131 Streptococcus mutans, consistent with their nonresponder behaviour [18, 19]. By contrast, 18 132 the lowtomoderate risk children develop caries as expected from eating and oral hygiene 19 133 habits [18]. Therefor the NICE study aims to evaluate the oral health of all participating 20 134 children in order to better understand the link between impaired immune maturation and the 21 135 risk for caries development. 22 23 24 136 Aims and objectives 25 26 27 137 Primary aim 28 138 The primary aim of the NICE birth cohort is to evaluate how multiple environmental 29 139 exposures such as lifestyle, diet, microbes and toxicants influence maturation of the immune 30 140 system and affect allergy development. 31 32 33 141 Secondary aims http://bmjopen.bmj.com/ 34 142 The secondary aim of NICE is to examine if these same environmental exposures influence 35 143 pregnancy outcomes, childhood growth, obesity, neuropsychological development, and caries 36 37 144 development. 38 39 145 Research questions 40 41 146 The NICE cohort has been designed to answer the following research questions: on September 28, 2021 by guest. Protected copyright. 42 147 i. Which nutrients affect the maturation of the infant’s immune system, allergy 43 148 development, oral health, neurophysiological development, growth and obesity? 44 149 ii. Can a metabolic fingerprint be identified that predicts disease risk during childhood? 45 150 iii. What role do the oral and gut microbiota play in immunological maturation, allergy 46 151 development, oral health, neurophysiological development, growth and obesity? 47 152 iv. What is the impact of early life exposure to toxicants, particularly toxic metals, on 48 153 immune maturation, allergy development, caries diseases, neurophysiological 49 154 development, growth and obesity? 50 155 v. Does maturation of the immune system during foetal life and infancy affect the risk of 51 156 allergy development, oral health or neurophysiological development? 52 53 157 vi. Can certain immune signatures be identified that are associated with immune 54 158 dysregulation, caries development or, conversely, with health? 55 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 159 vii. Is early life exposure to toxic metals and essential elements reflected in the dental 4 160 enamel? 5 161 viii. Which nutritional, environmental and lifestyle profiles are protective against caries in 6 162 the high versus the lowtomoderate caries risk groups and types? 7 163 ix. What is the impact of nutrition and toxic exposures on the epigenetic programming of 8 164 children in relation to immune maturation, allergy development, oral health, 9 165 neurophysiological development, growth and obesity? 10 166 x. What are the effects of perinatal exposures (diet, lifestyle and environmental 11 167 toxicants) on pregnancy outcomes such as gestational diabetes, preeclampsia, 12 168 pregnancy length, intraeuterine growth and birth weight? 13 14 15 169 Methods and Analysis 16 For peer review only 17 18 170 Study design and setting 19 171 The study design is a prospective longitudinal birth cohort study. All expectant parents 20 21 172 planning to give birth at the Sunderby Hospital in Luleå, during 2015 – 2018, will be invited 22 173 to participate. The Sunderby Hospital, with approximately 2,000 deliveries per year, is located 23 174 in Region Norrbotten in the very north of Sweden. Region Norrbotten makes up almost a 24 175 quarter of Sweden geographically, but has a population of only approximately 250,000 25 176 inhabitants. Expectant mothers are cared for at 22 maternity clinics across the region. Routine 26 177 ultrasound is performed at gestational week 1819 at either of three sites: at the Sunderby, 27 178 Kalix or Piteå hospitals. These three hospitals have combined catchment area of 28 179 approximately 200,000 inhabitants. 29 30 31 180 Recruitment and inclusion and exclusion criteria 32 181 All expectant parents living in the catchment area of the Sunderby Hospital will receive a http://bmjopen.bmj.com/ 33 182 concise information leaflet about the study at their first visit to their local maternity clinic in 34 183 gestation week 1012. Recruitment will take place in connection with the routine ultrasound at 35 184 gestational week 1819. All expectant parents will be given detailed information about the 36 37 185 study. The parents who wish to participate will be requested to provide informed consent. 38 186 They will subsequently receive an acknowledgement letter and assigned a study ID number. 39 187 40 188 The criteria for participation in the study will be as follows: the intention to give birth at the 41 189 Sunderby Hospital, the ability to communicate in written and spoken Swedish and current on September 28, 2021 by guest. Protected copyright. 42 190 habitation in Region Norrbotten. Families will be excluded from the study if they move from 43 191 Norrbotten before the child is born. Families who move outside of Norrbotten after birth will 44 192 be allowed to continue participation in the study if so wished and followup appointments will 45 193 be arranged where possible. 46 47 48 194 Estimated sample size 49 195 The NICEcohort will be based on the conservative inclusion of children to be delivered at the 50 196 Sunderby Hospital with runin during the period 20150115 – 20180228 (Table 1). The aim 51 197 is to include 650 families. Assuming 25% allergy rate among 650 families we estimate the 52 53 198 effects that can be detected with 80% power and 5% significance level in univariate analysis 54 199 of marginal effects. For continuous variables an effect size (Cohen's d) of 0.25 can be 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 200 detected and for dichotomous exposures ranging in prevalence among allergic patients (5%, 4 201 10%, 20%, 50%) the corresponding odds ratios that can be detected are (5.3, 2.9, 2.0, 1.7). 5 6 202 Study outcomes 7 8 203 Primary outcome 9 204 The primary outcome in NICE is immune maturation in the neonatal period and its effect on 10 205 development of sensitization and allergy over a followup period of four years. 11 12 206 Secondary outcomes 13 14 207 The secondary outcomes are: 15 208 pregnancy outcomes (preeclampsia, gestational diabetes, gestational week at birth, 16 209 intraeuterineFor growth peer and birth weight). review only 17 210 child growth (including overweight and obesity). 18 211 neurophysiological development (test of motor, cognitive and behavioural function). 19 212 caries and oral health 20 21 213 22 Biological sample collection 23 214 Mother, fathers and infants participating in the NICEcohort will be repeatedly followedup 24 215 from pregnancy until 4 years of age with both questionnaires and biological sample collection 25 216 (Figure 1). Mothers will fill out questionnaires and/or give biological samples at gestational 26 217 week 18, 28 and 34, at delivery and at 1, 4 and 12 months postpartum. Fathers will fill out 27 218 questionnaires and/or give biological samples at gestational week 18 and 48 hours and 12 28 219 months postpartum. All children will be followed up with biological samples or clinical 29 30 220 examinations at birth and at 48 hours, 4, 12, 18 and 48 months of age. The flow chart of the 31 221 study is shown in figure 1 and a detailed overview of the collection of biological samples and 32 222 questionnaires is shown in table 2. 33 http://bmjopen.bmj.com/ 34 223 Subcohort 35 224 Approximately one in four recruited families will be offered the possibility to participate in a 36 225 subcohort with more extensive sampling of mother and infant (Table 2). Mothers and children 37 226 in the subcohort will be followed up additionally at 1 months of age. Also, children in the 38 227 subcohort will collect feces samples at week 1, 2, 4 and 6 and at month 4 and 6. The 39 228 assignment of families to the subcohort will be performed at delivery, based on the feasibility 40 229 of extended sampling at that timepoint. 41 on September 28, 2021 by guest. Protected copyright. 42 230 Blood samples 43 231 Blood samples collected from the parents will include: 10 mL/5 mL SST II Advance serum 44 232 tube, 10mL/3mL EDTA tube (all fromBecton Dickinson, Plymouth, UK), 6 mL Naheparin 45 233 trace elementfree tube (Greiner bioone, Kremsmünster, Austria), and 10 mL RNAstabilised 46 234 blood collection Tempus™ tubes (Applied Biosystems, Foster City, CA), the latter from 47 235 mothers only. The mothers will be asked to be fasting for 8 h before study visits. 48 49 236 50 237 At birth, venous and arterial cord blood will be collected from the umbilical cord into small 51 238 500 µL microtainer EDTA tubes (Becton Dickinson) using a syringe. Mixed umbilical cord 52 239 blood will be collected into another 3 mL and 6 mL EDTA tube (the 3 mL tube will be stored 53 240 at room temperature for subsequent flow cytometric analysis) and a 6 mL Naheparin trace 54 241 elementfree tube (Greiner bioone). Midwives will follow WHO guidelines with regards to 55 56 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 242 umbilical cord clamping, that is, cord clamping will be carried out more than one minute after 4 243 the birth or when cord pulsation has ceased. 5 244 6 245 Venous blood will be collected from the infants by the study nurses into 3 mL EDTA tubes 7 246 (Becton Dickinson) using a Microneedle (Sarstedt, Numbrecht, Germany) or Eclipse Signal 8 247 Blood Collection needle (Becton Dickinson). Blood will be drawn from the dorsal side of the 9 248 hand at 48 h, 1 month, 4 months and from the hand or arm at 12 months and 48 months of 10 249 age. 11 12 250 Urine samples 13 251 Infant’s urine at 4 months of age will be collected in a urine collection bag placed around the 14 252 penis or vagina under the infant’s nappy at the start of a study visit, with the bag being 15 253 collected at the end of the study visit. Urine samples at 4 yearsofage will be collected in a 16 For peer review only 17 254 potty lined with a plastic bag. Midstream urine samples, taken at a random time point during 18 255 the day, will be collected from adults. 19 20 256 Placenta 21 257 After delivery, the placenta will be examined according to local guidelines by the delivery 22 258 personal and placed in a plastic bag and stored at 4°C. The placenta will be collected by the 23 259 study nurses and processed at the laboratory according to the outline in figure 2. 24 25 260 Breast milk 26 261 Mothers will be asked to express 30 mL of breast milk in the morning before the second 27 262 breastfeeding occasion of the day, at 1 and 4 months postpartum. Samples will be collected 28 263 in a clean bowl or pump and poured into plastic centrifuge tubes for immediate storage at 29 264 20°C at home. Mothers will be requested to bring the frozen tubes with them at the 4 months 30 265 followup and reminded to keep the samples in a cooling bag or wrapped in paper with a 31 266 cooling block to prevent thawing of the samples during transport to the hospital. The samples 32 267 will be transferred directly to a 80°C freezer on arrival to the research laboratory. If thawing http://bmjopen.bmj.com/ 33 268 appears to have occurred during transport, this will be noted in the sample preparation 34 269 protocol. 35 36 270 Meconium 37 271 Meconium will be sampled when voided using a sterile spatula and put in a sterile container 38 39 272 that will be immediately stored at 4°C until aliquoting at the research laboratory. 40 41 273 Faeces on September 28, 2021 by guest. Protected copyright. 42 274 Faecal samples will be collected from the infant’s nappy by the parents at home using a sterile 43 275 spatula. Faeces will be put in sterile tubes prelabelled with the child’s study ID. Parents will 44 276 be requested to write the date of sampling on the tubes and then place the samples in the 45 277 freezer as soon as possible and to bring it to the hospital at the next followup. Participants 46 278 will be requested to keep the samples in a cooling bag or wrapped in newspaper to prevent 47 279 thawing of the samples during transport to the hospital. The stool samples will be transferred 48 280 directly to a 80°C freezer on arrival to the research laboratory. If thawing appears to have 49 281 occurred during transport, this will be noted in the sample preparation protocol. 50 51 282 Saliva 52 283 Approximately 1.52 mL saliva will be collected from mothers, by spitting into a plastic tube. 53 284 Infant saliva (approximately 0.5 mL) will be collected from the buccal cavity using a plastic 54 285 Pasteur pipette. 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 286 Buccal swabs 4 287 Buccal swabs (eSwab, Copan, Bresia, Italy) will be collected by swiping the swab tip pass 5 288 beneath the upper lip, and the inside of the cheeks, taking care to avoid touching the teeth. 6 289 The swabs will be stored at 4°C until centrifugation at 12,000 g for 10 min. Excess fluid will 7 290 be removed from the bacterial pellet before storage. 8 9 291 Hair specimens 10 292 At 48 months hair specimens will be cut by stainless steel scissors from the occipital part of 11 293 the head, beneath some covering hair, as close to the scalp as possible. 12 13 14 294 Handling, transport and storage of biological samples 15 295 At gestational week 28, samples will be collected at the maternity clinics. The RNA 16 296 Tempus™ bloodFor samples andpeer urine samples review will be stored onlyimmediately at 4°C, whilst blood 17 297 samples will be kept at room temperature for 30 min before centrifugation. All samples will 18 298 be stored at 4°C until transport and will be subsequently transported at 410°C to the hospital 19 20 299 laboratory, either the same day or the following workday. 21 300 22 301 Samples taken at delivery and at 48 h will be stored at 4°C and processed within 212 h if 23 302 sampled on weekdays and within 1265 h if sampled on weekends. Samples taken at the 24 303 followup visits at 1, 4, 12 and 48 months of age will be collected by the NICEcohort study 25 304 nurses and handled the same work day. All biological samples will stand for 1530 min in 26 305 room temperature until centrifugation of the blood. After centrifugation of the blood, all 27 306 samples will be stored at 4°C until aliquoting (performed on a cooling block) and interim 28 307 storage at 20°C (for a maximum of 12 weeks). Once filled, storage boxes will be transferred 29 308 to longterm storage at 80°C. Hair samples will be collected in paper envelopes and stored at 30 309 room temperature. A laboratory protocol will be filled in for all samples where all transport 31 310 and preparation times will be noted in order to be able to account for the time from sampling 32

311 to freezer in all downstream analyses. http://bmjopen.bmj.com/ 33 34 35 312 Collection of participant data 36 37 313 All study participants will be closely followedup during pregnancy from inclusion until the 38 314 delivery from physician’s notes in their medical records. Data from the medical pregnancy 39 315 records will be analysed and collected in the database. 40 316 41 317 At inclusion, both parents will be asked to fill out one online questionnaire each regarding on September 28, 2021 by guest. Protected copyright. 42 318 allergic heredity and environment. When the infant has reached 18 months of age the second 43 319 online questionnaire, the “Infant environment” questionnaire, regarding the infant’s current 44 320 living environment and lifestyle will be sent. The third online questionnaire, the “Child 45 321 environment” questionnaire will be sent out to the participants when the infant has reached 48 46 322 months of age and concerns the infant’s environment and lifestyle (Table 2 and 3). 47 48 323 Assessment of dietary intake in mothers and children 49 324 The mothers will receive a webbased, food frequency questionnaire (MealQ) by email 50 325 during pregnancy in gestational week 34 and at 1 and 4 months postpartum, consisting of 51 52 326 questions regarding their diet during the previous month [20, 21] (Table 2 and 4). Reminders 53 327 will be sent out by email one and two weeks after the initial email and, alongside the last 54 328 reminder, a text message will be sent to their cell phones. The questionnaires will have a 55 329 mealbased and interactive format, i.e. only those who report a relevant frequency of 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 330 consumption of a certain food or food group will receive followup questions, to promote ease 4 331 of use and minimize the time required to answer the questionnaire. Depending on the number 5 332 of followup questions, the MealQ questionnaire contains 102174 questions. 6 333 7 334 When the child reaches 12 months and 48 months of age, the parents fill out a food frequency 8 335 questionnaire about the infant’s food habits (TodMealQ and KidMealQ, respectively), 9 336 which will be similar to the MealQ but adopted for young children [22]. Parents from 10 337 families where both parents have attended the followup visits will be asked to participate in a 11 338 subcohort with extended sampling, including a MealQ questionnaire for each of the parents 12 339 at 12 months of age of the child. 13 14 340 15 341 To measure total energy expenditure, all dietary questionnaires will be combined with the 16 342 activity questionnairesFor ActiveQ peer [23] or KidActiveQreview [24]. only 17 343 18 344 The questionnaires MealQ and KidMealQ have been validated in their original form [2022] 19 345 but all dietary questionnaires were modified to better fit the specific needs of the study of 20 346 evaluation of fat quality, consumption of fish from polluted areas, sugar and probiotic use. 21 347 Modifications are detailed in Supplemental file 1. Nutritional composition of the diet will be 22 348 calculated from the dietary questionnaires using the database of nutrient content published by 23 349 the Swedish National Food Agency. The nutrient conversion for the questionnaires will be 24 350 calculated by computer programs developed specifically for those questionnaires. 25 26 351 Diary for recoding introduction of complementary foods to the infant diet: 27 352 Practices of introduction of complementary foods and formulas to the infant’s diet will be 28 353 assessed each month during the first 12 months of life by a webbased questionnaire; 29 30 354 “Infections and feeding” (Table 2). The questions include the infant’s intake of breast milk 31 355 from breast or bottle, and whether certain foods have been introduced, including formula 32 356 (percentages of total feeding and brand name), gruel (regular, with/without gluten or lactose, 33 357 soya based), porridge (regular, without gluten or lactose) and different food groups http://bmjopen.bmj.com/ 34 358 (potatoes/roots, fruit/juice, berries, nuts/almonds, peanuts/peanut oil, bread/crackers/cookies, 35 359 butter, margarine, oil, milk, icecream, sour milk, yoghurt, meat, fish and egg), as well as if 36 360 there are any food the child has started to eat more than twice per week that is not included 37 361 above. Additional questions will be asked about probiotics, vitamin D, omega3 or other 38 362 supplements and frequency of canned versus homemade food. 39 363 40 364 Parents will be asked to estimate the proportion of commercial versus homemade baby food. on September 28, 2021 by guest. Protected copyright. 41 365 In order to be able to adjust the statistical analyses for reverse causation, the parents will be 42 366 also asked the questions: ‘during the past month, have you actively avoided to give any type 43 44 367 of food due to concern of allergy?’ and ‘if so what food items have been excluded from the 45 368 child’s diet?’. 46 47 369 Outcome measures 48 49 370 Infections during year one 50 371 Infectious episodes will be recorded monthly by the parents during the first 12 months using 51 372 the abovementioned webbased questionnaire. If an infectious episode is recorded, followup 52 373 questions will be posed concerning duration of the illness, body temperature, breathing 53 374 problems, antibiotic treatment, medications, hospitalisation etc. 54 55 56 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 375 Lung- function test 4 376 Lungfunction will be assessed at 4 and 12 months, using respiratory resistance with 5 377 interrupter technique (MicroRint, MicroMedical Ltd, UK). The method estimates alveolar 6 378 pressure (Palv) by performing a rapid occlusion of the airways during the respiratory circle at 7 379 rest, assuming that Palv will equal mouth pressure (Pmo). Respiratory resistance with 8 380 interrupter technique (RInt) is a noninvasive method that requires minimal cooperation and 9 381 can be used in evaluating airway obstruction in uncooperative and small children [25, 26]. At 10 382 48 months lung function will be measured by standardized Impulse Oscillometry. 11 12 383 Skin prick test 13 384 Sensitization to common allergens will be assessed by skin prick testing at 12 and 48 months 14 385 ofage. Testing will be performed on the volar side of the lower arm and will be carried out 15 386 according to the standards of the Subcommittee on Skin Tests of the European Academy of 16 387 Allergy and ClinicalFor Immunology peer [27]. Sixreview different standardized only allergens (egg, milk, birch, 17 18 388 timothy, cat and dog) will be applied with a lancet; histamine dihydrochloride (10 mg/mL) 19 389 will be used as positive control (Soluoprick, ALK Nordic A/S, Denmark). The test will be 20 390 considered positive for a particular allergen if the mean wheal diameter is at least 3 mm after 21 391 15 min. 22 23 392 Allergy diagnosis 24 393 Allergy will be diagnosed at 12 and 48 months by a specialist in paediatric allergology. The 25 394 diagnosis will be based on a clinical history and a clinical examination using the same 26 395 methodology as in the BAS cohort [28], enabling direct comparisons. The clinical protocol 27 396 will be carefully standardized and complies with the diagnostic protocols of our previous 28 397 ALLERGYFLORA study [29] and FARMFLORA [4] study. 29 398 30 399 The following diagnostic criteria will be used for clinical allergic diagnosis at 12 and 48 31 400 months of age: 32 http://bmjopen.bmj.com/ 33 401 Atopic eczema, atopic dermatitis (AD): 34 402 Various diagnostic criteria for AD have been proposed and validated. For the NICE cohort 35 403 followup at 12 and 48 months we choose to employ the simplified criteria proposed by 36 37 404 Williams et al. with high sensitivity and specificity [3032]. 38 405  major criteria (=must have): an itchy condition or parental report of 39 406 scratching/rubbing by the child 40 407  plus three or more of the following criteria: 1) personal history of itchiness in skin 41 408 creases such as cheeks, folds of the elbows, behind knees, front of ankles or around on September 28, 2021 by guest. Protected copyright. 42 409 the neck 2) personal history of asthma or hayfever or other atopic disease or a history 43 410 of atopic disease among a firstdegree relatives 3) a personal history of generally dry 44 411 skin in the previous year 4) visible eczema on the cheeks, folds of the elbows, behind 45 412 knees, front of ankles or around neck 5) a debut before 24 months of age. 46 47 413 Food allergy 48 414 A diagnosis of food allergy requires: a medical history of allergic reactions, on at least two 49 415 consecutive occasions when eating a specific food allergen. The allergic reaction disappears 50 416 on elimination of the allergen. For early food reaction, allergic symptoms will appear <2h 51 417 with rush, itching, urticaria, angioedema, gastrointestinal pain, vomiting, oral symptoms, 52 53 418 asthma or anaphylactic reaction. Late food reaction (>2h) will present eczema, 54 419 gastrointestinal reactions with pain, diarrhoea, obstipation and weight loss. For all suspected 55 420 cases of food allergy, the diagnosis will be confirmed by eliminationprovocation, i.e. the 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 421 child will be given the suspected food allergen and symptoms will be recorded. The 4 422 eliminationprovocation procedure will be carried out at home following local guidelines if 5 423 the symptoms are mild. If this is not feasible, or if the case is unclear, the child will be 6 424 admitted to the paediatric ward at the Sunderby Hospital for repeated open food provocation 7 425 until the diagnosis is clarified (food allergic, or healthy). Open provocation will start with a 8 426 dose determined by the individual child’s food allergen reaction and will be followed by 9 427 increasing doses of specific food allergen given with 30 min interval in 35 steps. The child 10 428 will stay at the ward for 2 hours after the last dose and parents will thereafter report any 11 429 symptoms during the following days. 12 13 430 Allergy to animals 14 431 A diagnosis of allergy to animal dander requires: a medical history of symptoms from the 15 432 eyes and/or nose and/or asthma on at least two consecutive occasions when exposed to a 16 For peer review only 17 433 specific animal dander (e.g. dog, cat, rabbit, horse). The reaction disappears when the animal 18 434 allergen is removed. 19 20 435 Pollen allergy 21 436 A diagnosis of pollen allergy requires: a medical history of symptoms from eyes and/or nose 22 437 and/or asthma on two consecutive occasions when exposed during pollen season. The reaction 23 438 disappears in areas with lower pollen content or following medication. 24 25 439 Asthma 26 440 At 12 months: An asthma diagnosis requires at least one of the following: 1) The presence of 27 441 wheeze between infections 2) Persistent wheeze for at least 4 weeks 3) At least one period of 28 442 wheeze during an infectious disease with concomitant allergic disease (AD, food allergy or 29 443 allergic rhinoconjunctivitis) or 4) Three episodes of wheeze during infectious disease, with 30 444 no concomitant allergic disease. 31 445 At 48 months: An asthma diagnosis requires at least one of the following: The presence of 1) 32 446 wheeze within the last 12 months (during or between infections) or 2) ongoing controller http://bmjopen.bmj.com/ 33 447 medication. 34 448 If there is any doubt concerning the asthma diagnosis at 12 or 48 months, the child will be 35 36 449 admitted to the children’s clinic at the Sunderby Hospital for a second clinical followup for 37 450 doctor´s diagnosed asthma. 38 39 451 Neuropsychological and neurobehavioral examination 40 452 At 48 months of age the parents will be asked to answer questionnaires and the child will 41 453 undergo examination with focus on neuropsychological behaviour. The questionnaires will on September 28, 2021 by guest. Protected copyright. 42 454 consist of parentreported Child Behaviour Checklist (CBCL), Adaptive Behaviour 43 455 Assessment System (ABAS) and Social Responsiveness Scale (SRS). Cognitive development 44 456 will be tested facetoface with Wechsler Preschool and Primary Scale of Intelligence 45 457 (WPPSI) and a physical assessment of motor skills with Peabody Development Motor Scales 46 458 (PDMS; by a physiotherapist). 47 48 459 Oral health examination. 49 460 Caries will be recorded in children at 48 months of age by one to three experienced and 50 461 calibrated dentists (intra and interexaminer kappa ≥ 0.98) using a mirror and probe as the 51 462 number of decayed (enamel caries included), filled surfaces in the primary dentition (defs) 52 463 [1719]. At the same visit, caries will also be recorded for the parents using also two bitewing 53 464 radiographs as the number of decayed (enamel caries included), filled surfaces in the 54 55 465 permanent dentition (DeFS). Whole saliva and buccal swabs will be collected from the 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 466 parents. Complementary questionnaires on oral behavior data, such as oral hygiene, intake 4 467 frequency of sweets and sugary drinks, use of fluorides and oral hygiene, will be taken for 5 468 children and parents. Enamel discolorations and irregularities on visual inspection and 6 469 probing will registered and exfoliated primary teeth will be collected by parents in prepared 7 470 tubes. 8 9 10 471 Analyses 11 12 472 Immune analysis 13 473 Blood samples will be analysed by flow cytometry regarding numbers and percentages of T 14 474 and B cells and their subsets and their markers of naivety, activation and memory [33, 34]. A 15 475 saliva sample will be collected from the mother close to the time of birth and from the infant 16 476 at 4 months for theFor analysis peer of secretory IgAreview and salivary microbiota only [35]. 17 18 477 Genetic analyses 19 478 Genomic DNA will be extracted from blood collected in EDTA tubes from both mothers, 20 479 fathers and infants for analysis of genetic variation. Blood will also be collected from the 21 480 mothers in Tempus™ Blood RNA Tubes for stabilization and isolation of total RNA from 22 481 whole blood for gene expression analysis. 23 24 482 Metabolomics 25 483 To study metabolic alterations in the participants, nontargeted metabolomics profiling will be 26 484 employed on plasma samples by liquid chromatography mass spectrometry (LCMS) and gas 27 485 chromatography mass spectrometry (GSMS) analysis [36]. 28 29 486 Transcriptomics 30 487 Whole transcriptome analysis will be performed on placenta tissue from those placentas 31 488 sampled within 4 h from birth and on maternal blood collected in Tempus™ RNA stabiliser 32

489 tubes at gestation, at birth and during lactation. http://bmjopen.bmj.com/ 33 34 490 Total RNA will be analysed and its quality will be checked using a bioanalyzer. RNA 35 491 microarray analysis will be performed by next generation sequencing and the raw data will be 36 492 analysed using R biostatistical computing platform RStudio GUI. Gene set enrichment 37 493 analysis (GSA) will be performed to visualize regulated biological processes [37]. 38 39 494 Epigenetics 40 495 Wholeepigenome genespecific methylation analysis (Illumina EPIC) will be performed on 41 496 selected samples of the placenta, cord blood, and blood of children at 48 months of age to on September 28, 2021 by guest. Protected copyright. 42 497 identify persisting genespecific changes in DNA methylation. Significant DNA methylation 43 498 markers will be validated by pyrosequencing in all children [38, 39]. 44 45 499 Fatty acid and vitamin A and D composition in blood cell membranes and plasma 46 500 Blood cell fatty acid composition will be analysed in unwashed red and white blood cells. 47 501 Fatty acids from blood cells will be extracted [40] and analysed using gas chromatography 48 502 flame ionisation (GCFID) [41]. Concentrations of vitamin A and D will be determined in 49 503 plasma with LCMS/MS [42, 43]. 50 51 504 Determination of toxic metals, other elements and micronutrients 52 505 Concentration of toxic and essential elements will be measured in serum (Ca, Mg, Zn, Se), 53 506 erythrocytes (Cd, Mn, MeHg, Pb) and urine (As, Cd, U and I) using Inductively Coupled 54 55 507 Plasma Mass Spectrometry (ICPMS), developed for highthroughput multiple element 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 508 analyses [4446]. Speciation of urinary As (metabolites of inorganic arsenic) will be 4 509 performed by HPLCICPMS. Urinary F will be measured using an ionselective electrode 5 510 (ISE). 6 7 511 Oral and gut microbiota analyses 8 512 Infant faeces, oral swab and saliva will be collected for later analysis of microbiota 9 513 composition using DNA based sequencing methods [47]. 10 11 12 514 Data analysis 13 515 Most of the questionnaires will be webbased and data will be downloaded directly form the 14 516 surveys to SPSS files. All data will be cleaned and entered into a secure database stored 15 517 within the hospital domain and handled and protected with the same confidentiality as patient 16 518 journal data. AFor statistician peer at the Unitreview for Research andonly Innovation, Department of 17 519 Development, Region Norrbotten, will be responsible for database management. Requests 18 19 520 from study researchers for analysis of cohort data require detailed analysis plan is approved 20 521 by the steering committee. All data transferred to the Universities will be blinded so that 21 522 researchers cannot link individual data to any specific individual 22 523 23 24 524 Statistical analyses 25 525 Univariate and multivariate analysis and data mining, including cluster and hierarchical 26 526 models, will be used, depending on data and outcome measures. To examine general 27 527 maternal, paternal and infant demographic data and distribution of outcomes, descriptive 28 528 analysis will be performed initially. Statistical methods linking outcomes to collected or 29 529 measured data will depend on the of the data: Chi2 tests will be used for categorical 30 530 data, Student’s ttests and ANOVA (and corresponding nonparametric methods when 31 531 applicable, such as MannWhitney U and KruskallWallis tests) for continuous data as well as 32 532 Pearson and Spearman correlations. Questionnaire data will be analyzed by factor analysis to http://bmjopen.bmj.com/ 33 533 investigate interactions between questionnaire variables and identify potential constructs. 34 35 534 Similar to PCA, factors/constructs can be interpreted as latent variables underlying observed 36 535 total data variability. To identify independent predictors of outcomes among questionnaire 37 536 and demographic data (both original data and factor scores), generalized linear mixed models 38 537 (ANOVA, logistic regressions and conditional logistic regressions) capable of controlling for 39 538 possible confounders will be applied. 40 539 41 540 Metabolomics data will be generated and preprocessed at the Chalmers Mass Spectrometry on September 28, 2021 by guest. Protected copyright. 42 541 Infrastructure (www.chalmers.se/en/researchinfrastructure/CMSI/Pages/default.aspx). To 43 542 identify molecular predictors related to outcomes, metabolomics data will be analyzed using 44 543 predominantly supervised multivariate methods. We will employ partial least squares (PLS) 45 544 and random forest techniques incorporated into nested crossvalidation frameworks to 46 545 effectively manage the highdimensional nature of untargeted metabolomics data. These new 47 546 techniques also perform automatic, unbiased variable selection, suitable for identifying 48 49 547 biomarkers associated with outcome variables. The combined nested crossvalidation with 50 548 unbiased variable selection greatly reduces the risk of false positive findings while improving 51 549 modelling accuracy [48]. Outputs from multivariate models can, as mentioned above, be 52 550 incorporated into mixed models to be able to adjust for confounders in associations between 53 551 molecular predictors and outcomes. 54 552 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 553 Statistical analyses will be performed in two steps: pilot analyses and final analyses. Pilot 4 554 analyses refer to analysis of a subset of samples collected during 20152016, and final 5 555 analyses refer to data from samples collected from the whole study (20152018). 6 556 7 8 9 557 Patient and Public Involvement 10 558 The study was based on healthy volunteers. Patients and public were neither involved in the 11 559 design of the study nor in the development of the research questions or outcome measures. 12 560 Neither were patients and public involved in the recruitment or conduction of the study. The 13 561 results will be disseminated to the study participants on a cohort basis on yearly workshops. 14 562 No individual results will be given to the participants. 15 16 For peer review only 17 563 NICE consortium 18 19 20 564 Steering committee 21 565 The Steering Committee, consisting of MD, PhD, PI Anna Sandin, Professor AnnSofie 22 23 566 Sandberg and Professor Agnes Wold, is responsible for administrative and financial 24 567 coordination and monitoring the progress of the cohort. The Steering Committee will meet 25 568 twice a year to handle the allocation of resources to the research projects, the recruitment of 26 569 personnel and the scientific direction. 27 28 570 Research groups 29 30 571 Research groups involved in the NICEcohort are shown in table 5. 31 32 33 572 Collaborators http://bmjopen.bmj.com/ 34 573 Within the NICEcohort there are close collaborators with a range of expertise. For 35 574 transcriptomics analysis and systems biology of the NICE samples we have a close 36 575 collaboration with Professor Jens Nilsen at Systems Biology at the Department of Biology 37 576 and Biological Engineering at Chalmers University of Technology in Gothenburg, Sweden. 38 577 Robert Lundqvist, statistician at the Research and Innovation Unit at Region Norrbotten, will 39 578 advise on statistical analyses and will be also responsible for the management of the NICE 40 41 579 cohort database. on September 28, 2021 by guest. Protected copyright. 42 43 580 Ethics and dissemination 44 581 The study will be conducted in accordance with the Helsinki Declaration and has been 45 46 582 approved by the Regional Ethical Review Board in Umeå, Sweden. The participants will be 47 583 informed about the aims and the requirements of the NICEcohort, and will be provided both 48 584 written and verbal information. They will be requested to complete and sign a consent form 49 585 before being enrolled in the study. Once the infant is born, the parents will again requested to 50 586 give consent for their child’s participation in the study. All participants have the right to 51 587 withdraw from the study at any point and have their data removed from all study 52 588 documentation. 53 589 54 590 All data will be securely stored according to European laws and password protected and 55 591 accessible by the research team only. All personal data will be blinded so that researchers 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 592 cannot link individual data to any specific individual. Information resulting from the study 4 593 will be disseminated to nonscientific audiences in the region where the study has been 5 594 carried out to keep health professionals and interested members of the public abreast of the 6 595 findings. Results from the cohort will be disseminated through peer reviewed journals, 7 596 preferably open access journals to ensure that all readers from the scientific society, the 8 597 clinical professionals as well as the general public will have open access to the publication. 9 598 Results will also be communicated by the researchers on scientific conferences. 10 11 12 599 Authors’ contributions 13 600 A. Sandin, AS. Sandberg and A. E. Wold initiated the study and are responsible for the 14 601 conception and design of the work. All authors have been involved in the design of the cohort 15 602 and is involved in data analysis and interpretation. F. Murray and A. Sandin are responsible 16 603 for data collection.For M. Barman peer and A. reviewRoss drafted the first only version of the manuscript. K. 17 604 Broberg, K. Jonsson, M. Kippler, N. Strömberg, M. Vahter, A. E. Wold, AS. Sandberg and 18 19 605 A. Sandin was involved in writing different parts of the manuscript. AI. Bernardi, S. Bölte, 20 606 B. Hesselmar, B. Jacobsson, H. Rabe, and F. Sjöberg contributed to the critical revision of the 21 607 manuscript. M Barman was responsible for finalising the last version of the manuscript. All 22 608 authors read and approved the manuscript before submission. 23 24 25 609 Acknowledgements 26 610 We would like to acknowledge study nurses Marjut Larsson, Ulrika Börlin and study assistant 27 611 nurse Emma LundinRåberger, study midwifes Louise Lindgren and Lisa Sundén and study 28 612 administrator Elvira Sandin for their fantastic work with recruitment, follow up assessments, 29 613 sample preparation and data collection. We also acknowledge Staffan Nilsson, associate 30 614 professor in mathematical statistics, Mathematical Sciences, Chalmers University of 31 615 Technology, for performing the power calculations and Carl Brunius, associate professor in 32 616 Food and Nutrition Science, Department of Biology and Biological Engineering, Chalmers http://bmjopen.bmj.com/ 33 617 University of Technology for assistance with the section on data analysis. We also thank all 34 618 personnel at the maternity clinics, antenatal clinics and delivery ward for informing and 35 619 recruiting patients and collecting samples. We deeply thank all the families who participate in 36 37 620 this study. We gratefully acknowledge our collaborators: Professor Jens Nilsen, at Systems 38 621 Biology at the Department of Biology and Biological Engineering at Chalmers University of 39 622 Technology in Gothenburg, Sweden; Robert Lundqvist, statistician at the Research and 40 623 Innovation Unit at Region Norrbotten; Susanne Lager, researcher, PhD, Department of 41 624 Women's and Children's Health, Uppsala University; Linda EnglundÖgge, MD, PhD, on September 28, 2021 by guest. Protected copyright. 42 625 Department of Obstetrics and Gynaecology, Sahlgrenska University Hospital, Gothenburg, 43 626 Sweden; Karin Larsson, researcher, PhD, Food and Nutrition Science, Departments of 44 627 Biology and Biological Engineering, Chalmers University of Technology, Sweden; and Maria 45 628 Hallingström, Midwife, Department of Obstetrics and Gynaecology, Sahlgrenska University 46 629 Hospital, Gothenburg, Sweden. 47 48 49 630 Funding 50 631 Funding has been received from the Swedish Research Council (VR) project no 5212013 51 632 3154, Swedish Research Council for Health, Working Life and Welfare (FORTE) project no 52 633 20140923, the Västra Götaland Region RUN 612061815, Åke Wibergs Stiftelse, Jane and 53 634 Dan Olsson Foundation, the Research and Innovation Unit at Region Norrbotten and through 54 55 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 635 a regional agreement between Umeå University and Västerbotten County Council in the field 4 636 of Medicine, Odontology, and Health (711251). 5 6 637 Competing interests 7 8 638 No competing interests are declared. 9 10 11 639 Patient consent 12 640 Patient consent will be obtained for participation in this study, and in the case of minors, 13 641 consent will be given by the parents 14 15 16 642 Ethics approvalFor peer review only 17 643 The NICEcohort is approved by the Regional Ethical Review board in Umeå, Dept. for 18 644 Medical Research, Sweden: 2013/1831M, supplementary applications: 201428132M, 2015 19 645 7132M, 201530432M and 201623232M. 20 21 22 646 Provenance and peer review 23 24 647 This work has not been commissioned, and has been externally peerreviewed. 25 26 648 Open access 27 28 649 This is an Open Access article distributed in accordance with the terms of the Creative 29 650 Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt 30 651 and build upon this work, for commercial use, provided the original work is properly cited. 31 652 See: http://creativecommons.org/licenses/by/4.0/ 32 33 http://bmjopen.bmj.com/ 34 653 References 35 654 1. Berggren, S., et al., Parents with overweight children two and five years of age did not 36 655 perceive them as weighing too much. Acta Paediatr, 2017. apa.14174. 37 656 2. Wang, M., et al., Reduced diversity in the early fecal microbiota of infants with atopic 38 657 eczema. J Allergy Clin Immunol, 2008. 121(1): p. 12934. 39 658 3. Sandin, A., et al., Faecal short chain fatty acid pattern and allergy in early childhood. 40 659 Acta Paediatr, 2009. 98(5): p. 8237. on September 28, 2021 by guest. Protected copyright. 41 660 4. Jonsson, K., et al., Fat intake and breast milk fatty acid composition in farming and 42 661 nonfarming women and allergy development in the offspring. Pediatr Res, 2016. 79(1 43 44 662 1): p. 11423. 45 663 5. Dietert, R.R., Developmental Immunotoxicity, Perinatal Programming, and 46 664 Noncommunicable Diseases: Focus on Human Studies. Adv Med, 2014. 2014: p. 47 665 867805. 48 666 6. Ahmed, S., et al., Arsenic exposure and cell-mediated immunity in pre-school children 49 667 in rural Bangladesh. Toxicol Sci, 2014. 141(1): p. 16675. 50 668 7. Barman, M., et al., High levels of both n-3 and n-6 long-chain polyunsaturated fatty 51 669 acids in cord serum phospholipids predict allergy development. PLoS One, 2013. 52 670 8(7): p. e67920. 53 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 671 8. Gilman, S.E., et al., Socioeconomic disadvantage, gestational immune activity, and 4 672 neurodevelopment in early childhood. Proc Natl Acad Sci U S A, 2017. 114(26): p. 5 673 67286733. 6 674 9. Grether, J.K., et al., Prenatal and Newborn Immunoglobulin Levels from Mother- 7 675 Child Pairs and Risk of Autism Spectrum Disorders. Front Neurosci, 2016. 10(1662 8 676 4548): p. 218. 9 677 10. Meltzer, A. and J. Van de Water, The Role of the Immune System in Autism Spectrum 10 678 Disorder. Neuropsychopharmacology, 2017. 42(1): p. 284298. 11 679 11. Vuong, H.E. and E.Y. Hsiao, Emerging Roles for the Gut Microbiome in Autism 12 680 Spectrum Disorder. Biol Psychiatry, 2017. 81(5): p. 411423. 13 14 681 12. Gardner, R.M., et al., Environmental exposure to metals and children's growth to age 15 682 5 years: a prospective cohort study. Am J Epidemiol, 2013. 177(12): p. 135667. 16 683 13. Kippler, ForM., et al., Early-lifepeer cadmium review exposure and childonly development in 5-year-old 17 684 girls and boys: a cohort study in rural Bangladesh. Environ Health Perspect, 2012. 18 685 120(10): p. 14628. 19 686 14. Hamadani, J.D., et al., Critical windows of exposure for arsenic-associated 20 687 impairment of cognitive function in pre-school girls and boys: a population-based 21 688 cohort study. Int J Epidemiol, 2011. 40(6): p. 1593604. 22 689 15. Skroder, H., et al., Early-Life Selenium Status and Cognitive Function at 5 and 10 23 690 Years of Age in Bangladeshi Children. Environ Health Perspect, 2017. 125(1552 24 691 9924). 25 692 16. Collaboration, N.C.D.R.F., Worldwide trends in body-mass index, underweight, 26 27 693 overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population- 28 694 based measurement studies in 128.9 million children, adolescents, and adults. Lancet, 29 695 2017. 390(10113): p. 26272642. 30 696 17. Kallestal, C., The effect of five years' implementation of caries-preventive methods in 31 697 Swedish high-risk adolescents. Caries Res, 2005. 39(1): p. 206. 32 698 18. Stromberg, N., et al., Genetic- and Lifestyle-dependent Dental Caries Defined by the 33 699 Acidic Proline-rich Protein Genes PRH1 and PRH2. EBioMedicine, 2017. http://bmjopen.bmj.com/ 34 700 26(Supplement C): p. 3846. 35 701 19. Esberg, A., et al., Streptococcus Mutans Adhesin Biotypes that Match and Predict 36 702 Individual Caries Development. EBioMedicine, 2017. 24: p. 205215. 37 703 20. Christensen, S.E., et al., Two new meal- and web-based interactive food frequency 38 704 questionnaires: validation of energy and macronutrient intake. J Med Internet Res, 39 705 2013. 15(6): p. e109. 40 41 706 21. Christensen, S.E., et al., Relative validity of micronutrient and fiber intake assessed on September 28, 2021 by guest. Protected copyright. 42 707 with two new interactive meal- and Web-based food frequency questionnaires. J Med 43 708 Internet Res, 2014. 16(2): p. e59. 44 709 22. Delisle Nystrom, C., et al., Validation of an Online Food Frequency Questionnaire 45 710 against Doubly Labelled Water and 24 h Dietary Recalls in Pre-School Children. 46 711 Nutrients, 2017. 9(1). 47 712 23. Bonn, S.E., et al., Active-Q: validation of the web-based physical activity 48 713 questionnaire using doubly labeled water. J Med Internet Res, 2012. 14(1): p. e29. 49 714 24. Bonn, S.E., et al., Feasibility of a novel web-based physical activity questionnaire for 50 715 young children. Pediatr Rep, 2012. 4(4): p. e37. 51 716 25. Zuriarrain, R., et al., Assessing the measurement of airway resistance by the 52 717 interrupter technique. Arch Argent Pediatr, 2013. 111(6): p. 495501. 53 54 718 26. Rech, V.V., et al., Airway resistance in children measured using the interrupter 55 719 technique: reference values. J Bras Pneumol, 2008. 34(18063756). 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 720 27. Skin tests used in type I allergy testing Position paper. Sub-Committee on Skin Tests of 4 721 the European Academy of Allergology and Clinical Immunology. Allergy, 1989. 5 722 44(Suppl 10:159.). 6 723 28. Sandin, A., B. Bjorksten, and L. Braback, Development of atopy and wheezing 7 724 symptoms in relation to heredity and early pet keeping in a Swedish birth cohort. 8 725 Pediatr Allergy Immunol, 2004. 15(4): p. 31622. 9 726 29. Hesselmar, B., et al., Early fish introduction is associated with less eczema, but not 10 727 sensitization, in infants. Acta Paediatr, 2010. 99(12): p. 18617. 11 728 30. Williams, H.C., et al., The U.K. Working Party's Diagnostic Criteria for Atopic 12 729 Dermatitis. I. Derivation of a minimum set of discriminators for atopic dermatitis. Br J 13 14 730 Dermatol, 1994. 131(3): p. 38396. 15 731 31. Williams, H.C., et al., The U.K. Working Party's Diagnostic Criteria for Atopic 16 732 Dermatitis.For III. Independent peer hospital review validation. Br J onlyDermatol, 1994. 131(3): p. 406 17 733 16. 18 734 32. Williams, H.C., et al., The U.K. Working Party's Diagnostic Criteria for Atopic 19 735 Dermatitis. II. Observer variation of clinical diagnosis and signs of atopic dermatitis. 20 736 Br J Dermatol, 1994. 131(3): p. 397405. 21 737 33. Lundell, A.C., et al., High proportion of CD5+ B cells in infants predicts development 22 738 of allergic disease. J Immunol, 2014. 193(2): p. 5108. 23 739 34. Strombeck, A., et al., High proportions of FOXP3(+) CD25(high) T cells in neonates 24 740 are positively associated with allergic sensitization later in childhood. Clin Exp 25 741 Allergy, 2014. 44(7): p. 94052. 26 27 742 35. Sandin, A., et al., High salivary secretory IgA antibody levels are associated with less 28 743 late-onset wheezing in IgE-sensitized infants. Pediatr Allergy Immunol, 2011. 22(5): 29 744 p. 47781. 30 745 36. Savolainen, O.I., A.S. Sandberg, and A.B. Ross, A Simultaneous Metabolic Profiling 31 746 and Quantitative Multimetabolite Metabolomic Method for Human Plasma Using 32 747 Gas-Chromatography Tandem Mass Spectrometry. J Proteome Res, 2016. 15(1): p. 33 748 25965. http://bmjopen.bmj.com/ 34 749 37. Soni, N.K., et al., Splenic Immune Response Is Down-Regulated in C57BL/6J Mice 35 750 Fed Eicosapentaenoic Acid and Docosahexaenoic Acid Enriched High Fat Diet. 36 751 Nutrients, 2017. 9(1). 37 752 38. Broberg, K., et al., Arsenic exposure in early pregnancy alters genome-wide DNA 38 753 methylation in cord blood, particularly in boys. J Dev Orig Health Dis, 2014. 5(4): p. 39 754 28898. 40 41 755 39. Kippler, M., et al., Sex-specific effects of early life cadmium exposure on DNA on September 28, 2021 by guest. Protected copyright. 42 756 methylation and implications for birth weight. Epigenetics, 2013. 8(5): p. 494503. 43 757 40. Masood, A., K.D. Stark, and N. Salem, Jr., A simplified and efficient method for the 44 758 analysis of fatty acid methyl esters suitable for large clinical studies. J Lipid Res, 45 759 2005. 46(10): p. 2299305. 46 760 41. Barman, M., et al., Serum fatty acid profile does not reflect seafood intake in 47 761 adolescents with atopic eczema. Acta Paediatr, 2014. 103(9): p. 96876. 48 762 42. Barman M, et al., Serum levels of Vitamin A and Atopic Rhinoconjunctivitis in 49 763 Swedish adolescents. Food Sci Nutr The, 2017. 3(1): p. 014019. 50 764 43. Barman, M., et al., No association between allergy and current 25-hydroxy vitamin D 51 765 in serum or vitamin D intake. Acta Paediatr, 2015. 104(4): p. 40513. 52 766 44. Lu, Y., et al., Alkali dilution of blood samples for high throughput ICP-MS analysis- 53 54 767 comparison with acid digestion. Clin Biochem, 2015. 48(3): p. 1407. 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 768 45. Rydbeck, F., et al., Urinary iodine concentrations of pregnant women in rural 4 769 Bangladesh: a longitudinal study. J Expo Sci Environ Epidemiol, 2014. 24(5): p. 504 5 770 9. 6 771 46. FrenckMestre, C., M. Besson, and J. Pynte, Finding the locus of semantic satiation: 7 772 an electrophysiological attempt. Brain Lang, 1997. 57(3): p. 40622. 8 773 47. Sjoberg, F., et al., Low-complexity microbiota in the duodenum of children with newly 9 774 diagnosed ulcerative colitis. PLoS One, 2017. 12(10): p. e0186178. 10 775 48. Hanhineva, K., et al., Discovery of urinary biomarkers of whole grain rye intake in 11 776 free-living subjects using nontargeted LC-MS metabolite profiling. Molecular 12 777 Nutrition & Food Research, 2015. 59(11): p. 23152325. 13 14 778 15 779 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from 20 20

ollow ollow up f

BMJ Open http://bmjopen.bmj.com/

on September 28, 2021 by guest. Protected copyright. 014 014 2015 2016 2017 2018 2019 2020 2021 2022

2 For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

Table 1: Overview 1: recruitment the study Overview NICE of Table and 781 781 Data collection collection Data 48 months follow follow up 48 months 12 months follow follow up 12 months Children born born Children Recruitment of patients of patients Recruitment Project start, planning planning start, Project

780 780 Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 22 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 782 5 Table 2: Overview of sample collection in NICE. GW GW Birth 48h 1m 2m 3m 4m 5m 6m 7m 8m 9m 10m 11m 12m 18m 48m 6 18/281 34 7 Questionnaires 8 "Allergic heredity and 2 9 environment" "Infant environment" 10

11 "Child environment" 12 "Infections and feeding”

13 Dietary Assessment 14 MealQ + ActiveQ , mothers 15 MealQ + ActiveQ, father 16 TodMealQ + TodActiveQ, infant For peer review only 17 KidMealQ + KidActiveQ, infant 18

19 Biological samples, mothers 20 Blood

21 Urine

22 Saliva 23 Placenta 24 Breast milk 25 Biological samples, fathers 26 27 Blood

28 Urine 29 Mouth swabs

30 Biological samples, all children

31 Blood 32 Urine 33 http://bmjopen.bmj.com/ Saliva 34 Faeces 35

36 Mouth swabs

37 Meconium

38 Hair 39 Biological samples, only

40 subcohort children

on September 28, 2021 by guest. Protected copyright. 41 Blood

42 Faeces 43 Mouth swabs 44 Clinical examination, children 45 RiNT 46

47 SPT

48 Doctor’s diagnosis of allergy 49 Neurologic test 50 Growth and obesity 51 Oral health 52 783 1Questionnairs distributed in gestational week 18. Biological collected in gestational week 28. 53 784 2Distributed to both mothers and fathers. 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 785 All studyrelated visits, with the exception of the first sampling of mothers at gestational 4 786 week 28 at the maternity clinics, will take place at the Sunderby Hospital. Midwives at the 5 787 maternity ward will be responsible for all sampling at the birth timepoint and assist with 6 788 collection of samples from the neonates at 48h. For all other time points post48 h the same 7 789 clinical team, consisting of two specialized neonatal nurses, will meet all participating 8 790 families. The same paediatric allergy specialist will examine all infants at 12 months and 48 9 791 months of age to diagnose any allergy and/or allergen sensitisation. At 48 months, the 10 792 children will also undergo an assessment of neuropsychological development, anthropometric 11 793 measurements as well as an oral health examination. 12 794 Abbreviations: GW = gestational week, SPT = skin prick test, RiNT = respiratory resistance 13 14 795 with interruption technique. 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 796 Table 3: Overview of questionnaires distibuted in the NICE- 5 6 797 cohort 7 Questionnaire Time Answered by Questions about 8 Q1: “Allergic GW 2025 Mothers and Education, profession, nationality, 9 heredity and partners answer allergic heredity, smoking habits, 10 environment” one questionnaire alcohol habits, residence, heating, 11 each ventilation, “mould”, siblings, 12 pets. 13 Q2: “Infant 18 months Parents answers Questions to both parents 14 environment” about the family regarding: profession, education, 15 ethnicity, civil status, smoking 16 For peer review only 17 habits, hunting habits, residence. 18 Questions about the infant: 19 feeding, vaccinations, travels, 20 daycare, behaviour, siblings, pets. 21 Q3: “Child 48 months Parents answers Questions to both parents 22 environment” about the family regarding: profession, education, 23 civil status, smoking habits, 24 hunting habits, residence 25 Questions about the infant: 26 feeding, vaccinations, travels, 27 daycare, behaviour , siblings, pets, 28 29 798 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 799 Table 4: Overview of the dietary assessments in NICE 5 Questionnaire Time Answered by 6 MealQ1+ ActiveQ1 GW 34 Mothers 7 8 MealQ2+ ActiveQ2 1 month Mothers 9 MealQ3 + ActiveQ3 4 months Mothers 10 TodMealQ+ TodActiveQ 12 months Parents answer for the infant 11 KidMealQ+ KidActiveQ 48 month Parents answer for the infant 12 800 GW: Gestational week 13 801 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 802 Table 5: Overview of research sites involved in the NICE- 5 6 803 consortium 7 Institution, Department, Group Principal investigator Main research interests in the 8 of the site NICE-cohort 9 Umeå University, Clinical Anna Sandin, MD, PhD, Environmental exposures in 10 Sciences, Paediatrics. Principal Investigator for relation to immune maturation 11 the NICEcohort. and allergy outcomes. 12 Chalmers University of AnnSofie Sandberg, Nutritional factors in relation to 13 Technology, Biology and Professor allergy, growth and neurological 14 Biological Engineering, Food and function. Metabolomics to 15 Nutrition Science. predict risk of disease. Systems 16 For peer review only 17 biology. 18 Gothenburg University, Clinical Agnes Wold, Professor Immune maturation, microbiota 19 Microbiology. and allergy development. 20 Karolinska Institute, Institute of Marie Vahter, Professor Metal exposure and epigenetics 21 Environmental Medicine. in relation to growth and 22 neurological function. 23 Gothenburg University, Bo Jacobsson, Professor Pregnancy exposures and 24 Sahlgrenska Academy, Obstetrics outcome with special interest in 25 and Gynaecology. placenta. 26 Umeå University, Department of Nicklas Strömberg, Nutritional, immune, microbiota 27 28 Odontology/Cariology. Professor and metal factors in caries 29 development. 30 Karolinska Institute, Department Sven Bölte, Professor Neuropsychological and 31 of Women’s and Children’s neuropsychiatric assessments 32 Health, Division of 33 Neuropsychiatry, Centre of http://bmjopen.bmj.com/ 34 Neurodevelopmental Disorders 35 (KIND). 36 804 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 805 Figure legends 5 806 Figure 1: Flow chart of the NICE-cohort 6 807 7 808 Figure 2: Schematic overview of placenta collection at delivery. 8 809 Samples from the placenta will be collected for analysis of histology, DNA, RNA, proteins, 9 3 10 810 metabolomics, fatty acids, toxicants and essential elements. First 4 pieces, 4 cm of villous 11 811 tissue including decidua will be collected from the basal plate from 4 separate places of the 12 812 placenta. The decidua will be removed stored in a separate cryotube. The remaining villous 13 813 tissue will be divided into 20 smaller pieces, 4 pieces of 3 mm Ø that will be put in 14 814 RNAlater® and 16 pieces that will be put 4 and 4 in separate cryotubes. Two different cross 15 815 sections will be then collected: one crosssection measuring approximately 2x2 cm taken 16 816 within 4 cm fromFor the cord peerinsert will be placedreview in formalin. Anotheronly crosssection formed as a 17 817 V taken from the cord insert to the end of the placenta will be sampled with specific metal 18 818 free ceramic knife and put in to plastic bags. The crosssections will be taken with care to 19 819 avoid selecting a tissue section that has previously been sampled for villous tissues. At one 20 820 cm from the cord insert a 2 cm piece of the umbilical cord will be sampled and stored in 21 821 formalin. From the same end of the cord 2 pieces, 1 cm each in length, will be sampled and 22 23 822 placed in a cryotube. Also, samples from the membrane will be collected: pieces of the 24 823 amnion and chorion membranes will be placed in separate cryotubes, while one piece of 25 824 membrane with both the membranes still attached to eachother (“double membrane”) will be 26 825 placed in formalin. 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from Page 28 of 30 27 27 BMJ Open http://bmjopen.bmj.com/ on September 28, 2021 by guest. Protected copyright. For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

826 826 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 29 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 Flow chart of the NICE-cohort 46 47 190x275mm (300 x 300 DPI) 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 30 BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Figure 2: Schematic overview of placenta collection at delivery. 32 33 254x190mm (300 x 300 DPI) http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 30 BMJ Open

1 2 3 Supplementary file 1: Modifications of the dietary questionnaires Meal-Q, BMJ Open: first published as 10.1136/bmjopen-2018-022013 on 21 October 2018. Downloaded from 4 5 TodMeal-Q and KidMeal-Q 6 7 1. Modification of the nutrition calculation program to distinguish between different 8 fat contents of "Milk, sour milk or yogurt" and "Other milk products, e.g. cream 9 or crème fraiche", when consumption patterns of these food items will be asked about, it is 10 11 followed by a question if you usually choose light products. 12 13 2. Addition of questions regarding game meat and lamb, fish species before and during 14 pregnancy, organic food purchases, and gluten and lactose in the 15 adult questionnaire. These questions were not included in the nutrition calculation. 16 17 3. Addition of a follow-up question to those respondents who report that they eat 18 For peer review only 19 flaxseeds in the adult questionnaire, plus corresponding modification of the 20 nutrition calculation program (i.e. different nutritional content of crushed 21 and whole flaxseeds, respectively). 22 23 4. Addition of questions regarding the participants’ consumption pattern of products with 24 probiotics in the adult survey (does not affect the nutrition calculation program). 25 26 27 5. Modification of the nutrition calculation program to distinguish between sweet and 28 and "light" drinks. The question of how often you drink "soda, cider, 29 fruit syrups, juice" on page 4 was coupled with the presence of light products 30 and the alternative "Soft drinks, or juice" on page 14. 31 32 6. Addition of questions regarding different types of "Infant Formula" to small children and 33 34 subsequent modifications of the nutritional calculation program. 35 36

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml