Next Article in Journal
Curcumin, Hormesis and the Nervous System
Previous Article in Journal
Severe Weight Loss and Its Association with Fatigue in Old Patients at Discharge from a Geriatric Hospital
Previous Article in Special Issue
Testing the Cow’s Milk-Related Symptom Score (CoMiSSTM) for the Response to a Cow’s Milk-Free Diet in Infants: A Prospective Study
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:

Statistical Approaches in the Studies Assessing Associations between Human Milk Immune Composition and Allergic Diseases: A Scoping Review

Oleg Blyuss
Ka Yan Cheung
Jessica Chen
Callum Parr
Loukia Petrou
Alina Komarova
Maria Kokina
Polina Luzan
Egor Pasko
Alina Eremeeva
Dmitrii Peshko
Vladimir I. Eliseev
Sindre Andre Pedersen
Meghan B. Azad
Kirsi M. Jarvinen
Diego G. Peroni
Valerie Verhasselt
Robert J. Boyle
John O. Warner
Melanie R. Simpson
12,13 and
Daniel Munblit
add Show full author list remove Hide full author list
Wolfson Institute of Preventive Medicine, Queen Mary University of London, London EC1M 6BQ, UK
Department of Paediatrics and Paediatric Infectious Diseases, Institute of Child’s Health, Sechenov First Moscow State Medical University (Sechenov University), 123337 Moscow, Russia
Department of Paediatrics, Imperial College London, London W2 1PG, UK
N. Polyakov Institute of Geotechnical Mechanics on the NAS of Ukraine, 49005 Dnipro, Ukraine
Library Section for Medicine and Health Sciences, NTNU—Norwegian University of Science and Technology, 7030 Trondheim, Norway
Department of Pediatrics and Child Health, Children’s Hospital Research Institute of Manitoba, University of Manitoba, Winnipeg, MB R3E 3P4, Canada
Division of Pediatric Allergy and Immunology & Center for Food Allergy, University of Rochester School of Medicine and Dentistry, Rochester, New York, NY 14642, USA
Department of Clinical and Experimental Medicine, Section of Paediatrics, University of Pisa, 56126 Pisa, Italy
School of Molecular Sciences, University of Western Australia, Perth, WA 6009, Australia
inVIVO Planetary Health, Group of the Worldwide Universities Network (WUN), West New York, NJ 10704, USA
National Institute for Health Research, Collaboration for Leadership in Applied Health Research and Care for NW London, London SW10 9NH, UK
Department of Public Health and General Practice, NTNU – Norwegian University of Science and Technology, 7030 Trondheim, Norway
Clinic of Laboratory Medicine, St Olavs Hospital, 7030 Trondheim, Norway
Solov’ev Research and Clinical Center for Neuropsychiatry, 115419 Moscow, Russia
Author to whom correspondence should be addressed.
These authors contributed equally to this paper.
Nutrients 2019, 11(10), 2416;
Submission received: 2 September 2019 / Revised: 27 September 2019 / Accepted: 1 October 2019 / Published: 10 October 2019
(This article belongs to the Special Issue Nutrition, Diet, and Allergic Diseases)


A growing number of studies are focusing on the associations between human milk (HM) immunological composition and allergic diseases. This scoping review aims to identify statistical methods applied in the field and highlight pitfalls and unmet needs. A comprehensive literature search in MEDLINE and Embase retrieved 13,607 unique records. Following title/abstract screening, 29 studies met the selection criteria and were included in this review. We found that definitions of colostrum and mature milk varied across the studies. A total of 17 out of 29 (59%) studies collected samples longitudinally, but only 12% of these used serial (longitudinal) analyses. Multivariable analysis was used in 45% of the studies, but statistical approaches to modelling varied largely across the studies. Types of variables included as potential confounding factors differed considerably between models. Discrimination analysis was absent from all studies and only a single study reported classification measures. Outcomes of this scoping review highlight lack of standardization, both in data collection and handling, which remains one of the main challenges in the field. Improved standardization could be obtained by a consensus group of researchers and clinicians that could recommend appropriate methods to be applied in future prospective studies, as well as already existing datasets.

1. Introduction

The prevalence of allergic disease has been rising worldwide in the past decades [1]. This pattern is particularly evident in westernised and urbanised countries [2], but recent data suggest that it is also an emerging problem in the developing world [3]. “The Hygiene Hypothesis” has been proposed as the main theory to explain the increase in prevalence of allergic disease [4], pointing at a reduced repertoire of specific immunological responses to different organisms due to reduced stimulation of the immune system by microbial antigens associated with modern living.
Human breast milk is a known contributor to neonatal infection prevention and recent data show that it may also reduce the risk of non-communicable diseases. It is possible that the pre-biotic properties of human milk (HM) contribute to the latter by enhancing the establishment of a disease impairing gut microbiome [5]. A systematic review by Lodge and co-authors suggested protection against asthma development, with weak evidence of association between breastfeeding and eczema [6] with heterogeneity in methodology reported across studies. Some experts have suggested that conflicting results may be linked to variation in HM composition [7,8,9]. Furthermore, some biologically active components in HM are suspected to be modified by interventions in the maternal diet and lifestyle during late pregnancy and lactation [9,10], subsequently influencing infant health outcomes [11,12].
Many studies assessing HM composition have focused on diverse allergic outcomes such as allergic sensitisation, eczema, food allergy and/or wheeze/asthma development in infancy and there is a need for systematic analysis of evidence for each condition independently [13]. Although several systematic reviews were recently published on this topic [14,15,16], appropriate meta-analysis of the data was not possible due to methodological inconsistencies between the studies.
It has been recently demonstrated that even when the same dataset is analysed independently by different research groups, methodological approaches and outcomes of the statistical analysis vary significantly [17]. In the experimental work, 29 research teams, involving 61 analysts, used a wide range of statistical models, and 21 unique combinations of covariates, which led to the estimated effect size ranging from 0.89 to 2.93 [18]. When study authors discussed approaches and results with members of the represented research groups, no consensus emerged on a single, best approach. This illustrates the importance of a proper justification of the statistical approach selected in a given setting and need for standardisation of statistical analyses.
Serial (longitudinal) samples (e.g., colostrum and mature milk, or longitudinal mature milk samples) are often collected in HM research, however, they are rarely analysed in a serial fashion. The importance of this approach has recently attracted increased attention in other disciplines, most notably oncology. The incorporation of longitudinal algorithms in a cancer-screening program resulted in a substantial increase in the predictive value of risk models [19].
This scoping review provides a critical appraisal of statistical approaches used in studies investigating associations between HM immunological composition and allergic disease outcomes, with an emphasis on serial analysis.

2. Materials and Methods

Our methods were based on Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR). A protocol outlining the process was prepared prior to this scoping review. The protocol has not been registered at PROSPERO as it does not currently accept registrations for scoping reviews, literature reviews or mapping reviews. This work, however, is complementary to the ongoing series of systematic reviews looking at available evidence on HM immunological composition and health outcomes and registered at PROSPERO (PROSPERO 2019 CRD42019126893 and PROSPERO 2019 CRD42019126894).

2.1. Search Strategy

In the process of conducting the systematic reviews mentioned in the previous section, an extensive electronic search was performed in the bibliographic databases MEDLINE and Embase using both thesaurus and free-text terms to identify bibliographic records involving HM and immunological composition, such as cytokines, chemokines and immunoglobulins (see Supplementary Material for detailed description of the literature search). The literature search was last updated on 11 December 2018. All records were imported into EndNote reference manager and all duplicates removed from the library. For the purpose of this scoping review, we employed the EndNote library to identify all records examining associations between HM immunological composition and allergic diseases.

2.2. Eligibility Criteria and Selection of Articles

For the purpose of this research, studies of all designs were included if the following criteria were met: (1) reported original data; (2) clinical study of mother–infant dyads; (3) the study had an epidemiological design: observational studies (i.e., pregnancy cohort study, birth cohort study, human prospective study or randomized controlled trial), during pregnancy or lactation and interventional studies; (4) included a quantitative assessment of HM immunological constituents (cytokines, chemokines and/or immunoglobulins); and (5) investigated associations between HM immunological constituents and at least one allergic disease, or allergic sensitization, in a child with follow-up from 6 months to 18 years.
Reviews, conference abstracts, editorials, letters to the editor, case reports and/or case series were excluded from the analysis. Specific immunoglobulins in HM was not part of this research.
To reduce potential selection bias, eight independent investigators (C.P., J.C., L.P., K.C., P.L., E.P., A.K. and M.K.) reviewed all titles and abstracts identified by the search for inclusion. Four researchers (C.P., J.C., L.P. and K.C.) independently reviewed full texts of all publications selected for data extraction. Any disagreements were resolved through discussion involving additional reviewers (O.B. and D.M.) until consensus was reached.

2.3. Data Extraction

The data from each study were extracted in duplicate, tabulated, and included the following fields: author(s) and year of publication; descriptive information concerning the study design; country and setting; number and timing of sample collection; details of statistical analysis; adjustment for potential confounders; outcome definition; age of outcome assessment.

3. Results

3.1. Synthesis

Based on the search strategy, a total of 13,607 unique records were identified and screened for eligibility (Figure 1). A total of 29 studies met our stated eligibility and selection of full-text articles criteria and we proceeded to full-text assessment and data extraction. All included studies were published between 1986 and 2018.

3.2. Sample Collection and Analysis

Authors reported colostrum/early milk collection within the first 10 days of life [20,21,22,23,24,25,26,27,28,29,30,31,32,33,34], and mature milk between seven days [35,36] and nine months [37] postpartum, with most of the studies collecting one-month milk. The number of immunological markers measured varied between one [25,38,39,40,41] and 50 [30] in different studies, with a median of four across the studies.
A total of 17 out of 29 studies collected samples at two [21,23,24,25,26,27,28,29,31,32,33,34,42,43] or more [25,37,38,41] timepoints. Only two [31,38] of these studies reported serial data analysis, and a single study [44] used a dimensionality reduction approach as part of its statistical analysis (Table 1).

3.3. Statistical Methods and Confounders

In total, 13 out of 29 studies analysed data using univariable analysis [20,21,23,26,32,34,37,38,39,41,42,43,45]. The most commonly used approach included the Mann–Whitney [21,23,27,31,33,34,39,42,43,45] and t-test [22,23,25,27,37,45]. Multivariable analysis was used in 16 studies [22,24,25,27,28,29,30,31,33,35,36,40,44,46,47,48], with logistic regression being the preferred approach to modelling [22,24,27,28,31,40,46,47,48]. Other techniques included Cox regression [30,44], generalized estimating equations (GEE) [35,36], binomial GLmulti [29] and least absolute shrinkage and selection operator (LASSO) [29].
Adjustment for potential confounding factors or use of covariates was reported in 17 studies [22,24,25,27,28,29,30,31,35,36,38,40,43,44,46,47,48], and in all studies since 2013 [27,28,29,30,31,35,36,38,40,43,44,46,47,48], apart from Hogendorf et al. [39]. Maternal atopy [24,28,29,30,31,35,44,46,47,48], child gender [25,30,31,35,44,46,48] and maternal smoking were the most commonly used confounders across the studies (Table 2).
Association between HM immunological markers and allergic outcomes was reported as a difference in means [20,22,25,37,38,45,46] or medians [21,23,27,31,34,39,42,43], odds [22,24,26,27,28,29,31,40,46,47,48], risk [35,36] or hazardous [30,44] ratios. In a single study, the levels of cytokines were divided into tertiles and association between infant sensitization and levels of immunological markers in HM was reported descriptively [32]. No studies reported discrimination analysis (receiver operating characteristic (ROC) curve, area under the curve (AUC)), with a single study using classification measures (sensitivity, specificity) [38]. Approaches to statistical analysis are summarised in Table 3.

4. Discussion

Human milk research aims to detect multiple biologically active components and associations with health outcomes. We are still far from fully understanding the complex composition of HM and its potential consequences for infant health. Emerging evidence generated by prospective cohorts may improve our understanding of the composition role of HM in health and disease. To generate plausible results, which later can be used in meta-analyses, it is highly important to apply the most appropriate and effective methodology.
Lack of standardisation is still a major issue in HM research and results from our scoping review highlight this. Among the 29 identified studies examining associations between immune constituents of HM and allergic diseases, we found that even though most studies collected samples longitudinally, very few performed serial analysis. Definitions of colostrum were very broad and often overlapped with mature milk between the studies. Potential confounding factors differed considerably between models and no studies used discrimination analysis, with only one using classification measures.
Three stages in human lactation are normally identified and explained by changes in specific components of the breast milk (e.g., whey proteins, lactose). These compositional changes likely reflect the changing physiological needs of the infant, and can be stratified into colostrum, transitional milk and mature milk. Unfortunately, there is little alignment in the precise definition of the terms mentioned above. According to various sources, colostrum is the special milk secreted in the first 2–3 days [49] or 1–5 days [50] after delivery; transitional milk is usually produced 7–14 [49] or 5–21 [50] days postpartum; and mature milk after two weeks [49] or >21 days [50] post-partum. These discrepancies cause significant heterogeneity in sample collection timing and HM definition between the studies. Improved standardisation will allow for better harmonisation of datasets.
Serial collection is widely used in the field of human milk research. Regrettably, four of 10 studies failed to collect HM samples longitudinally, at different time points. Even several of the recent studies collected HM samples only once [30,35,44].
Out of 17 studies collecting serial samples, only 12% (2/17) performed serial analysis using appropriate methods. For example, Jarvinen et al. used ANOVA for repeated measurements studying relationships between HM IgA and cow’s milk allergy [38], and Morita and co-authors reported associations between Transforming Growth Factor (TGF)β ratio (mature milk/colostrum) and development of eczema in offspring [31]. Longitudinal analysis provides additional insights into the personalized predictive models and some promising approaches were recently successfully adopted in cancer research [51], showing that serial models significantly outperform non-serial in terms of risk prediction. Applying similar approaches in the field of HM research may potentially allow the development of predictive models for the development of a variety of outcomes, including allergic diseases. Current lack of longitudinal analysis is surprising and is possibly linked with the lack of funding, with many studies failing to utilise some of the information collected (i.e., links between sequential sample collections). In an ‘ideal world’ scenario, sample collection should be longitudinal, but this may not always be feasible, particularly in large cohort studies where HM collection and analysis is not the primary focus. With that being said, sample collection limited to a single time point, restricts statistical analysis to the comparison with population characteristics as some individual’s previous information cannot be taken into account. Potential approaches to be used may include mixed models, which are commonly applied to the analysis of repeated measurements [52], methods that analyse trend indices to evaluate the dynamics [51], as well as more sophisticated techniques, including Bayesian changepoint models [53,54,55], and recurrent neural networks [56]—one of the most prominent deep learning techniques utilizing serial measurements. The listed approaches could be used to describe serial patterns in infants, with or without allergic diseases, thus enabling improved discrimination.
A wide selection of statistical methods was applied in the included studies. The spectrum ranged from simple comparison of two groups using the t-test and/or Mann–Whitney to regression models and dimensionality reduction approaches. In total, 41% of the studies (12/29) failed to conduct multivariable analysis, which may negatively impact the validity of the reported data. Exclusion of potentially important variables from the analysis may lead to misleading results. Since 2013, almost all studies have reported multivariable analysis use, which is very promising. In agreement with the study by Silberzahn et al. [18], we found that different groups tend to choose very different statistical methods for assessing the relationship between HM immune active molecules and allergic outcomes. While there is always a variety of statistical approaches that could be used in each setting, such diversity makes merging data-pooled estimates problematic. We therefore encourage adopting methods that generate characteristics that allow comparison in potential future meta-analyses such as odds ratio, relative risk and hazard ratio. This would also require standardization of units of concentration and choice of transformation or normalization methods prior to statistical analyses.
Similar trends were found in relation to confounding factors adjustment or use of covariates, with only 59% of the studies (17/29) considering potential confounders, and all the studies published since 2013 providing adjustment. The selection of potential confounders varied significantly between the studies, with maternal allergy, infant gender and maternal smoking being among the most frequently used. Interestingly, alcohol consumption during pregnancy and/or lactation has not been used as a potential confounder in the modelling and is of interest to be collected in the future. Appropriate use of statistical methods, as well as correct adjustment for potential confounding factors, is highly important in yielding valid results and avoiding possible bias. Silberzahn et al. recently demonstrated that application of identical statistical methods may lead to opposing results depending on the selection of confounding factors [18]. Future research in the field should provide adjustment for potential confounding factors, as there are data suggesting maternal and environmental characteristics may significantly influence both HM immune composition and allergic disease [57], however, a systematic review of available evidence is still needed.
In the studies assessing multiple biologically active molecules in a limited number of samples it is always hard to assess whether associations reflect real causal relationships or are just related to multiple hypothesis testing. This is certainly the case in HM research, where associations between the levels of different HM constituents with disease development are tested. There is no reliable way to prevent this and multiple testing correction methods (e.g., Bonferroni, Holm, Benjamini and Hochberg) are normally applied in order to reduce the risks of invalid interpretation. Unfortunately, a very limited number of reviewed studies adjusted the results for multiple testing. This implies that outcomes of some studies may be overstated, as not taking into account the possible impact of multiple simultaneous testing can greatly increase the probability of false positive findings [58]. We highly recommend use of the multiple testing technique in future HM research.
Reviewed studies of HM immune composition have assessed the association between the levels of various constituents and allergic diseases development, but none of them performed discrimination analysis, even though odds/risk/hazard ratios were reported in some cases. Use of discrimination analysis (e.g., ROC curves, AUC), and classification measures (e.g., sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV)), would determine whether immunologically active molecules in HM may be used as a predictive tool for allergic outcomes development. We stress the importance of not only reporting the effect sizes of the associations, but also corresponding discrimination characteristics, in order to allow direct comparison of models and populations across the studies. We acknowledge, however, that ROC analyses can be informative, but is not necessarily the goal of all HM studies. Identification of HM components association with health outcomes is often needed to understand disease mechanisms.
The strength of our review is its comprehensive scope and attention to statistical analysis detail. The limitations are related to the methodology of the scoping review and include the possibility that the review may have missed some relevant studies due to database selection, exclusion of grey literature from the search and exclusion of studies not published in English. For the purpose of this review the ‘immunological composition’ is limited to cytokines, chemokines and immunoglobulins. We acknowledge that other components (e.g., human milk oligosaccharides, fatty acids and cells) might have immunological functions. Since our scoping review is focused on the approach to data handling and analysis, this limitation is unlikely to influence the results.

5. Conclusions

Although some improvements were seen in the past five years, there was a high degree of heterogeneity across the studies regarding the methodology of data collection and analysis. This scoping review assessed statistical approaches used in the field and critically appraised available evidence. We found that there is a lack of agreement in both, data collection (variations in definitions of human milk “maturity”, lack of longitudinal sample collection) and data analysis (lack of multivariable analysis, serial analysis, use of discrimination analysis/classification measures, adjustment for important confounding factors). Better funding could possibly lead to improvements in methodology and longitudinal samples collection.
Lack of standardization is still one of the main challenges in the field, which makes meta-analysis difficult. We call national and international societies focused on human milk research to establish a consensus group, involving clinicians, researchers, methodologists and biostatisticians to produce proposals for standardization which could then be applied to already available datasets as well as future prospective studies.

Supplementary Materials

The following are available online at, Supplemental material: Search strategy.

Author Contributions

Conceptualization, O.B., A.E., D.G.P., M.R.S. and D.M.; data curation, K.Y.C., J.C., C.P., L.P., A.K., M.K., P.L., E.P., S.A.P. and D.M.; investigation, K.Y.C., J.C., C.P., L.P., A.K., M.K., P.L., E.P. and M.R.S.; methodology, O.B., A.E., V.I.E., S.A.P., J.O.W., M.R.S. and D.M.; supervision, D.P. and D.M.; writing—original draft, O.B., K.Y.C., J.C., C.P., L.P., M.R.S. and D.M.; writing—review and editing, O.B., K.Y.C., J.C., C.P., L.P., M.B.A., K.M.J., D.G.P., V.V., R.J.B., S.A.P., J.O.W., M.R.S. and D.M.


This research received no external funding.

Conflicts of Interest

With regards to this research the authors declare no conflicts of interest. The views expressed in the paper are those of the authors and not those of the NIHR or Department of Health (UK).


  1. Asher, M.I.; Montefort, S.; Bjorksten, B.; Lai, C.K.; Strachan, D.P.; Weiland, S.K.; Williams, H.; The ISAAC Phase Three Study Group. Worldwide time trends in the prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and eczema in childhood: Isaac phases one and three repeat multicountry cross-sectional surveys. Lancet 2006, 368, 733–743. [Google Scholar] [CrossRef]
  2. Williams, H.; Robertson, C.; Stewart, A.; Ait-Khaled, N.; Anabwani, G.; Anderson, R.; Asher, I.; Beasley, R.; Bjorksten, B.; Burr, M.; et al. Worldwide variations in the prevalence of symptoms of atopic eczema in the international study of asthma and allergies in childhood. J. Allergy Clin. Immunol. 1999, 103, 125–138. [Google Scholar] [CrossRef]
  3. Kung, S.J.; Steenhoff, A.P.; Gray, C. Food allergy in Africa: Myth or reality? Clin. Rev. Allergy Immunol. 2014, 46, 241–249. [Google Scholar] [CrossRef] [PubMed]
  4. Strachan, D.P. Hay fever, hygiene, and household size. BMJ 1989, 299, 1259–1260. [Google Scholar] [CrossRef] [PubMed]
  5. Victora, C.G.; Bahl, R.; Barros, A.J.; Franca, G.V.; Horton, S.; Krasevec, J.; Murch, S.; Sankar, M.J.; Walker, N.; Rollins, N.C.; et al. Breastfeeding in the 21st century: Epidemiology, mechanisms, and lifelong effect. Lancet 2016, 387, 475–490. [Google Scholar] [CrossRef]
  6. Lodge, C.J.; Tan, D.J.; Lau, M.X.; Dai, X.; Tham, R.; Lowe, A.J.; Bowatte, G.; Allen, K.J.; Dharmage, S.C. Breastfeeding and asthma and allergies: A systematic review and meta-analysis. Acta Paediatr. 2015, 104, 38–53. [Google Scholar] [CrossRef] [PubMed]
  7. D’Alessandro, A.; Scaloni, A.; Zolla, L. Human milk proteins: An interactomics and updated functional overview. J. Proteome Res. 2010, 9, 3339–3373. [Google Scholar] [CrossRef] [PubMed]
  8. Agarwal, S.; Karmaus, W.; Davis, S.; Gangur, V. Immune markers in breast milk and fetal and maternal body fluids: A systematic review of perinatal concentrations. J. Hum. Lact. 2011, 27, 171–186. [Google Scholar] [CrossRef]
  9. Munblit, D.; Peroni, D.G.; Boix-Amoros, A.; Hsu, P.S.; Land, B.V.; Gay, M.C.L.; Kolotilina, A.; Skevaki, C.; Boyle, R.J.; Collado, M.C.; et al. Human milk and allergic diseases: An unsolved puzzle. Nutrients 2017, 9, 894. [Google Scholar] [CrossRef]
  10. Seppo, A.E.; Kukkonen, A.K.; Kuitunen, M.; Savilahti, E.; Yonemitsu, C.; Bode, L.; Jarvinen, K.M. Association of maternal probiotic supplementation with human milk oligosaccharide composition. JAMA Pediatr. 2019, 173, 286–288. [Google Scholar] [CrossRef]
  11. Peroni, D.G.; Pescollderungg, L.; Piacentini, G.L.; Rigotti, E.; Maselli, M.; Watschinger, K.; Piazza, M.; Pigozzi, R.; Boner, A.L. Immune regulatory cytokines in the milk of lactating women from farming and urban environments. Pediatr. Allergy Immunol. 2010, 21, 977–982. [Google Scholar] [CrossRef] [PubMed]
  12. Holmlund, U.; Amoudruz, P.; Johansson, M.A.; Haileselassie, Y.; Ongoiba, A.; Kayentao, K.; Traore, B.; Doumbo, S.; Schollin, J.; Doumbo, O.; et al. Maternal country of origin, breast milk characteristics and potential influences on immunity in offspring. Clin. Exp. Immunol. 2010, 162, 500–509. [Google Scholar] [CrossRef] [PubMed]
  13. Boix-Amoros, A.; Collado, M.C.; Van’t Land, B.; Calvert, A.; Le Doare, K.; Garssen, J.; Hanna, H.; Khaleva, E.; Peroni, D.G.; Geddes, D.T.; et al. Reviewing the evidence on breast milk composition and immunological outcomes. Nutr. Rev. 2019, 77, 541–556. [Google Scholar] [CrossRef] [PubMed]
  14. Khaleva, E.; Gridneva, Z.; Geddes, D.T.; Oddy, W.H.; Colicino, S.; Blyuss, O.; Boyle, R.J.; Warner, J.O.; Munblit, D. Transforming growth factor beta in human milk and allergic outcomes in children: A systematic review. Clin. Exp. Allergy 2019, 49, 1201–1213. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  15. Waidyatillake, N.T.; Dharmage, S.C.; Allen, K.J.; Lodge, C.J.; Simpson, J.A.; Bowatte, G.; Abramson, M.J.; Lowe, A.J. Association of breast milk fatty acids with allergic disease outcomes-a systematic review. Allergy 2018, 73, 295–312. [Google Scholar] [CrossRef] [PubMed]
  16. Doherty, A.M.; Lodge, C.J.; Dharmage, S.C.; Dai, X.; Bode, L.; Lowe, A.J. Human milk oligosaccharides and associations with immune-mediated disease and infection in childhood: A systematic review. Front. Pediatr. 2018, 6, 91. [Google Scholar] [CrossRef] [PubMed]
  17. Silberzahn, R.; Uhlmann, E.L. Crowdsourced research: Many hands make tight work. Nature 2015, 526, 189–191. [Google Scholar] [CrossRef] [PubMed]
  18. Silberzahn, R.; Uhlmann, E.L.; Martin, D.P.; Anselmi, P.; Aust, F.; Awtrey, E.; Bahník, Š.; Bai, F.; Bannard, C.; Bonnier, E.; et al. Many analysts, one data set: Making transparent how variations in analytic choices affect results. Adv. Methods Pract. Psychol. Sci. 2018, 1, 337–356. [Google Scholar] [CrossRef]
  19. Drescher, C.W.; Shah, C.; Thorpe, J.; O’Briant, K.; Anderson, G.L.; Berg, C.D.; Urban, N.; McIntosh, M.W. Longitudinal screening algorithm that incorporates change over time in ca125 levels identifies ovarian cancer earlier than a single-threshold rule. J. Clin. Oncol. 2013, 31, 387–392. [Google Scholar] [CrossRef] [PubMed]
  20. Saarinen, K.M.; Vaarala, O.; Klemetti, P.; Savilahti, E. Transforming growth factor-beta1 in mothers’ colostrum and immune responses to cows’ milk proteins in infants with cows’ milk allergy. J. Allergy Clin. Immunol. 1999, 104, 1093–1098. [Google Scholar] [CrossRef]
  21. Böttcher, M.F.; Jenmalm, M.C.; Björkstén, B. Cytokine, chemokine and secretory iga levels in human milk in relation to atopic disease and iga production in infants. Pediatr. Allergy Immunol. 2003, 14, 35–41. [Google Scholar] [CrossRef] [PubMed]
  22. Savilahti, E.; Siltanen, M.; Kajosaari, M.; Vaarala, O.; Saarinen, K.M. Iga antibodies, tgf-beta1 and -beta2, and soluble cd14 in the colostrum and development of atopy by age 4. Pediatr. Res. 2005, 58, 1300–1305. [Google Scholar] [CrossRef] [PubMed]
  23. Rigotti, E.; Piacentini, G.L.; Ress, M.; Pigozzi, R.; Boner, A.L.; Peroni, D.G. Transforming growth factor-1 and interleukin-10 in breast milk and development of atopic diseases in infants. Clin. Exp. Allergy 2006, 36, 614–618. [Google Scholar] [CrossRef] [PubMed]
  24. Böttcher, M.F.; Abrahamsson, T.R.; Fredriksson, M.; Jakobsson, T.; Björkstén, B. Low breast milk tgf-β2 is induced by lactobacillus reuteri supplementation and associates with reduced risk of sensitization during infancy. Pediatr. Allergy Immunol. 2008, 19, 497–504. [Google Scholar] [CrossRef] [PubMed]
  25. Pesonen, M.; Kallio, M.J.T.; Siimes, M.A.; Savilahti, E.; Ranki, A. Serum immunoglobulin a concentration in infancy, but not human milk immunoglobulin a, is associated with subsequent atopic manifestations in children and adolescents: A 20-year prospective follow-up study. Clin. Exp. Allergy 2011, 41, 688–696. [Google Scholar] [CrossRef] [PubMed]
  26. Kuitunen, M.; Kukkonen, A.K.; Savilahti, E. Impact of maternal allergy and use of probiotics during pregnancy on breast milk cytokines and food antibodies and development of allergy in children until 5 years. Int. Arch. Allergy Immunol. 2012, 159, 162–170. [Google Scholar] [CrossRef] [PubMed]
  27. Ochiai, S.; Shimojo, N.; Morita, Y.; Tomiita, M.; Arima, T.; Inoue, Y.; Nakaya, M.; Uehara, N.; Sato, Y.; Mori, C.; et al. Cytokine biomarker candidates in breast milk associated with the development of atopic dermatitis in 6-month-old infants. Int. Arch. Allergy Immunol. 2013, 160, 401–408. [Google Scholar] [CrossRef] [PubMed]
  28. Simpson, M.R.; Rø, A.D.B.; Grimstad, Ø.; Johnsen, R.; Storrø, O.; Øien, T. Atopic dermatitis prevention in children following maternal probiotic supplementation does not appear to be mediated by breast milk tslp or tgf-β. Clin. Transl. Allergy 2016, 6, 27. [Google Scholar] [CrossRef] [PubMed]
  29. Munblit, D.; Treneva, M.; Peroni, D.G.; Colicino, S.; Chow, L.Y.; Dissanayeke, S.; Pampura, A.; Boner, A.L.; Geddes, D.T.; Boyle, R.J.; et al. Immune components in human milk are associated with early infant immunological health outcomes: A prospective three-country analysis. Nutrients 2017, 9, 532. [Google Scholar] [CrossRef] [PubMed]
  30. Berdi, M.; de Lauzon-Guillain, B.; Forhan, A.; Castelli, F.A.; Fenaille, F.; Charles, M.A.; Heude, B.; Junot, C.; Adel-Patient, K. Immune components of early breastmilk: Association with maternal factors and with reported food allergy in childhood. Pediatr. Allergy Immunol. 2019, 30, 107–116. [Google Scholar] [CrossRef]
  31. Morita, Y.; Campos-Alberto, E.; Yamaide, F.; Nakano, T.; Ohnisi, H.; Kawamoto, M.; Kawamoto, N.; Matsui, E.; Kondo, N.; Kohno, Y.; et al. Tgf-β concentration in breast milk is associated with the development of eczema in infants. Front. Pediatr. 2018, 6, 162. [Google Scholar] [CrossRef] [PubMed]
  32. Huurre, A.; Laitinen, K.; Rautava, S.; Korkeamaki, M.; Isolauri, E. Impact of maternal atopy and probiotic supplementation during pregnancy on infant sensitization: A double-blind placebo-controlled study. Clin. Exp. Allergy 2008, 38, 1342–1348. [Google Scholar] [CrossRef] [PubMed]
  33. Prescott, S.L.; Wickens, K.; Westcott, L.; Jung, W.; Currie, H.; Black, P.N.; Stanley, T.V.; Mitchell, E.A.; Fitzharris, P.; Siebers, R.; et al. Supplementation with lactobacillus rhamnosus or bifidobacterium lactis probiotics in pregnancy increases cord blood interferon-gamma and breast milk transforming growth factor-beta and immunoglobin a detection. Clin. Exp. Allergy 2008, 38, 1606–1614. [Google Scholar] [CrossRef] [PubMed]
  34. Tomicic, S.; Johansson, G.; Voor, T.; Bjorksten, B.; Bottcher, M.F.; Jenmalm, M.C. Breast milk cytokine and iga composition differ in estonian and swedish mothers-relationship to microbial pressure and infant allergy. Pediatr. Res. 2010, 68, 330–334. [Google Scholar] [CrossRef] [PubMed]
  35. Soto-Ramírez, N.; Boyd, K.; Zhang, H.; Gangur, V.; Goetzl, L.; Karmaus, W. Maternal serum but not breast milk il-5, il-6, and il-13 immune markers are associated with scratching among infants. Allergy Asthma Clin. Immunol. 2016, 12, 25. [Google Scholar] [CrossRef] [PubMed]
  36. Soto-Ramírez, N.; Karmaus, W.; Yousefi, M.; Zhang, H.; Liu, J.; Gangur, V. Maternal immune markers in serum during gestation and in breast milk and the risk of asthma-like symptoms at ages 6 and 12 months: A longitudinal study. Allergy Asthma Clin. Immunol. 2012, 8, 11. [Google Scholar] [CrossRef] [PubMed]
  37. Savilahti, E.; Tainio, V.M.; Salmenpera, L.; Arjomaa, P.; Kallio, M.; Perheentupa, J.; Siimes, M.A. Low colostral iga associated with cow’s milk allergy. Acta Paediatr. Scand. 1991, 80, 1207–1213. [Google Scholar] [CrossRef] [PubMed]
  38. Järvinen, K.M.; Laine, S.T.; Järvenpää, A.L.; Suomalainen, H.K. Does low IGA in human milk predispose the infant to development of cow’s milk allergy? Pediatr. Res. 2000, 48, 457–462. [Google Scholar] [CrossRef] [PubMed]
  39. Hogendorf, A.; Stańczyk-Przyłuska, A.; Siniewicz-Luzeńczyk, K.; Wiszniewska, M.; Arendarczyk, J.; Banasik, M.; Fendler, W.; Kowalski, M.; Zeman, K. Is there any association between secretory IGA and lactoferrin concentration in mature human milk and food allergy in breastfed children. Med. Wieku Rozwoj. 2013, 17, 47–52. [Google Scholar] [PubMed]
  40. Joseph, C.L.M.; Havstad, S.; Bobbitt, K.; Woodcroft, K.; Zoratti, E.M.; Nageotte, C.; Misiak, R.; Enberg, R.; Nicholas, C.; Ezell, J.M.; et al. Transforming growth factor beta (tgfβ1) in breast milk and indicators of infant atopy in a birth cohort. Pediatr. Allergy Immunol. 2014, 25, 257–263. [Google Scholar] [CrossRef]
  41. Machtinger, S.; Moss, R. Cow’s milk allergy in breast-fed infants: The role of allergen and maternal secretory iga antibody. J. Allergy Clin. Immunol. 1986, 77, 341–347. [Google Scholar] [CrossRef]
  42. Kalliomäki, M.; Ouwehand, A.; Arvilommi, H.; Kero, P.; Isolauri, E. Transforming growth factor-β in breast milk: A potential regulator of atopic disease at an early age. J. Allergy Clin. Immunol. 1999, 104, 1251–1257. [Google Scholar] [CrossRef]
  43. Ismail, I.H.; Licciardi, P.V.; Oppedisano, F.; Boyle, R.J.; Tang, M.L.K. Relationship between breast milk scd14, tgf-β1 and total iga in the first month and development of eczema during infancy. Pediatr. Allergy Immunol. 2013, 24, 352–360. [Google Scholar] [CrossRef] [PubMed]
  44. Jepsen, A.A.; Chawes, B.L.; Carson, C.G.; Schoos, A.M.M.; Thysen, A.H.; Waage, J.; Brix, S.; Bisgaard, H. High breast milk il-1β level is associated with reduced risk of childhood eczema. Clin. Exp. Allergy 2016, 46, 1344–1354. [Google Scholar] [CrossRef] [PubMed]
  45. Rautava, S.; Kalliomaki, M.; Isolauri, E. Probiotics during pregnancy and breast-feeding might confer immunomodulatory protection against atopic disease in the infant. J. Allergy Clin. Immunol. 2002, 109, 119–121. [Google Scholar] [CrossRef] [Green Version]
  46. Oddy, W.H.; Halonen, M.; Martinez, F.D.; Lohman, I.C.; Stern, D.A.; Kurzius-Spencer, M.; Guerra, S.; Wright, A.L. Tgf-β in human milk is associated with wheeze in infancy. J. Allergy Clin. Immunol. 2003, 112, 723–728. [Google Scholar] [CrossRef]
  47. Snijders, B.E.; Damoiseaux, J.G.; Penders, J.; Kummeling, I.; Stelma, F.F.; van Ree, R.; van den Brandt, P.A.; Thijs, C. Cytokines and soluble cd14 in breast milk in relation with atopic manifestations in mother and infant (koala study). Clin. Exp. Allergy. 2006, 36, 1609–1615. [Google Scholar] [CrossRef]
  48. Orivuori, L.; Loss, G.; Roduit, C.; Dalphin, J.C.; Depner, M.; Genuneit, J.; Lauener, R.; Pekkanen, J.; Pfefferle, P.; Riedler, J.; et al. Soluble immunoglobulin a in breast milk is inversely associated with atopic dermatitis at early age: The pasture cohort study. Clin. Exp. Allergy 2014, 44, 102–112. [Google Scholar] [CrossRef]
  49. World Health Organization. Infant and Young Child Feeding; WHO: Geneva, Switzerland, 2009. [Google Scholar]
  50. Darragh, A.; Lönnerdal, B. Human milk. In Encyclopedia of Dairy Sciences, 2nd ed.; Elsevier: London, UK, 2011; pp. 581–590. [Google Scholar]
  51. Blyuss, O.; Burnell, M.; Ryan, A.; Gentry-Maharaj, A.; Marino, I.P.; Kalsi, J.; Manchanda, R.; Timms, J.F.; Parmar, M.; Skates, S.J.; et al. Comparison of longitudinal ca125 algorithms as a first-line screen for ovarian cancer in the general population. Clin. Cancer Res. 2018, 24, 4726–4733. [Google Scholar] [CrossRef]
  52. Tomasko, L.; Helms, R.W.; Snapinn, S.M. A discriminant analysis extension to mixed models. Stat. Med. 1999, 18, 1249–1260. [Google Scholar] [CrossRef]
  53. Slate, E.H.; Turnbull, B.W. Statistical models for longitudinal biomarkers of disease onset. Stat. Med. 2000, 19, 617–637. [Google Scholar] [CrossRef]
  54. Skates, S.J.; Pauler, D.K.; Jacobs, I.J. Screening based on the risk of cancer calculation from bayesian hierarchical changepoint and mixture models of longitudinal markers. J. Am. Stat. Assoc. 2001, 96, 429–439. [Google Scholar] [CrossRef]
  55. Mariño, I.P.; Blyuss, O.; Ryan, A.; Gentry-Maharaj, A.; Timms, J.F.; Dawnay, A.; Kalsi, J.; Jacobs, I.; Menon, U.; Zaikin, A. Change-point of multiple biomarkers in women with ovarian cancer. Biomed. Signal Process. Control 2017, 33, 169–177. [Google Scholar] [CrossRef] [Green Version]
  56. Vazquez, M.A.; Marino, I.P.; Blyuss, O.; Ryan, A.; Gentry-Maharaj, A.; Kalsi, J.; Manchanda, R.; Jacobs, I.; Menon, U.; Zaikin, A. A quantitative performance study of two automatic methods for the diagnosis of ovarian cancer. Biomed. Signal Process. Control 2018, 46, 86–93. [Google Scholar] [CrossRef] [PubMed]
  57. Munblit, D.; Boyle, R.J.; Warner, J.O. Factors affecting breast milk composition and potential consequences for development of the allergic phenotype. Clin. Exp. Allergy 2015, 45, 583–601. [Google Scholar] [CrossRef] [PubMed]
  58. Gyorffy, B.; Gyorffy, A.; Tulassay, Z. The problem of multiple testing and its solutions for genom-wide studies. Orv. Hetil. 2005, 146, 559–563. [Google Scholar]
Figure 1. Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) chart describing study selection process.
Figure 1. Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) chart describing study selection process.
Nutrients 11 02416 g001
Table 1. Serial sample collection and serial data analysis in the studies assessing associations between human milk (HM) immunological composition and allergic diseases.
Table 1. Serial sample collection and serial data analysis in the studies assessing associations between human milk (HM) immunological composition and allergic diseases.
Author, YearCountryNumber of ParticipantsNumber of HM Samples CollectedSerial Sample CollectionNumber of
Immune Markers Measured
Serial Analysis 1Dimension Reduction/
Clustering 2
Machtinger 1986 [41]USA57NRYes1NoNo
Savilahti 1991 [37]Finland161102 C (NR)
204 MM (2,6,9 mo)
Kalliomaki 1999 [42]Finland4743 C (NR)
38 MM (3 mo)
Saarinen 1999 [20]Finland315315 C (1–4 d) No5N/ANo
Jarvinen 2000 [38]Finland87NR MM
(2w, 1, 3, 6 mo)
Rautava 2002 [45]Finland62NR ▪▪ MM (3 mo)No2N/ANo
Bottcher 2003 [21]Sweden5353 C (4 d)
47 MM (1 mo)
Oddy 2003 [46]USA243243 MM (2 w)No4N/ANo
Savilahti 2005 [22]Finland228228 C (1–4 d)No4N/ANo
Rigotti 2006 [23]Italy2222 C (3 d)
22 MM (1 mo)
Snijders 2006 [47]Netherlands315315 MM
(1 mo)
Bottcher 2008 [24]Sweden109109 C (<3 d)
109 MM (1 mo)
Huurre 2008 [32]FinlandBetween
118 and 126
58 С (1 d)
68 (1 mo)
Prescott 2008 [33]New Zealand105239 MM (7d, 3, 6 mo)Yes8NoNo
Tomicic 2010 [34]Estonia, Sweden 9999 C (0-4 d)
99 MM (1 mo)
Pesonen 2011 [25]Finland169169 C (5 d)
286 MM
(2, 6 mo)
Kuitunen 2012 [26]Finland364364 C (0–3 d)
321 MM (3 mo)
Soto-Ramírez 2012 [36]United States of America115115 MM
(1–8 w)
Hogendorf 2013 [39]Poland8484 MM (NR) No1N/ANo
Ismail 2013 [43]Australia, Malaysia, UK 79158 MM (7, 28d)Yes3NoNo
Ochiai 2013[27]Japan9898 C (4–5 d)
98 MM (1 mo)
Orivuori 2013 [48]Austria, Finland, France, Germany and Switzerland610610 MM (2 mo) No2N/ANo
Joseph 2014 [40]USA304304 MM
(1 mo)
Jepsen 2016 [44]Denmark223223 MM
(1 mo)
Simpson 2016 [28]Norway259255 MM (10 d)
247 MM
(3 mo)
Soto-Ramírez 2016 [35]United States of America115115 MM
(1–8 w)
Munblit 2017 [29]UK, Italy, Russia398398 C (6 d)
153 MM (4–6 w)
Morita 2018 [31]Japan9696 C (5 d)
96 MM (1 mo)
Berdi 2019 [30]France263263 C (2–6 d)No50N/ANo
Abbreviations; C, colostrum; D, days; Mo, months; MM, mature milk; NA, not applicable; NR, not reported; W, weeks. 1 Serial analysis was considered as positive if any attempts were undertaken to handle data as serial measurements rather than single time-point variables; 2 Latent class analysis (LCA), Principal component analysis (PCA), interactions and scanning electron microscopy (SEM) were used in the analysis; an average of 3.5 mature milk samples were obtained per mother; ▪▪ 62 mother–infant pairs participated in the study.
Table 2. Confounding factors/covariates reported in the reviewed studies, assessing associations between HM immunological composition and allergic diseases.
Table 2. Confounding factors/covariates reported in the reviewed studies, assessing associations between HM immunological composition and allergic diseases.
Confounding FactorsFrequencyReference
Maternal atopy11[24,28,29,30,31,35,38,44,46,47,48]
Child gender7[25,30,31,35,44,46,48]
Maternal smoking6[25,28,30,35,44,46]
Breastfeeding duration4[22,25,43,44]
Maternal age4[30,35,44,47]
Number of siblings4[25,28,43,47]
Family history of atopy3[22,25,35]
Site of collection3[29,30,48]
Exposure to other children2[44,46]
Maternal educational level2[30,46]
Mode of delivery2[43,44]
Sibling atopy2[29,30]
Colostrum collection time/infant age2[29,38]
Birth weight1[46]
BMI before pregnancy1[30]
Breastfeeding by 1 month and Transforming Growth Factor (TGF)β ratio1[31]
Gestational age1[46]
Household income1[44]
Household pets1[43]
Introduction of food during first year of life1[48]
Maternal consumption of acetaminophen during pregnancy1[35]
Maternal infection1[47]
Maternal marital status1[35]
Maternal race1[35]
Mother’s alcohol use (3rd trimester)1[44]
Mother’s antibiotic use (3rd trimester)1[44]
Na+/K+ ratios1[24]
Season of birth1[35]
Season of breast milk collection1[47]
Study treatment1[24]
Time interval between births1[47]
Vaginal or urinary infections during pregnancy1[35]
Yoghurt and antibiotic consumption during pregnancy1[43]
Table 3. Statistical approaches to data handling, adjustment for potential confounding factors and use of discrimination and classification measures in the studies assessing associations between HM immunological composition and allergic diseases.
Table 3. Statistical approaches to data handling, adjustment for potential confounding factors and use of discrimination and classification measures in the studies assessing associations between HM immunological composition and allergic diseases.
Author, YearUnivariable/MultivariableAnalysisStatistical Method of HM Marker/Outcome AssessmentConfounders 1Association ReportingDiscrimination Analysis 2Yes/NoClassification Measures 3
Machtinger 1986 [41]UnivariableChi-squared, Fisher’sNoProportions NoNo
Savilahti 1991 [20]Univariablet-testNoMean differencesNoNo
Kalliomaki 1999 [42]UnivariableKruskal–Wallis,Mann–WhitneyNoMedian differencesNoNo
Saarinen 1999 [20]UnivariableANOVANoMean differencesNoNo
Jarvinen 2000 [38]UnivariableANOVA for repeated measurementYesMean differenceNoYes
Rautava 2002 [45]Univariablet-test,Mann–WhitneyNoMean differenceNoNo
Bottcher 2003 [21]UnivariableChi-squared, Fisher’s, Mann–WhitneyNoMedian differencesNoNo
Oddy 2003 [46]MultivariableChi-squared,multivariable logistic regressionYesMean differences, Odds RatiosNoNo
Savilahti 2005 [22]MultivariableIndependent samples t-test, multivariable logistic regressionYesMean differences, Odds RatiosNoNo
Rigotti 2006 [23]UnivariableMann–Whitney, independent samples, t-testNoMedian differencesNoNo
Snijders 2006 [47]MultivariableMultivariable logistic regressionYesOdds RatiosNoNo
Bottcher 2008 [24]MultivariableLogistic regressionYesOdds ratios NoNo
Huurre 2008 [32]UnivariableDescriptiveNoDescriptiveNoNo
Prescott 2008 [33]MultivariableMann–Whitney, Chi-squared, Fisher’s, logistic regression ▪NoNRNoNo
Tomicic 2010 [34]UnivariableMann–WhitneyNoMedian differencesNoNo
Pesonen 2011 [25]MultivariableTwo-tailed unpaired t-testYesMean differencesNoNo
Kuitunen 2012 [26]UnivariableChi-squared, ANOVA, Mantel–HaenszelNoGeometric means, Odds ratiosNoNo
Soto-Ramírez 2012 [36]MultivariableLog-linear regression,generalized estimating equationsYesRisk RatioNoNo
Hogendorf 2013 [39]UnivariableMann–WhitneyNoMedian differencesNoNo
Ismail 2013 [43]UnivariableMann–WhitneyYesMedian differencesNoNo
Ochiai 2013[27]MultivariableFisher’s, t-test, Mann­Whitney, multivariable logistic regressionYesProportions, Odds Ratios, Median differencesNoNo
Orivuori 2013 [48]MultivariableMultivariable logistic regressionYesOdds RatiosNoNo
Joseph 2014 [40]MultivariableLogistic regressionYesOdds ratiosNoNo
Jepsen 2016 [44]MultivariableCox regressionYesHazard ratiosNoNo
Simpson 2016 [28]MultivariableLogistic regressionYesOdds RatiosNoNo
Soto-Ramírez 2016 [35]MultivariableGeneralized estimating equationsYesRisk RatioNoNo
Munblit 2017 [29]MultivariableBinomial GLmulti, LASSOYesOdds RatiosNoNo
Morita 2018 [31]MultivariableMann–Whitney, multivariable logistic regressionYesOdds Ratios, Median differencesNoNo
Berdi 2019 [30]MultivariableCox regressionYesHazard ratiosNoNo
Abbreviations: LASSO, least absolute shrinkage and selection operator. 1 Adjustment for potential confounding factors in the model; 2 measure(s) of discrimination (receiver operating characteristic (ROC) curve, area under the curve (AUC) or C-statistic, log-rank, D-statistic); 3 sensitivity, specificity, predictive values, net reclassification improvement or cut-off points were reported; ▪ it is unclear whether logistic regression was used for assessment of associations between immunological markers and health outcomes.

Share and Cite

MDPI and ACS Style

Blyuss, O.; Cheung, K.Y.; Chen, J.; Parr, C.; Petrou, L.; Komarova, A.; Kokina, M.; Luzan, P.; Pasko, E.; Eremeeva, A.; et al. Statistical Approaches in the Studies Assessing Associations between Human Milk Immune Composition and Allergic Diseases: A Scoping Review. Nutrients 2019, 11, 2416.

AMA Style

Blyuss O, Cheung KY, Chen J, Parr C, Petrou L, Komarova A, Kokina M, Luzan P, Pasko E, Eremeeva A, et al. Statistical Approaches in the Studies Assessing Associations between Human Milk Immune Composition and Allergic Diseases: A Scoping Review. Nutrients. 2019; 11(10):2416.

Chicago/Turabian Style

Blyuss, Oleg, Ka Yan Cheung, Jessica Chen, Callum Parr, Loukia Petrou, Alina Komarova, Maria Kokina, Polina Luzan, Egor Pasko, Alina Eremeeva, and et al. 2019. "Statistical Approaches in the Studies Assessing Associations between Human Milk Immune Composition and Allergic Diseases: A Scoping Review" Nutrients 11, no. 10: 2416.

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop