Likely causes of type 1 (mainly childhood) diabetes



The following is a supplement to an article on breastfeeding versus formula feeding, which can be found at  This is supplemental to part 4 of that article, but it is also of interest separately.




4.c.1:  Toxins that are strongly associated with childhood diabetes and to which many young children are extraordinarily exposed, via a modifiable pathway


4.c.2:  No other suggested risk factors for childhood diabetes are considered to be as strong


4.c.3: The highest incidence rates and increases of childhood type 1 diabetes have been in the parts of the world where exposures to PCBs and dioxins have been highest.  The differences in PCB/dioxin exposures seem to combine with differences in prevalence of the main pathway of infant exposure of those toxins:  breastfeeding.


4.c.4: National links between childhood diabetes and breastfeeding in recent decades, possibly with no exceptions.

Increases then stability, first in breastfeeding and later in childhood diabetes, in three countries.


4.c.5: Another principal pattern in many countries:  initial low levels in breastfeeding and childhood diabetes, then rapid increases, first in breastfeeding and later in childhood diabetes. 


4.c.6: Continuous global increase of type 1 diabetes, not satisfactorily explained by the other suggested risk factors:     


4.c.7: Explanations that have been suggested for the geographic differences in type 1 diabetes prevalence, leading to a need for a better explanation:   

High diabetes seems to follow high breastfeeding, whether the latitude generalization applies or does not apply:  Type 1 diabetes is also low where breastfeeding is low.


4.c.8: Many especially large national increases in type 1 diabetes, with compatible especially large corresponding increases in breastfeeding: 


4.c.9: Exceptional national decrease in type 1 diabetes, and exceptional matching decrease in breastfeeding: 


4.c.10: An early upturn and dramatic increase in Sweden, first in breastfeeding and later in type 1 diabetes, separated by a predictable lag:


4.c.11: Other noteworthy examples of type 1 diabetes varying in the same direction and at times compatible with earlier variations in breastfeeding





The AAP’s 2022 breastfeeding document states that “extensive data confirm” that diabetes occurs less frequently among children who were breastfed.11  The possible validity of that statement should be considered in light of the following: 

  (a) the recognized low quality and frequent refutation of observational studies (part 2 of this article); such studies formed the basis of the AAP’s breastfeeding statement;11,12,13

  (b) there have been epidemic increases in diabetes among children and later among young adults following the dramatic increases in breastfeeding in the U.S. 10,13, 81, 191 and in Europe.25 and



4.c.1Toxins that are strongly associated with childhood diabetes and to which many young children are extraordinarily exposed, via a modifiable pathway:


We should consider the findings of a 2017 U.S. study:  the incidence of type 1 diabetes has increased among children by 1.9% per year between 2001 and 2015, during which years the disease has decreased among adults.216  Of further interest, three studies have reported that “the most rapid increase” of type 1 diabetes has taken place in the 0-4 age group.217  So it is crucial to consider toxins to which very young children, but not adults, have been increasingly exposed.

The above should direct attention toward PCBs and dioxins, since very young children are the part of the population with by far the greatest exposures to those toxins we should bear in mind the exceptionally high concentrations of PCBs and dioxins in typical human milk1 and also the finding in a Dutch study that the daily toxic-equivalency intake of PCBs and dioxins per kilogram of body weight is 50 times higher in breast-fed infants than in adults.220  Results concerning dioxins, compatible with the above, were stated in U.S. and German studies.218,219  It appears that PCBs and dioxins received via breastfeeding are a perfect fit when looking for environmental toxins that

  a) are closely linked with diabetes (Appendix A),

  b) are especially heavily ingested by very young children and much less so by adults, and

  c) have been ingested by infants in greatly increasing amounts in recent decades, in line with increases that have taken place in diabetes and other diseases in children. (In, see Figure 2 left re increasing exposures in U.S. and refs. no. 142, and 159 re Europe.)


Aside from PCBs and dioxins in human milk, It would be surprising if any other recognized possible causes of diabetes could be named that qualify regarding any two of the above important considerations, not to mention all three.


4.c.2:  No other suggested risk factors for childhood diabetes are considered to be as strong:


The link between type 1 diabetes and toxins including PCBs and dioxins should be seen in light of what has been said about other risk factors for this disease.  Substantial searches in both PubMed and Google Scholar for “strong risk factors for type 1 diabetes” bring up no study in which the term “strong” or an equivalent descriptor is used in relation to a non-genetic risk factor for type 1 diabetes.  So the authoritative statement of “striking dose-response”  relationship between diabetes and toxins including PCBs and dioxins185 seems to set those toxins well above the rest of the group of suggested risk factors for this disease.

Given the above, there would seem to be ample reason for serious concern about the undisputed presence in human milk of PCBs and dioxins in concentrations that exceed established safe levels many times over.1  Nevertheless, the AAP apparently ignores the matter.  The word “toxin” does not appear anywhere in the AAP’s 2022 policy statement on breastfeeding.11,12



4.c.3:  The highest incidence rates and increases of childhood type 1 diabetes have been in the parts of the world where exposures to PCBs and dioxins have been highest: 


The highest incidence rates and increases of childhood type 1 diabetes in the world have been found in northern Europe, and many other countries in the northern hemisphere have very high rates of that disease, but no countries in the southern hemisphere have very high rates of type 1 diabetes.198  In relation to that, we should bear in mind that

  a) more than 80% of the world’s industrialized regions that are sources of dioxins and PCBs are located in the northern hemisphere,234 

  b) there is a very strong link between PCB and dioxin exposures on the one hand and diabetes on the other hand (Appendix A), and

  (c) type 1 diabetes began to increase when PCBs and dioxins were greatly increasing in many environments (ref. 217 and part 3 of   


Researchers have also observed that incidence of type 1 diabetes seemed to be higher in the northern than in the southern part of North America.198  This is compatible with differences in breastfeeding, the pathway by which developing children are most heavily exposed to PCBs and dioxins.1  According to OECD breastfeeding data, looking at the only metric (exclusive breastfeeding for three months) that is provided for all three principal countries of North America, Canada’s breastfeeding rate is about 50% higher than that of the U.S., and Mexico’s rate is about 70% lower than that of the U.S.195  This, combined with data for breastfeeding differences within the U.S.,237 forms a high-to-low, north-to-south progression of breastfeeding rates, which is compatible with the reported north-south differences in type 1 diabetes in North America.



4.c.4:  Links between childhood diabetes and breastfeeding in recent decades, possibly with no exceptions:


Finland, Sweden, and Norway, all very northern, have all been at the world’s highest level of incidence and increases of childhood type 1 diabetes.198,236  Those countries have also all stood out in breastfeeding, having had increases and incidences of that practice that have been among the very highest in Europe.206,207,208,209 (also see chart below.)  


Fig. 4




What appears to be a downturn in Finland’s data after 1985 is deceptive -- the apparently declined data points represent exclusive breastfeeding, whereas the earlier figures include partial breastfeeding.201 


Increases then stability, first in breastfeeding and later in childhood diabetes:


There were many years of increases in childhood type 1 diabetes in Finland, Sweden, and Norway, followed by a stabilized trend in that disease in those countries during the 2000s.236,192  Similarly, breastfeeding had increased greatly in those same countries in earlier decades;207,208,209 and breastfeeding also became stable in those countries by the late 1990s. (see chart above)



4.c.5:  Another principal pattern:  initial low levels in both breastfeeding and childhood diabetes, then rapid increases, first in breastfeeding and later in childhood diabetes: 


When discussing type 1 diabetes reported in European populations, the authors of a 2019 study pointed out, “…areas beginning with moderate-low incidences... exhibit a steep escalation of their annual incidences….192  This relationship was also observed by another study team.242  Looking at Figure 4 above, it is easy to see that same kind of pattern becoming apparent in breastfeeding:  a large number of countries with relatively low breastfeeding rates as of the early 1990s, but those rates steeply escalating after 1992.  Consider the compatibility between the post-1992 rapid increases in lactational exposures to PCBs and dioxins1 and the later rapid increases in type 1 diabetes,192 bearing in mind that type 1 diabetes is a mainly childhood disease for which PCBs and dioxins are strong risk factors. (Appendix A)




4.c.6:  Continuous global increase of type 1 diabetes, not satisfactorily explained:


A 2021 study reported that “despite over 40 years of investigation, with multiple, international case-control, cohort, and prospective studies, we are still in search of those critical environmental triggers for type 1 diabetes.”  Another study summarized that although some studies suggest associations (of environmental factors) with the risk of Type I diabetes, their causative role has not been shown.  It is also difficult to show that any of these environmental factors has changed in such a way that a continuous global increase in the incidence of Type I diabetes would be easily explained. 242 (italics added)


Therefore it is of interest to note that, as described elsewhere, exposures to PCBs and dioxins (especially via lactation1) have increased very considerably during recent decades;2 the increase in the breastfeeding that transfers large amounts of those toxins to infants has been so substantial and so extended that it could largely explain a continuous global increase in the incidence of Type I diabetes.  And especially large and/or rapid increases in type 1 diabetes have been associated with especially large and/or rapid increases in breastfeeding (part 4.c.8).  A good case can be made that the strong associations of type 1 diabetes with earlier lactational exposures to PCBs and dioxins are causal -- see Appendix A.


4.c.7: Explanations that have been suggested for the geographic differences in type 1 diabetes prevalence have been deficient, showing the need for a better explanation:   



A research team that pointed out many geographic differences in incidence of diabetes stated that, for various reasons, the differences cannot be explained on the basis of differences in ascertainment.198


When considering hypotheses that have been suggested about factors that might explain geographic differences in child diabetes (northern latitude, ethnic/genetic susceptibilities, etc.), the countries whose data supports or conflicts with those hypotheses are of interest because of what they have in common.  Data for the conflicting examples, as well as for the majority countries, all seems to point to high or increasing breastfeeding as a common characteristic that is linked with high or increasing diabetes.  See below.


High diabetes seems to follow high breastfeeding, whether the latitude generalization applies or does not apply: 


The principal Scandinavian countries have already been discussed in this regard. (see part 4.c.4 as far as Figure 4).  


The second highest incidence of child diabetes in the world has been found in Sardinia,198 despite that island’s being in the far south of Europe; so it is of interest that Sardinia is exceptional in having unusually prolonged breastfeeding. (See Appendix B


Kuwait has exceptionally high childhood diabetes among Asian countries even though it is not at all northern,200 and Kuwait’s breastfeeding rates are also the highest in its region.228


Low diabetes, low breastfeeding:  A 1993 study reported Northern Greece to be the place with the lowest type 1 diabetes incidence in Europe.198  And exclusive breastfeeding at 6 months has been found to be at only 0.7% among Greek women, less than one-tenth as high as the equivalent breastfeeding rate in most European countries.222


 Estonia is genetically similar to Finland as well as also of northern latitude, but it has incidence of childhood Type 1 diabetes less than a third as high as that of Finland,198 as reported in a 1993 study.  As of the early 1990s, Estonia had a 6-month breastfeeding rate only a third as high as that of Finland. (see Figure 4).159



A 2001 study compiled data on childhood type 1 diabetes in 31 countries of Europe, and the Former Yugoslav Republic of Macedonia was reported to have the lowest diabetes incidence of all 31 countries, at 3.2 cases per 100,000, compared with 21.2 and higher in Finland, Norway and Sweden.204  So it is of interest that a 2000 WHO document on infant feeding in Europe pointed to only one country at the low end of the breastfeeding rate spectrum; that low-end breastfeeding country was the Former Yugoslav Republic of Macedonia.226


Expanding the scope to include the entire world, the country with the lowest type 1 diabetes rate reported for any nation was Venezuela.200  So it is of interest that, in World Bank data, Venezuela’s breastfeeding percentage was almost at the bottom.  The few countries that were even lower in the breastfeeding data were not among the ones for which diabetes data had been published, with the single exception of the Dominican Republic;231 the Dominican Republic was near-bottom in world breastfeeding and fourth from the bottom in the world diabetes data.200  So, of the countries for which type 1 diabetes was published, the very lowest and the fourth from the bottom both had breastfeeding rates that were at the bottom level in world breastfeeding data.





4.c.8:  Many especially large national increases in type 1 diabetes, with especially large corresponding increases in breastfeeding: 


Type 1 diabetes tripled among the youngest age group (0-<5) in Austria between 1985 and 2005.232  So it is of interest that breastfeeding at 6 months in Austria also tripled between 1985 and 2005 (see Figure 4 earlier).


A 1999 study pointed out that, among the countries of eastern Europe, Hungary had the most rapid increase of type 1 diabetes (8.5% per year);242 the disease tripled in the youngest age group over the period 1989-2009.230    Therefore it is of interest to see the extraordinarily rapid increases of breastfeeding in Hungary in Figure 4 earlier.


The average standardized incidence of type 1 diabetes in Polish children more than tripled between 1989 and 2004, from 5.4 to 17.7.223  So it is of interest that breastfeeding at 6 months in Poland increased from 13% in 1988 to 60% in 1997.224


  The average incidence rate of type 1 diabetes among 0-4-year-old Estonian children rose from 3.6 to 15.7 per 100,000 between 1983-1990 and 1999-2006.229  Notice in Figure 4 how compatible that dramatic increase of diabetes is with the increase of breastfeeding in Estonia over the relevant years.


Other remarkably rapid increases of type 1 diabetes (5% to 8% per year) occurred in the Czech Republic, Slovakia, and Romania in 1989-2003:223   In years that would have been relevant to those diabetes increases, there were similar remarkable increases in breastfeeding at 6 months in those same countries:  from 15.5% in 1980225 to 37% in 2002 in Slovakia,159 from 9% in 1993 to 32% in 2002 in the Czech Republic;159 and from an estimated 12% in the early 1980s227 to 35% in 2001 in Romania.159



4.c.9:  Exceptional decrease in type 1 diabetes, and exceptional matching decrease in breastfeeding: 


According to a 2011 study of recent international data for type 1 diabetes, observations from Sweden indicated that, contrary to analyses in most other countries, there were signs of “declining rates in the youngest children in recent years, in particular those born after the year 2000.” 230  Another study team made essentially the same observation.246  So it is of interest to notice in Figure 4 earlier that, for the post-1995 period, there was only one decline of breastfeeding that was noteworthy, standing out both in magnitude and in the fact that the decline was visible in multiple data points:  that decline was in Sweden, beginning slowly in 1999 and accelerating in 2004.  That makes a good fit with the very unusual decline of type 1 diabetes among Sweden’s youngest children after 2000.


4.c.10:  An early upturn and dramatic increase in Sweden, first in breastfeeding and later in type 1 diabetes, separated by a predictable lag:


Breastfeeding at 6 months in Sweden more than tripled, from 6% in 1972 to 20% in 1976/77 and continued to increase to 50% in the 1980s.208,159  Given the toxins that are both strongly associated with diabetes185 and also present in high concentrations in human milk,1 that breastfeeding history was very compatible with what subsequently took place in Sweden in type 1 diabetes:  incidence of that once-rare disease “began to increase drastically in the 1980s,” according to Swedish researchers.239   The lag of a decade or so after the steep upturn in breastfeeding, before type 1 diabetes dramatically increased, was in line  with what could have been expected; onset of that disease occurs especially around puberty.  



4.c.11:  Other examples of type 1 diabetes varying in the same direction and at compatible times with variations in breastfeeding:


 An expert reporting on the international increases of type 1 diabetes pointed out that, although the increases had been predominantly in Europe, some increases had recently been reported outside Europe, mentioning only one example:  Kuwait, with incidence rate of seventh highest in the world.  Therefore it is of interest to note that this exceptional country in diabetes increase and incidence has also stood out in its breastfeeding increase and incidence rate:  

  -- exclusive breastfeeding at six months increased six-fold in Kuwait since 2006/2007 (as reported in a 2021 study), and

  -- Kuwait’s breastfeeding rate at six months was reported to be highest among the six member states of its region (the Gulf Cooperation Council).228


Iceland is the northern-most nation of Europe, but it has relatively low child diabetes, thus conflicting with the hypothesis about northern latitude and low temperatures leading to diabetes.  But, unlike the other Scandinavian countries, with their consistently very high breastfeeding rates (see Figure 4), 6-month breastfeeding in Iceland has been only irregularly high,159 and its exclusive breastfeeding has been about 30% lower than that of Finland and Sweden;201 also important, Iceland has very low levels of the pollution that has been entering human milk in most developed countries since mid-20th century.202,211 its population density is very low, and it has no neighbor countries from which to receive pollutants via wind transport. 


Ethnic Differences:  A study conducted at five locations in the U.S. between 2002 and 2012 found that type 1 diabetes among non-Hispanic white youths increased 1.2% annually during those years, while it increased at 4.2% and 2.2% per year among Hispanic and black youths.243  Therefore it is of interest that breastfeeding at six months doubled among non-Hispanic whites between 1989 and 2007, whereas it approximately tripled and quadrupled among Hispanics and blacks, respectively, during that same period.244  The increases in diabetes were not closely proportionate to the increases in breastfeeding at six months -- there are obviously other variables that cause differences, including exclusiveness of breastfeeding and ethnic differences in susceptibility to effects of toxins; associations between type 1 diabetes and exposures to toxins including PCBs and dioxins have been found to be particularly strong among Mexican Americans.185  In any case, the above examples provide support for a generalization to the effect that large increases in exposures to a potent source of PCBs and dioxins1 are associated with later increases in childhood type 1 diabetes; and especially large increases in exposures to a potent source of PCBs and dioxins are associated with especially large later increases in diabetes.  For reasons why those associations could reasonably be regarded as causal, see Appendix A.



To see the complete article to which the above is a supplement, as well as author information, go to






Appendix AAdditional evidence linking PCBs and dioxins (and therefore their major pathway to infants -- breastfeeding1) with diabetes:


Type 1 diabetes is essentially a malfunctioning of the immune system; and increasing evidence implicates inappropriate activity of the immune system in the origins of type 2 diabetes.214  Therefore a discussion of increases in diabetes should consider the strong evidence of adverse effects on the immune system of PCBs and dioxins, which are present in typical human milk in high concentrations:1

According to a report written by a nine-member team for the European Commission Directorate-General for the Environment, dioxins "produce a vast number of effects on the immune system, often at very low doses."2b  According to an EPA document, PCBs "suppress the immune system;" and animal studies were unable to identify a level of PCB exposure that did not cause effects on the immune system.46  


Substantial other evidence has also indicated that PCB and dioxin exposures are linked with greatly increased risk of diabetes.190, 20,184   A 2006 study, by a team that included five MD degrees and 7 PhD degrees among the group, investigated the relationship between risk of diabetes and serum concentrations of a mixture of six organic pollutants including three types of PCBs and dioxins.  Their research was based on the 1999-2002 U.S. National Health and Examination Survey. The study team found “striking dose-response relations between serum concentrations of (the pollutants) and the prevalence of diabetes.” 185  We should note that dose-response relationships in general are recognized to provide good evidence of causality, when considering the report that dose-response relationships were “striking” in this particular case.  Of the six pollutants studied, PCB exposures showed by far the strongest correlation with diabetes.  The authors said that their finding was basically consistent with earlier prospective studies, despite some differences, citing seven references for that.  They also noted that “the idea that dioxin exposure may cause diabetes is in line with the known biology of these pollutants.”


The especially strong dose-response relationships found in this study should be seen in light of the following:

  a) this study is apparently the first to investigate effects of mixtures of these toxins, as opposed to effects of individual toxins; exposures to mixtures is clearly much more similar to what is normal for general populations, compared with exposures to individual toxins;

  b) this study focused attention on dose-response effects at lower (background) concentrations of toxins, compared with most previous epidemiological studies’ devoting attention mainly to differences in high (occupational) exposures; the authors noted reasons why effects of extended differences in exposures at low concentrations may be especially significant.


The authors also provided five reasons why the strong relationship between these toxins and diabetes should be considered to be causal, including that (a) their finding is basically consistent with multiple other studies, and (b) the idea that dioxin exposure may cause diabetes is in line with the known biology of those pollutants.


A 2020 study team summarized the research on this topic thus far saying, “Dioxin and dioxin-like pollutants have consistently been associated with an increase in diabetes risk in humans.” 240




Appendix BSardinia’s exceptional breastfeeding and type 1 diabetes:

In a study of breastfeeding across geographical areas in Italy, breastfeeding exclusiveness and duration in the various regions was assessed according to three different criteria, and Sardinia was the highest region according to all three measures.205  In other studies, Sardinia was the only geographic location mentioned as having unusually prolonged breastfeeding, including when characteristics and practices were summarized internationally.213   

 It might not be random coincidence that Sardinia has been reported to have the second highest rate of type 1 diabetes internationally and also type 1 diabetes rates 3-5 times higher than in continental Italy.233

Sardinia is authoritatively reported to have distinct language and culture, and “the relative isolation of the island has preserved many traditions;” 202 this may explain why breastfeeding continued to be notably prolonged in Sardinia into modern times, even though breastfeeding beyond 12 months has been rare in the rest of Europe.215




    1) van den Berg et al., WHO/UNEP Global Surveys of PCDDs, PCDFs and DDTs in human milk and benefit-risk evaluation breastfeeding, Arch Toxicol. 2017, at  

   For considerable additional scientific basis for seeing PCBs, dioxins and PBDEs in human milk in the 2000s to be far greater than established safe levels, see references no. 23, 37h and 37j in See ref. 37h at that link re PBDEs.


   2) See part 3 and Figure 2 and accompanying text in .


  159) Go to Click on HFA Explorer.  On the left, choose Line chart, and uncheck the indicator categories other than HFA.  Under Filter through the subjects, click on Pick a subject, and scroll to the bottom and click on Deselect all.. Then scroll back up and choose "Maternal and child health."  Then under Indicators select Proportion of infants breastfed at (6) or (3) months. Then Select countries. Finally, click on "Line chart" at the bottom.  Pass the cursor across the chart to see data identified by country. (Identification of countries may require entering only a few countries at a time.)  Note: A much more easily accessible source for most of the same data can be found in Figure 4.


  185) Lee D-H et al., A strong dose-response relation between serum concentrations of persistent organic pollutants and diabetes: Results from the National Health and Nutrition Examination Survey 1999-2002, at


  192) Xia Y et al., Incidence and trend of type 1 diabetes and the underlying environmental determinants, Diabetes Metab Res Rev, 2019, at    

  See p. 5.  Discussing type 1 diabetes incidence, refers to “stabilization after years of the upward trend” In Norway, Finland and Sweden “since the first decade of the 21st century.”

  When reporting the steep escalation of annual incidences of type 1 diabetes in many countries, this 2019 article used the present tense and was clearly mainly discussing increases that had been occurring during relatively recent years.


  195) OECD Family database, Social Policy Division - Directorate of Employment, Labour and Social Affairs, Chart CO1.5.B, at

  198) Karvonen M et al., A review of the recent epidemiological data on the worldwide incidence of Type 1 (insulin-dependent) diabetes mellitus, Diabetologia, 1993, at

  Note that increase is discussed in the section, Temporal trends in incidence, and incidence is discussed elsewhere.

  200) The Diamond Project Group:  Incidence and trends of childhood Type 1 diabetes worldwide 1990-1999, Diabetic Medicine, 2006, at  See Figure 1 re Venezuela lowest.

  201) Yngve A and Sjostrom M, Breastfeeding in countries of the European Union and EFTA:current and proposed recommendations, rationale, prevalence,duration and trends, Public Health Nutrition, 2001

  202) Tamanini and Valenciano, The Global Green Economy Index, GGEI 2016, Figure 1, p. 14, at

  204) Patterson CC et al., Is childhood-onset Type 1 diabetes a wealth-related disease? Diabetologia, 2001, at

  205) Giovannini M et al., 22. Breastfeeding Across Geographical Areas in Italy, Pediatric Research (2005) at

   206) Kocturk T and Zetterstrom, Thoughts about rates of breastfeeding,  Acta Pediatrica, 2007,


  207) Ludvigsson J, The jury is still out on possible links between cows’ milk and type 1 diabetes, Acta Paediatrica, 2019, at


  208) Hofvander and Sjolin, Breast feeding trends and recent information activities in Sweden, Acta Paediatrica, 1979, at


  209) Grovslien and Gronn, Donor Milk Banking and Breastfeeding in Norway, J Hum Lact, 2009,  Figure 1

  211) Iceland in figures 2018, Statistics Iceland 2018, at   See p. 19, Gross domestic product by industries.

  212) Encyclopedia Britannica article, Sardinia island, Italy, at

  213) Pes et al., Why are Sardinians the shortest Europeans?  A journey through genes, infections, nutrition, and sex;  Am J Physical Anthropol, 2017, at  Page 9, below Figure 2.

  Poulain M et al., A Population Where Men Live As Long As Women: Villagrande Strisaili, Sardinia, Journal of Aging Research, 2011, above Figure 2, where the discussion is about international conditions and practices, at  

  Comparison with other Italian regions:  Raftakis M, Why were infants dying and what were they dying from? Infant mortality patterns in the Greek urban centre of Hermoupolis, Syros (1860-1940), The History of the Family, 2021, at  Part 2, p. 407

  214) Pickup JC, Inflammation and Activated Innate Immunity in the Pathogenesis of Type 2 Diabetes, Diabetes Care, 2004, at

  216) Rogers MAM et al., Fluctuations in the incidence of type 1 diabetes in the United States from 2001 to 2015: a longitudinal study, BMC Medicine, 2017, at

  217) Gale EAM, The Rise of Childhood Type 1 Diabetes in the 20th Century, Perspectives in Diabetes|,  at

  222) Iliadou M et al., Measuring the Effectiveness of a Midwife-led Education Programme in Terms of Breastfeeding Knowledge and Self-efficacy, Attitudes Towards Breastfeeding, and Perceived Barriers of Breastfeeding Among Pregnant Women, Mater Sociomed, 2018, at


  223) Jarosz-Chobot et al., Rapid increase in the incidence of type 1 diabetes in Polish children from 1989 to 2004, and predictions for 2010 to 2025, Diabetologia, 2011, at


  224) Anderson JK et al., The training needs of doctors working in England and in Poland with breastfeeding women, Archives of Medical Science, 2007, at,19,9012,1,1.html


  225) Velkova S, History of breastfeeding in Slovakia until Corona crisis, Int J Health New Tech Soc Work 2022, Table 2, at


  226) Fleischer-Michaelsen K et al., Feeding and Nutrition of Infants and Young Children:  Guidelines for the WHO European Region with Emphasis on the Former Soviet Countries, Chapter 1, p. 28 ff., at  To summarize some of that publication’s contents:  In the 1980s in the member states of the Soviet bloc (which included Estonia and Poland), breastfeeding was encouraged but going beyond partial breastfeeding was strongly discouraged.

  Re Macedonia, see p. 31.

   227) There appears to be no published data about breastfeeding in Romania in the 1980s, but a reasonable approximation of 12% can be obtained from data for its two neighbors, Hungary and Poland, which had 6-month breastfeeding rates of 12.7% in 1980 and 11% in 1985 respectively;159 low rates of breastfeeding at 6 months were apparently normal in the 1980s in the then member states of the Soviet bloc, which included all of the above countries.226


  228) Alwaiian T et al., Factors Associated with Exclusive Breastfeeding in the State of Kuwait, Research Square, 2021,



  229) Teeaar T et al., Increasing incidence of childhood-onset type 1 diabetes mellitus among Estonian children in 1999-2006. Time trend analysis 1983-2006, Pediatric Diabetes, 2010, see Fig. 1, at  .


  230) Gyurus EK et al., Twenty-one years of prospective incidence of childhood type 1 diabetes in Hungary, Pediatric Diabetes, 2011, at 

  The overall increase in type 1 diabetes was 136% (increasing from 7.7 to 18.2 per 100,000), with a mean annual increase of 4.4%, as stated by the authors. In the youngest age group, the rate of increase was 6.2%, 41% higher than the overall increase,  136% x 1.41 = 192% increase in the youngest age group

   Re recent decrease in Sweden: See p. 21. The recent decrease in Sweden is also reported in Diabetes in America. 3rd edition. Chapter 2:  Prevalence and Incidence of Type 1 Diabetes Among Children and Adults in the United States and Comparison with Non-U.S. Countries, at


  231) World Bank Exclusive breastfeeding (% of children under 6 months) at

  232) Schober E et al., Steep increase of incidence of childhood diabetes since 1999 in Austria. Time trend analysis 1979-2005. A nationwide study, European Journal of Pediatrics, 2007, Figure 1, at


  233) Maahs et al., Chapter 1: Epidemiology of Type 1 Diabetes, Endocrinol Metab Clin North Am., 2010, at

  234) Fiedler H, Ch. 2. Polychlorinated Biphenyls (PCBs): Uses and Environmental Releases, in Polychlorinated Biphenyls (PCBs), 2009, p. 12, at


  236) Tamayo T et al., Diabetes in Europe: An update, IDF Diabetes Atlas, 2014, especially part 2.2.1, at


  237) CDC map of US counties by breastfeeding initiation rates, at

  239) Diabetologia. "Sweden has 2-3 times as many adolescents, young adults with type 1 diabetes as previously thought." ScienceDaily, 6 April 2014. <>.

   240) Hoyeck MP et al., Long-term metabolic consequences of acute dioxin exposure differ between male and female mice, Sci Rep., 2020,




  242) Onkamo et al., Worldwide increase in incidence of Type I diabetes -- the analysis of the data on published incidence trends, Diabetologia, 1999, at


  243) Lawrence JM et al., What do we know about the trends in incidence of childhood-onset type 1 diabetes? Diabetologia, 2019, doi:10.1007/s00125-018-4791-z


  244) Data from U.S. Surgeon General’s report13 shows 6-month breastfeeding rates to be 44.7%, 46% and 27.5% for non-Hispanic whites, Hispanics, and blacks, respectively among children born in 2007.

  Table 3-2 in Ch. 3, Who Breastfeeds in the United States?  Nutrition During Lactation,.Institute of Medicine (US), Committee on Nutritional Status During Pregnancy and Lactation, National Academies Press (US); 1991, at shows breastfeeding rates at 5-6 months of 22.7%, 15.0%, and 7.0%, for non-Hispanic whites, Hispanics, and blacks, respectively, among children born in 1989.


  245) Quinn LM et al., Environmental Determinants of Type 1 Diabetes: From Association to Proving Causality, Frontiers in Immunology, 2021,at

  246) Berhan Y et al., Thirty Years of Prospective Nationwide Incidence of Childhood Type 1 Diabetes, Diabetes (journal), 2011, at







Temporal trends in incidence