Nephrogenic Diabetes Insipidus in North America. The Hopewell Hypothesis

Title: Nephrogenic Diabetes Insipidus in North America. The Hopewell Hypothesis
Authors: Bode, Hans H.; Crawford, M.D., John D.
Publisher: New England Journal of Medicine
Date Published: January 01, 1969
Reference Number: 292

Reprinted with permission from Massachusetts Medical Society, 1969, Vol. 280(14):750-4 for educational use within the NDI community. No part of this article may be reproduced in any way without permission in writing from the publisher.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.

Nephrogenic diabetes insipidus was called to medical attention less than 25 years ago, when Waring et al.1 described six cases. Subsequently, Williams and Henry,2 noting the inability of these patients to form concentrated urine even after infusion of antidiuretic hormone in doses large enough to cause systemic side effects, proposed the theory of end-organ resistance. The presence of biologically active hormone in the serum and urine of affected patients has since been demonstrated by several workers.3-7

During the last 20 years a number of families with this disease have been seen at the Massachusetts General Hospital; although claiming no relation to one another, all have been descendants of early settlers in Nova Scotia. This kindled our interest to search historical records for an individual or single group of settlers who might have brought the defect to this continent. Previously reported pedigrees could not be explored far enough into the past to determine the ethnic origin and the port of entry of the progenitors, but several kindreds, like our own, led to the colonists of Nova Scotia.8,9 Local investigation soon proved that nephrogenic diabetes insipidus is still a rather common disorder in that maritime province. So-called "water drinkers" have been well known by the lay population, especially in Colchester County, ever since the area was settled by "Ulster Scotsmen" after the evacuation of the French in 1754.

According to the reports of Eaton10 and Miller,11 these immigrants were descendants of Scottish Presbyterians who, during the reign of James I of England (1603-1633), had tried to escape oppression by migrating to the Ulster Province of Ireland. In their new homes they preserved their folklore and tradition, and the clan was soon known as the Ulster Scott. Even today their descendants support their own historical society in Belfast, Ireland. In 1718 several of these families left for the new world; upon their arrival in Boston, however, being Presbyterians, they were refused permission to stay. Nonetheless, land was granted them for colonization at the northern border of Massachusetts in the still largely unexplored Merrimac Valley, now part of New Hampshire. There, they founded the settlements of Londonderry, Windham and Westchester, and graves of early settlers can still be found in the old cemetery at Derry.

The second wave of immigration followed the lure of Colonel Alexander McNutt, a government agent. The first of the ships to land in Halifax, the Hopewell, arrived on October 9, 1761, with somewhere between 200 and 300 passengers, depending on the historian.10,11 Many of these settled in Colchester County, Nova Scotia. McNutt also succeeded in reuniting a number of the prior group of Ulster-Scot immigrants with their newly arrived relatives in Colchester County when he resettled several families from the Merrimac Valley.

It is difficult to believe that a male affected with nephrogenic diabetes insipidus could have survived a journey over the ocean in the early eighteenth century. We read in the records11 that the ships were "scantily supplied" with food and water. On one occasion, at least, the passengers "were reduced to the necessity of eating the bodies of those who died. Even this resource failed them; and, at length, Mr. Fisher was selected to give up his life to preserve the lives of the rest." Providentially, he was saved by the arrival of unexpected help from another ship, and the record does not neglect to emphasize his later aversion to the sight of wasted food or water.

It seems quite probable that the trait for nephrogenic diabetes insipidus was brought to this continent by passengers aboard the Hopewell, and there is a suggestion, at least, that affected males were on the ship. One family, in which the disorder has cropped up in many descendants (Fig. 1), lost three children during the voyage, whereas their only daughter born before the trip survived. The circumstances of the deaths of those children are lost, but conceivably, they died of causes other than their special need for water, for Eaton10 states that the passengers, upon arrival in Halifax, were not allowed to disembark for several days because of a quarantine against small pox.

FIGURE: 1. Simplified Pedigree of Nova Scotian Settlers Arriving aboard the Hopewell.

The sibship of generation II plus four of the children of the sibship giving rise to the Fulmore and Morrison branches were on the Hopewell. Three of these children died during the voyage. Several of the Fulmore offspring were reputed to be "water drinkers." The diagonally shaded areas indicate that four fraternal nephews of the three brothers with nephrogenic diabetes insipidus in Generation IV were affected with the disease, suggesting at first gene transmission through the unaffected fathers. In each case, however, it was possible to establish consanguinity and thus the likelihood that the gene was transmitted through carrier females. The horizontally shaded areas in generations VI and VII indicate affected males who fathered sons reputed to be "water drinkers." These three sons were examined and shown to have normal concentrating ability.

Detailed historical and genealogic records of these early Nova Scotian settlers and their destinies are preserved in the folklore and in local writings, which contain quite accurate descriptions of the clinical picture of nephrogenic diabetes inspidus.12 The legend of the "water drinkers' curse" provides a delightful account of the origin of the disease, as well as an astonishingly precise description of its mode of inheritance. Briefly paraphrased, the legend runs as follows:

A gypsy woman and her son were traveling the road and became thirsty. Pausing at a well in front of the next house, the gypsy requested water for her son; the housewife refused, whereupon, the gypsy woman cast upon her a curse. Henceforth, the story goes, the woman's sons would be afflicted with a craving for water. The curse would be passed on by her daughters and revisited upon their sons for generations to come.

Even today there is reluctance to discuss the imprecation, for so surely did it affect the lives of the designated descendants that few of the present inhabitants of the area are unfamiliar with the malediction or its victims. In many cases their affliction is so damaging that "water drinkers" presently constitute more than 1 per cent of the male inmates of the Nova Scotia Training School. Mental retardation, a complication of hypertonic dehydration in infancy,13 was found in 39 of 44 patients evaluated at the Massachusetts General Hospital.

Already in the first medical report, Waring1 had suspected that he was describing an X-linked inherited disease, for two of his six males were brothers. Forssman,14 as well as several other authors,2,8,9,15,16 subsequently reported pedigree patterns compatible with this proposal. It was challenged, however, when Dancis3 described a severely affected female and again when Cannon17 reported a large Mormon pedigree with 679 subjects, including 55 affected males, 11 symptomatic females and 21 carriers. Cannon observed apparent male-to-male transmission on six occasions and suggested an autosomal dominant mode of inheritance with incomplete penetrance in the female. This interpretation was considered a possibility by Childs and Sidbury18 and was favored by Robinson and Kaplan.19

Re-examination of published data, together with an analysis of our own pedigrees, suggests that the original proposal of X-linked inheritance1 is both correct and in keeping with the traditional gypsy's tale. In Cannon's pedigree all six cases of male-to-male transmission occurred in early generations in which he was forced to rely on historical data for the diagnosis as well as for the exclusion of consanguinity. Except for these six transmissions his pedigree conforms to the rules of X-linkage.

The unreliability of recollection and historical data for the diagnosis of nephrogenic diabetes insipidus is illustrated in the large pedigree covering nine generations shown in Figure 1. Here, male-to-male transmission is suggested on five occasions. In two cases (Fig. 1, Generations VI-VII), however, the supposedly affected sons were shown to have normal urinary concentrating ability. In the three others (Fig. 1, Generations IV-V), consanguinity of the parents was documented. The almost universal lack of social restriction in the period of the early settlements, as well as the isolation of the Mormons and the Ulster Scotsmen, predisposed to a high rate of intermarriage. In addition, information on both Cannon's pedigree and that shown in Figure 1 is incomplete, and in neither case does it rule out the presence of more than one heterozygote in the oldest generation. This is, in fact, the likely explanation for the appearance of several apparently unrelated affected males born in different settlements of Colchester County before 1800.

X-linked gene transmission would dictate that all sons of affected males should be free of the defect and all daughters heterozygous. Accordingly, it would also be expected that half the offspring of heterozygous females should be gene carriers and that the other half should consist of healthy subjects. The published literature does not contain sufficient data on the descendants of prolific males affected with nephrogenic diabetes insipidus to test these propositions. In families studied by us, however, the progeny of eight affected males consisted of 12 sons, all of whom were healthy, and 25 daughters. Of the latter, 21 submitted to examination and were shown to be heterozygous. Similar studies were done on 221 offspring of 53 heterozygous females with the findings of 121 who were either affected males (70) or carrier females (51).

It is appropriate to pause here to explain that heterozygous females can be detected by means other than the geneticist's approach through studies of their progeny. The great majority of carrier females manifest a partial concentrating defect. This was first demonstrated when Carter and Simpkiss15 showed that carriers identified on genetic grounds, when subjected to 12 hours of fasting and thirsting, were unable to elaborate urine having a specific gravity above 1.018. In studies performed on several of our genetically determined heterozygotes similar results were obtained (Fig. 2). It is evident, however, that discrimination by this means is not perfect since there is some overlap between the carriers and the normal controls.

FIGURE 2. Concentration Ability of Males and Females Carrying the Defective Gene, Showing Urinary Osmolality and Specific Gravity after Water Deprivation Either for 12 Hours or, in the Severely Affected Persons, for the Maximum Time Tolerated.

The control urines were collected from spouses of the gene carriers, all of whom were easily able to tolerate water deprivation of 12 hours. The graph demonstrates the decreased average concentrating ability in the heterozygous female as well as the fact that neither specific gravity nor osmolality estimations permit a clear separation of carrier and normal females.

This overlap and the variability in concentrating power of the heterozygous female can be explained by the Lyon20 hypothesis of random X chromosome inactivation. This proposal has recently has been strongly supported by the observation of a girl affected quite as severely as any of the males.21 She was found to have a deletion of one X chromosome and thus was subject to selective X chromosome inactivation.22

The Ulster Scotsmen were well aware of the occasional symptomatic female whose polyuria they correctly portrayed as more pronounced after puberty. Their explanation, ascribing this to the greater susceptibility to suggestion of the weaker sex, is expressed in the following lines from Mrs. Rogers's novel12 about the "water drinkers": ". . . [The women] had heard about it and kind of fell into the habit unconscious-like. They don't need to have it, for the curse was only on their men."

Descendants of the early immigrants are now spread over the entire continent. The families of some of them who have helped so much in our study had migrated as far afield as Alaska and Florida. The genealogic data that they have contributed give strong support for the speculation that most, if not all, of the persons with nephrogenic diabetes insipidus are among the posterity of the original settlers in Nova Scotia. Furthermore, this information has enabled us to link the kindred reported by Walker9 in 1954 to one of our own families deriving from the Colchester area. A clue to the relation between the Ulster Scotsmen and Cannon's family is to be found in Miller's11 comment that a number of settlers had moved from Nova Scotia with the Mormons to Utah. The surnames of Cannon's oldest forebears, Williams, Fulmer and Marvin,23 are either identical or closely similar to those found among the early Nova Scotian settlers (Williams, Fulmore and Morrison — see Fig. 1), especially when it is taken into account that, at the time of the westward migration, illiteracy was widespread among both populations. The authors of a number of other pedigrees have been unable to trace back their kindreds to the period before the progenitors left the east coast. Perry,8 however, recently reported a kindred residing in Manitoba, Canada, and although these people lacked records to trace the family's origin, they proved to be well aware of the "water drinkers' curse."

Intriguing as is the proposal of a common origin of all North American families affected by nephrogenic diabetes insipidus, it is likely the "Hopewell hypothesis" can never be proved. Evidence is overwhelming, nonetheless, that in all reported cases, the defect is due to an X-linked gene, and new mutations giving rise to the disorder appear to be excessively rare. Perhaps most noteworthy of all is the fact that one finds in the folklore of the Ulster Scot an accurate description of the pattern of inheritance formulated more than a hundred years before 1865, when Gregor Mendel proposed his laws of heredity.

We are indebted to Miss Genevieve Cole, librarian of the Massachusetts General Hospital, for assistance in locating and obtaining, for our examination, the published records of the Ulster Scotsmen and to Mr. Ross Hill, of Onslow, Nova Scotia, whose interest in our project and uncanny knowledge of his neighbors, their family inter-relations and the location of church records and tombstones of their forebears was no less essential to the preparation of this report.


  1. Waring, A. G., Kajdi, L., and Tappan, V. Congenital defect of water metabolism. Am. J. Dis. Child. 69:323, 1945.
  2. Williams, R. H., and Henry, C. Nephrogenic diabetes insipidus: transmitted by females and appearing during infancy in males. Ann. Int. Med. 27:84-95, 1947.
  3. Dancis, J., Birmingham, J. R., and Leslie, S. H. Congenital diabetes insipidus resistant to treatment with pitressin. Am. J. Dis. Child. 75:316-328, 1948.
  4. Holliday, M. A., Burstin, C., and Harrah, J. Evidence that antidiuretic substance in plasma of children with nephrogenic diabetes insipidus is antidiuretic hormone. Pediatrics 32:384-388, 1963.
  5. Linneweh, F., Buchborn, E., and Dellbrück, B. Familiärer renaler Diabetes insipidus. Klin. Wchnschr. 35:321-327, 1957.
  6. Luder, J., and Burnett, D. Congenital renal tubular defect. Arch. Dis. Childhood 29:44-47, 1954.
  7. Macdonald, W. B. Congenital pitressin resistant diabetes insipidus of renal origin. Pediatrics 15:298-311, 1955.
  8. Perry, T. L., Robinson, G. C., Teasdale, J. M., and Hansen, S. Concurrence of cystathioninuria, nephrogenic diabetes insipidus and severe anemia. New Eng. J. Med. 276:721-725, 1967.
  9. Walker, N. F., and Rance, C. P. Inheritance of nephrogenic diabetes insipidus. Am. J. Human Genet. 6:354-358, 1954.
  10. Eaton, A. W. H. The Settling of Colchester County, Nova Scotia, by New England Puritans and Ulster Scotsmen. Series 3, Vol. 6. Ottawa, Canada: Royal Society of Canada, 1912. Pp. 221-265.
  11. Miller, T. Historical and Genealogical Record of the First Settlers of Colchester County. Halifax, Nova Scotia: A. & W. MacKinlay, 1873.
  12. Rogers, G. D. (McLeod). Joan at Halfway. New York: G. H. Doran, 1919.
  13. Ruess, A. L., and Rosenthal, I. M. Intelligence in nephrogenic diabetes insipidus. Am. J. Dis. Child. 105:358-363, 1963.
  14. Forssman, H. On hereditary diabetes insipidus, with special regard to sex-linked form. Acta med. Scandinav. Supp. 159:1-196, 1945.
  15. Carter, C., and Simpkiss, M. "Carrier" state in nephrogenic diabetes insipidus. Lancet 2:1069-1073, 1956.
  16. Schoen, E. J. Renal diabetes insipidus. Pediatrics 26:808-816, 1960.
  17. Cannon, J. F. Diabetes insipidus: clinical and experimental studies with consideration of genetic relationships. Arch. Int. Med. 96:215-272, 1955.
  18. Childs, B., and Sidbury, J. B., Jr. Survey of genetics as it applies to problems in medicine. Pediatrics 20:177-218, 1957.
  19. Robinson, M. C., and Kaplan, S. A. Inheritance of vasopressin-resistant ("nephrogenic") diabetes insipidus. Am. J. Dis. Child. 99:164-174, 1960.
  20. Lyon, M. F. Sex chromatin and gene action in mammalian X-chromosome. Am. J. Human Genet. 14:135-148, 1962.
  21. Herrin, J. T. Personal communication.
  22. Rohde, R. A., and Berman, N. Lyon hypothesis and further malformation postulates in chromosomal syndromes. Lancet 2:1169, 1963.
  23. Cannon, J. F. Personal communication.

    *From the Shriner's Burns Institute and the Children's Service Massachusetts General Hospital, and the Department of Pediatrics, Harvard Medical School (address reprint requests to Dr. Bode at the Children's Service, Massachusetts General Hospital, Boston, Mass. 02114).

    Supported in part by grants (T01-HD-00033, HD-00243 and K3-HD-14,824) from the National Institutes of Health, United States Public Health Service.

This translation by the NDI Foundation is to assist the lay reader. To provide a clear, accessible interpretation of the original article, we eliminated or simplified some technical detail and complicated scientific language. We concentrated our translation on those aspects of the article dealing directly with NDI. The NDI Foundation thanks the researchers for their work toward understanding and more effectively treating this disorder.
© Copyright NDI Foundation 2007 (JC)

Bode and Crawford noticed that all the patients admitted to the Massachusetts General Hospital for nephrogenic diabetes insipidus (NDI) were descendants of early settlers of Nova Scotia. This prompted them to search historical records for an individual or single group of settlers who might have brought NDI to the North American continent. They concluded that a group of Scots from the Ulster Province of Ireland who originally arrived in Nova Scotia in 1761 aboard the ship, Hopewell, was the group that brought NDI to the continent.

The settlers brought with them a legend of "the water drinker's curse." A gypsy asked an Ulster Scot (then living in Ireland) for a drink to quench the thirst of her son and herself. When the Scot refused, the gypsy cursed her, saying her sons would all be born with a craving for water, and the curse would be passed on by her daughters whose own sons would have the curse.

Bode and Crawford note that the legend contains an astonishingly precise description of NDI's X-linked mode of inheritance. One hundred years before Mendel proposed his laws of heredity, the Scots observed that the disorder was carried by women who, though they did not express the full symptoms of the disorder themselves, passed it on to their sons, who did express the full symptoms.

The authors noted that a great majority of the carrier females show some symptoms of NDI. That is, many are not able to concentrate their urine completely. Bode and Crawford attribute this to the females having a skewed random inactivation of healthy X chromosomes.

Descendants of these early immigrants are now spread over the entire North American continent. Though the authors doubt that their "Hopewell Hypothesis" can ever be proved, they suggest it is possible that all the North American cases of NDI share a common origin in the Ulster Scot immigrants.