AUDIOLOGECAL MANIFESTATlONS IN JUVENILE-ONSET DSABETICS Thesis for the Degree of Ph. D. MiCHIGAN STATE UNWERSITY KENNETH R. JOHNSON ‘ ‘ 1.970 ' u; T A E ;? ‘C L IB R A R Midtigan State University This is to certify that the thesis entitled AUDI OLOGI CAL MAN I FESTAT I CNS I N JUVENILE - ONSET DIABETICS presented by KENNETH R. .IGINSON has been accepted towards fulfillment of the requirements for PH. D. degree in AUDIOLOGY AND % SPEECH SCIENCES (/WW .4fi4/é/A Major prdfeésor ' Date AUgUSt 18; 1970 0-169 V BINDING BY HMS & SONS’ 300K BINDERY INC: l murmur" I . Ill-lull! e. ‘ I... :a ABSTRACT AUDIOLOGICAL MANIFESTATIONS IN JUVENILE - ONSET DIABE TICS By Kenneth R. Johnson Past studies have reported a relation between diabetes mellitus and impaired auditory function. However, these studies presented contradictory results. They reported from 0 to 81% of their samples as having a hearing loss in conjunction with diabetes. Very few of these studies have been concerned with juvenile -onset diabetes, and all but one of the studies have been European. The purpose of this study was to determine the performance of juvenile -onset diabetics on certain auditory tests. The questions posed were whether the age of the diabetic had an effect on test responses; whether the age at onset of the diabetes had an effect on test responses; and whether the duration of the diabetes had an effect on test responses. Thirty individuals between the ages of 12 and 45 years, who had had a medical diagnosis of diabetes mellitus prior to the age of Kenneth R. Johnson 25 years and who were on a daily therapy program of insulin, served as subjects. Their mean age was 24 years. After a thorough case history was takenand other possible causes of hearing loss were identified, these subjects were given a clinical battery of auditory. tests. All testing was performed in a two —room test suite. The test battery included conventional air - and bone - conduction pure —tone testing, tone decay and 8181 at three frequen- cies, speech reception threshold (SRT), speechdiscrimination at + 15 and + 40 dB sensation levels, sweep -frequency Bekesy tracings and brief -tone audiometry. Mean responses to all the tests were within normal limits. The poorest mean pure -tone threshold was 10. 3 dB hearing level, and there‘were no air -bone gaps greater than 5 dB. The mean speech reception threshold was 4. 3 dB hearing level, whereas the mean speech discrimination scores were 75. 4% at the + 15 dB sensation level and 98. 3% at the + 40 dB sensation level. The mean responses to the special pure -tone tests could be classified as negative. The mean tone decay scores were 5. 2, 4. 7 and 5. 3 dB respectively at 500, 1000 and 4000 Hz, whereas the mean SISI scores were 17. 0, 17.3 and 15. 3% at 1000, 2000 and 4000 Hz respectively. All thirty Bekesy audiograms were classified as Type I, and the Bekesy thresholds closely. approximated the pure -tone thresholds obtained Kenneth R. Johnson via conventional audiometry. In brief -tone audiometry the difference between the thresholds at the equivalent durations of 20 and 200 msec at 2000 Hz was 7.2 dB, whereas the difference at 4000 Hz was 5. 9 dB. The speech discrimination scores at the + 15 dB sensation level revealed a significant effect due to the age at the onset of diabetes. A significant difference was also found between the brief- tone audiometry scores at 4000 Hz for this population and a popula- tion of normal hearers from another study. The remainder of the data revealed no hearing loss and there were no other significant effects due to: (a) the duration of the diabetes, (b) the age at the onset of the diabetes, or (c) the current age of the diabetic. Further conclusions were made and recommendations for further research were discussed. AUDIOLOGICAL MANIFESTATIONS IN JUVENILE -ONSET DIABETICS By Kenneth R Jdohns on A THE SIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Audiology and Speech Sciences 1970 TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES . CHAPTER I. INTRODUCTION Purpose of the Study . . Importance of the Study . Definitions II. REVIEW OF BACKGROUND LITERATURE Diabetes Mellitus Diabetic Complications Juvenile -Onset Diabetes Previous Audiometric Studies with Diabetics III. EXPERIMENTAL PROCEDURE . , Subjects , , Auditory Tests Employed Equipment Instrumentation and Calibration Procedures IV. RESULTS AND DISCUSSION Test Results . Statistical Results Discussion ’11 Page vii «101$ 12 12 19 29 32 45 45 47 48 50 55 62 64 72 80 CHAPTER V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Summary Conclusions Recommendations for Further Research BIBLIOGRAPHY APPENDIX A. SUBJECT INFORMATION SHEET . B. SUMMARY OF RAW DATA . iii Page 83 83 84 86 90 95 97 LIST OF TAB LES 'Table Page 1. Incidence ofdiabetes by age. . . . . . . . . . . . 14 2. Previous investigations that have studied the hearingofdiabetics ............. . 42 3. Mean pure -tone air -conduction thresholds in dB HL and standard deviations for‘thirty subjects.....................65 4.. The means, ranges and standard deviations of the speech discrimination scores . . . . . . . . . 66 5. Mean tone decay scores and standard deviations . . . 67 6. Mean SISI scores, standard deviations and highest scoreat each frequency . . . . . . . . . . 67 7. Mean threshold differences, and their standard deviations, between the equivalent durations of 20and200msec.‘....;........... 69 8. Division of the thirty subjects into the appropriate groups for the one -way analyses -of -variance tests.......................74 9. Group means according to "age at onset" for the 3181 test at 4000 Hz, SRT, speech discrimination at a +15 dB SL and the difference between the 20 and 200 msec duration thresholds at 4000 Hz . . . 75 10. F ratios obtained from a one sway analysis -of— variance test between the variable "age at onset" and selected auditory variables . . . . . . . . . . 75 iv Table 11. 12. 13. 14. Page Group means according to "duration of diabetes" for the 8181 test at 4000 Hz, SRT, speech discrimination at a +15 dB SL and the difference between the 20 and 200 msec duration thresholds at4000Hz....................77 F ratios obtained from a one -way analysis -of - variance test between the variable "duration of diabetes" and selected auditory variables . . . . . . 78 Group means according to "current age" for the 8131 test at 4000 Hz, SRT, speech discrimination at a +15 dB SL, and the difference between the 20 and 200 msec duration thresholds at 4000 Hz . . . 79 F ratios obtained from a one -way analysis —of- variance test between the variable "current age" and selected auditory variables . . . . . . . . . . 80 Figure LIST OF FIGURES Block diagram of the brief -tone audiometry apparatus A diagram of the brief -tone signal Audiogram showing mean air -conduction thresholds for thirty subjects and the highest thresholds obtained at each frequency vi Page 53 55 63 CHAPTER I INTRODUCTION Diabetes mellitus is an ancient, molesting disease responsible for many complications, some of which can lead to blindness, kidney failure, lower-limb amputation and even death. Julius Hoffman, speaking about diabetes, said, "Even the most casual superficial glance at the field of diabetes will quickly reveal an outstanding and bewildering vastness and intricacy so that to attempt a global attack on the problems would quickly lead to futility and frustration. "1 Numerous books and articles have beenwritten about the history, etiology, cures and treatment, complications, etc. , of diabetes mellitus; but it has seldom been mentioned as being related in any way to changes in hearing. Even in otologic and audiologic textbooks there has been little mention of diabetes mellitus leading to a reductionin hearing. There is, however, sufficient evidence to make it suspicious. 1Julius Hoffman, "Peripheral Neuropathy in Children with Diabetes Mellims, " Acta Neurologica Scandinavica, Supplementum 8, 40 (1964), 7. According to Jorgensen and Buch, 1 the discussion of the relationship between diabetes mellitus and hearing loss has been going on for over 100 years. In those earlier years most of the interest was centered on diseases of the outer and middle ear because the diabetic was susceptible to infection and generally did not survive long enough to suffer the later complications that are seen today. However, after the discovery of insulin in 1922 and the advent of the sulfonamides and antibiotics, many of these cases werecured. Interest was then centered on the-inner ear and acoustic nerve, and some studies on hearing loss in diabetes followed, frequently with conflicting results. According to Axelsson and Fagerberg, 2 published studies have reported the incidence of hearing loss in diabetesfrom 0 to 81%. While the majority of these studies agreed that a hearing loss can and does result directly or indirectly from diabetes mellitus, an inherent deficiencyin these studies has been the lack of thorough audiometric evaluation. Spoken voice, whispered voice, tuning fork tests, etc. were routinely employed; but these tests are unable to yield the definitive information offered by pure -tone and speech 1M. Balslev J orgensen and Nils H. Buch, "Studies on Inner- Ear Function and Cranial Nerves in Diabetes, 7' Acta Oto -Laryngologica, 53 (1961), 350. 2 A. Axelsson and S. -E. Fagerberg, "Auditory Function in Diabetics, " Acta Oto-Laryngolojgica, 66 (1968), 50. audiometry. With these latter two tests, not only can the amount of hearing loss be determined at various frequencies, but also the site of the lesion can frequently be pinpointed with a high degree of accu- racy. Marshak and Anderson stated it as follows: "The response of juvenile -onset diabetics to specific audiometric tests may well pro- vide information as to the behavioral evidence of the probable presence and site of lesions within the auditory system. "1 A complete assessment of the audiological manifestations has been noticeably lacking in the previous studies. In fact, only one recent study has been found where speech audiometry was employed 9 as a diagnostic tool, 2 whereas only two studies, reported utilizing the Bekesy audiometer. These studies support the fact that a hearing loss does indeed exist in conjunction with diabetes mellitus, though not in 100% of the cases. These findings are in direct opposi- tion to a statement made by Sir Derrick Dunlop, in an opening address at the symposium of the Pfizer Foundation of the Post-Graduate 1Gabriel Marshak and Charles V. Anderson, "Bekesy Audiometry with Juvenile -Onset Diabetics, " Journal of Auditory Research, 8 (1968), 323 -324. 2Axelsson and Fagerberg, "Auditory Function. " 3Marshak and Anderson, "Bekesy Audiometry. " 4J. Zelenka and P. Kosak, "Disorder‘in Blood Supply of the Inner Ear as Early Symptom of Diabetic Angiopathy, " Journal of Laryngology and Otology, 79 (April 1965). Medical School, University of Edinburgh, in 1965. He said, ". . . diabetes is taking pride of place as being‘the disorder in which any doctor can find something to interest him whatever his specialty--apart perhaps from oto —rhinolaryngology. "1 The otologist should most certainly be interested in hearing losses brought on by diabetes, not only for diagnosis and treatment but also for research in determining the site of lesion and methods of prevention. Purpose of the Study This study was undertaken to examine audiometrically the auditory system in juvenile -onset diabetics up to the cochlear nucleus. More specifically this study proposed to determine whether or not individuals, who have been medically diagnosed as having diabetes prior to the age of 25 years, have any associated changes in their hearing that are not generally accepted as being normal. Within the limits imposed by this study, answers to the following questions were sought: 1. Do juvenile-onset diabetics demonstrate a hearing loss not accounted for by other causes? 1Sir Derrick Dunlop, "Opening Address, " Diabetes Mellitus, ed. by L. J. P. Duncan (Edinburgh: Edinburgh University Press, 1966), p. 2. 2. Does the age at the onset of diabetes mellitus have any effect on hearing? 3. Does the duration since the medical diagnosis of diabetes mellitus have an effect on hearing? 4. Does the age of the diabetic have an effect on hearing? Importance of the Study In reviewing the literature it became evident that there was a surprising lack of scientific research relative to the effect juvenile -onset diabetes has on hearing but a wealth of information on the complications of this disease, especially diabetic angiopathy and diabetic neuropathy. Although there have been studies dating back as far as 1899 relating hearing loss and diabetes, only one study, I done in Sweden, has attempted to study diabetics thoroughly by applying a series of modern audiological tests. That study and its results will be fully discussed in the following chapter along with other relevant studies. With few exceptions these studies have dealt with adult diabetics and have neglected those peoplewho acquire their diabetes as children. Juvenile -onset diabetes is generally regarded as being a more severe form of the disease and can, therefore, be looked at as a cause of more and serious complications. Danowski said: leelsson and Fagerberg, "Auditory Function." "Juvenile diabetics . . . compared to adults, present problems out of proportion to their small numbers. "1 If it is confirmed by this study that one of those problems is a hearing loss, then this study will be purposeful for at least four reasons. First, when a child develops diabetes mellitus, he can be forewarned of the unusual complications that accompany the disease plus the additional one of a hearing loss. Being forewarned, both the child, parents, audiologist, and physician could be better pre - pared to deal with the hearing loss. Second, armed with the knowl- edge that a hearing loss is imminent, a program of acceptance could begin along with an educational program of how to cope with the impending change in hearing. Third, it is possible that a medical research program could be directed toward controlling the extent of the hearing loss andthe concomitant complications. Finally, the otologist and audiologist would both benefit by having one more hear - ing loss removed from the wastebasket terminology of "sensori-neural unknown. " On the other hand, it may turn out that juvenile -onset diabetics acquire no abnormal changes in their hearing before they reach the age of 45 years. This knowledge would be most assuring 1T. S. Danowski, Diabetes Mellitus with Emphasis on Children and Young Adults (Baltimore: The Williams and Wilkins Co. , 1957), p. 135. and comforting to this particular population of diabetics because if there is one more problem they don't need, it is an auditory prob - lem. Regardless of the outcome of this study, it must be con- stantly reinforced that the findings must not be generalized to other diabetic populations. This study is dealing with only a specific group of diabetics, and to generalize to other diabetic populations would be both hazardous and foolhardy. If it is determined that the juvenile -onset diabetics. in this study have no abnormal changes in their hearing, then further studies must be undertaken to determine why there are reports in the litera- ture to the contrary. Definitions Hearing loss. --Any audiological manifestations that were deviant from standardized norms. On the pure -tone audiogram this was any loss greater than 25 dB hearing level (HL) obtained on a pure -tone audiometer calibrated to the 1964 norm established by the International Standards Organization at the octave frequencies from 125 to 8000 Hz, including 3000 Hz. In speech audiometry a hearing loss was a speech reception threshold poorer than 25 dB HL. In Bekesy audiometry thresholds below 25 dB were considered a hearing loss as were tracings that could not be classified as Type I. 1 For brief-tone audiometry a hearing loss was said to exist if there was less than a 5 dB difference between the obtained thresholds at the equivalent durations of 20 and 200 msec. Diabetes mellitus. - - A metabolic disorder in which the ability to oxidize carbohy- drates is more orlless completely lost, usually due to faulty pancreatic activity, especially of the islets of Langerhans, and consequent disturbance of the normal insulin mechanism. This produces hyperglycemia with resulting glycosuria and polyuria giving symptoms of thirst, hunger, emaciation and weakness and also imperfect combustion of fats with resulting acidosis, sometimes leading to dyspnea, lipemia, ketonuria, and finally coma. There may also be pruritus and lowered resistance to pyogenic infections. This disease is often referred to as sugar diabetes, and one suffer- ing from this disease is said to be diabetic. This study was concerned with individuals who have been medically diagnosed as having diabetes mellitus prior to the age of 25 years, and they were called juvenile- onset diabetics. 1James Jerger, "Bekesy Audiometry in Analysis of Auditory Disorders, " Journal of Speech and Hearing Research, 3 (1960), 278. 2Dorland' s Illustrated Medical Dictionary (24th ed. : 1955), pp. 410-411. Hyperglycemia. -—Abnormally increased content of sugar in the blood. 1 Glycosuria. -- The presence of an abnormal amount of glu- cose in the urine; especially the excretion of an abnormally large 2 amount of sugar (glucose) in the urine; . . . Polyuria. -— The passage of a large volume of urine in a given period. Acidosis. --A pathologic condition resulting from accumu- lation of acid or loss of base in the body, and characterized by increase in hydrogen ion concentration (decrease in pH). Dyspnea. --Difficult or labored breathing. 5 Lipid. --Any one of a group of organic substances which are insoluble in-water, but soluble in alcohol, ether, chloroform and other fat solvents and which have a greasy feel. . . . 6 l Ibid., p. 703. 2 . Ib1d., p. 623. 3Ibid., p. 1199. 4Ibid., p. 26. 5 . Ib1d., p. 459. 61bid., p. 841. 10 Lipemia. -- The presence of an excess of lipids in the blood. 1 Ketone. --Any compound containing the carbonyl group, CO. 2 Ketonuria. -- The presence of ketone (acetone) bodies in the urine. Acetone bodies. --. . . being intermediates in fat metabo- lism. Called also ketone bodies. 4 . . 5' Prur1tus. --Itch1ng. Pyogenic. - - Producing pus. 6 Insulin. --A protein hormone formed by the islet cells of Langerhans in the pancreas and secreted into the blood, where it regulates carbohydrate (sugar) metabolism. Also a preparation of lIbid. 2Ibid., p. 778. 3Ibid . 4Ibid. , p. 208. 51pm, p. 1236. 61bid., p. 1256. 11 the active principle of the pancreas, used therapeutically in diabetes . . . . 1 and somet1mes in other conditions. The above definitions are considered basic to a study involv— ing diabetes mellitus, and any other definitions that may be needed will be provided when necessary. 1Ibid., p. 746. CHAPTER II REVIEW OF BACKGROUND LITERATURE The author has two intentions in this chapter: first, to give the reader a basic understanding of the disease diabetes mellitus and its many complications and, second, to review the work of other authors who have been interested in the hearing of diabetics. After‘the explanation of diabetes mellitus there will be a review of the two main complications, angiopathy and neuropathy, that could possibly lead to a reduction in hearing. There will then follow a review of most of the studies that have taken a look at the hearing of individuals with diagnosed cases of diabetes mellitus. Diabetes Mellitus Diabetes mellitus is no new disease. The first evidence of it was recorded in 1500 B. C. , 1 but it wasn' t until the second century A.D. that a Greek physician named Aretaeus of Cappodokia gave a 1N. s. Papaspyros, The History of Diabetes Mellitus (Stuttgart: Georg Thieme Verlag, 1964), p. 8. 12 13 complete clinical description of diabetes. 1 He was also responsible for giving the disease. its present name. From the existing polyuria, as if the human had a "siphon" and ”run through" of body fluids, he applied the Greek base of diabetes. To this he added the Latin adjective mellitus which meant honeyed. Hoffmanz claims that "from the etymological point of view, this has remained unchanged. " Diabetes mellitus is a prevalent disease in the United States. In the late 1940' 8 there were approximately two million Americans with diabetes. By 1964 this figure had exceeded three million, 3 and another two million have it and do not yet know it. 4 Each year about . 150, 000 persons are diagnosed as being diabetic. 5 These figures indicate that about 1 out of every 60 Americans is diabetic. 6 1Ibid., p.. 1. 2Julius Hoffman, "Peripheral Neuropathy. in Children with Diabetes Mellitus, " Acta Neurologica Scandinavica, Supplementum 8, 40 (1964), 8. J 3U. S. Department of Health, Education, and Welfare, Diabetes Source Book, Public Health Service Publication No. 1168, May 1964, p. 1. 4George F. Schmitt, Diabetes for Diabetics (2d ed. ; Miami: Diabetes Press of America, 1968), p. 5. 5Garfield G. Duncan, ed. , Diseases of Metabolism (5th ed. ; Philadelphia and London: W. B. Saunders Co., 1964), p. 3. 6Schmitt, Diabetes for Diabetics, p. 5. 14 However, diagnosed diabetes is ten times more prevalent in persons aged 45 and over. 1 To further explain the relation between age and incidence of diabetes, Table 1 follows: Table 1. --Incidence of diabetes by age. Age Incidence 0-20 1 in 2500 20-40 1 in 1000 40-50 1 in 200 50-60 1 in 100 60-70 1 in 50 Tabled data from Schmitt, Diabetes for Diabetics, p. 5. In childhood diabetes the sex distribution is nearly even. The J oslin Clinic in Bostonhad seen 4054 cases of childhood diabetes; and 48. 6% of them were male, with the age of 11 years being the most common age of onset for both sexes. 2 However, after the age of 25 years there is a marked predilection for the female. 3 1Duncan, Diseases of Metabolism, p. 3. 2Elliot J oslin et a1. , Treatment of Diabetes Mellitus (10th ed. rev.; Philadelphia: Lea & Febiger, 1956), p. 656. 3Schmitt, Diabetes for Diabetics, p. 5. 15 In 1922 it was found that animal insulin was a safe and life -saving substitute for human insulin. Since then, the life expectancy of diabetics has greatly increased. In the pre - insulin era, about half of the adults, and nearly all children, died from diabetic coma, while many other diabetics died from other diabetic complications. Today the life expectancy of diabetics has been raised to two -thirds that of the general population. However, today diabetes ranks eighth as a cause of death in the United States while in 1900 it ranked twenty -seventh. This does not necessarily mean that diabetes is getting worse, but in part it is due to the longer life span of diabetics and also to the elimination of many of the other leading causes of death. Diabetes mellitus is very much with us today and, as will be seen, is the cause of many complications. Until a cure or effective preventive measure is found, diabetics and all their problems will continue to be seen. Prior to a discussion of diabetic complications, it would seem appropriate to review very basically the process of sugar metabolism in the human body. Fuel for the body comes from glucose, which is a form of sugar, also called dextrose. This glucose is manufactured in the liver from protein and fat and is also derived from the digestion of sugars and starches in the foods we eat. A certain amount of glucose is always present in the blood. To make 16 certain there is always a continuous supply of glucose, some of it is stored in the liver in the form of glycogen to be released when needed. The signal for the manufacture and release of glucose by the liver is insulin. This hormone is also responsible for glucose utilization by every tissue of the body. Julius Hoffman emphasized the importance of glucose thusly: Because glucose is an essential ingredient in the basic metabo- lism of all cells and presumably all elements within the cell, and therefore, very life itself, any abnormality in its produc- tion or utilization would bring about a malfunction of the organ and/or organism dependent on it. In persons with diabetes mellitus the signal for theliver to release and manufacture glucose and for the tissues of the body to utilize the glucose does not get through. This may be due to the fact that insulin is no longer being produced by the Islands of Langerhans in the pancreas, or it may be because its travel route is impeded elsewhere. Whatever the reason, the signal is not received. All currently held theories of the pathogenesis of diabetes begin with the premise that diabetes is characterized by an ineffective insulin mechanism for regulating glucose utilization. Inter- ference with the insulin mechanism could arise at any number of steps along the path from the stimulus for‘insulin secretion to final tissue responsiveness to the hormone. 1Hoffman, "Peripheral Neuropathy in Children, " p. 17 . 2 Charles J. Goodner, "Newer Concepts in Diabetes Mellitus, Including Management, " Disease-a-Month, September, 1965, p. 21. 17 So the diabetic, having an insulin deficiency, is unable to use sugar. When an untreated diabetic eats these foods, they accumu- late in his body because he has no adequate means of breaking them down. The sugar “accumulates in his blood, overflows into the kid- neys, and appears in the urine. Because the body cannot use sugars, it burns up quantities of its own protein and fat. As a result, the person begins to loseweight and certain poisonous acids accumulate in the body. These are called ketone and acetone bodies. Temporarily, the kidneys are able to get rid of the acids by using their own alkali reserves. When the alkali stockpile is exhausted, the lungs make an attempt to neutralize the acidity by "blowing off" the acids by an increased rate and depth of breathing. Finally, according to Schmitt, 1 when the kidneys and lungs are nolonger able to cope with the situation, the acids continue to build up; and finally the patient goes into ketosis, acidosis, coma and dies. He further stated that the major causes that bring on a coma are absence of or insufficient insulin, resistance to insulin, too much food, infections, vomiting and diarrhea, injuries and operations, shock, pregnancy, anesthesia, and emotional stress. 1Schmitt, Diabetes for Diabetics, p. 181. 2Ibid. 18 This elementary explanation of the metabolic disease diabetes mellitus already reveals that it is an extremely complex disease, and its usual complications have not yet been discussed. Before they are discussed, a few more general comments about diabetes seem appro- priate. The first comment is that the intelligence of the diabetic is unimpaired. This is indeed fortunate for many reasons, one of which is the prevalence of the disease. In fact, as mentioned earlier, diabetes is on the increase because diabetics are living longer. In the pre -insulin years young diabetics died very quickly, and success- ful diabetic pregnancies were so rare as to be clinical curiosities. Now young diabetics are living to breed and, according to J oslin, 1 the bulk of the evidence favors inheritance of the trait as a non -sex linked recessive. Schmitt2 stated that over 80% of diabetics have a positive family history. Other potential diabetics are the obese and those with advancing age. Clarke succinctly said, "It maywell be that gluttony not genes is the. problem» in the older age groups. "3 lDanowski, Diabetes Mellitus with Emphasis on Children, p. 131. 2 Schmitt, Diabetes forDiabetics, p. 4. 3C. -A. Clarke, "Genetic Aspects of Diabetes, " in Diabetes Mellitus, ed. by L. J. P. Duncan (Edinburgh: Edinburgh University Press, 1960), p.. 113. 19 Diabetic Complications Today diabetes can be controlled in different ways. To control diabetes usually refers to maintaining the amount of sugar found in the blood within normal limits and keeping little or no sugar in the urine. The purpose of controlling the blood -sugar level is to aid in minimizing the later diabetic complications. The method of control ~of diabetes depends on the severity of the disease. Those persons with a mild case of diabetes are able to control it simply by following a special diet. Others are able to combine diet and oral medication. This oral medication has the chemical name of tolbuta- mide and was released for use in the United States in 1957. This drug has no blood -sugar lowering effect in the absence of the pan- creatic beta cells (Islets of Langerhans) and is not used in the juvenile type of diabetes. It acts by stimulating an increased release of insulin from the pancreas. Finally, those individuals with the more severe cases of diabetes must use the combination of diet and insulin. Insulin has to be injected because it is a protein substance and would be broken up by digestive juices if taken orally. Control of diabetes is extremely important, not only‘to pre - vent diabetic coma but also to improve the chances of not developing other complications. In speaking about one complication, retinopathy, Jackson and Pickens reported that, 20 Observations have indicated that duration of the disease is not an important factor in the development of retinopathy. The degree of control of the diabetes was the only identifiable factor bearing a constantly significant relationship to the incidence and severity of this complication. No diabetic patient can be promised that he will not have complications such as retinopathy, peripheral neuritis, nephritis, cardiac failures, thrombosis, and gangrene. 2 He also has an affinity for certain other complications such as acute infections of the skin, the upper respiratory system, and the genito -urinary tract. However, besides these, Sheppard3 said the diabetic patient is also heir to all the complications of the nondiabetic. The majority of the chronic diabetic complications are being encountered with ever-increasing frequency today as compared with fifty years ago because of the much longer expected life span of the diabetic. This presents a real problem because once established, these chronic complications tend to progress. It is they, and not the diabetes per se, that incapacitate and threaten life. It should be noted at this point that no mention has been made of hearing loss being a complication of diabetes mellitus. There 1Duncan, Diseases of Metabolism, pp. 1104-1105. 2L. Benjamin Sheppard, Current Concepts of Diabetes Mellitus with Special Reference to Ocular Changes (Springfield: Charles C. Thomas, 1955). p. 72. 3Ibid. , p. 73. 21 seem to be two camps regarding this matter. One camp, referred to in Chapter I in a statement made by Sir Derrick Dunlop, feels that audition is not affected by the disease. This camp is also represented by Abraham R. Hollender, who authored Office Practice of Otolaryn- gology. In this book he stated, "Actually, unlike the relationship of diabetes to the eye, there are no characteristic ear, nose, and throat involvements which can be contributed directly to the disease. "1 The other camp, those who feel there can be an associated auditory problem with diabetes, is. somewhat larger; and one enthu - siastic member of this camp, Jorgensen, stated, "Diabetes mellitus as the cause of partial deafness is often mentioned in the textbooks of otology as well as internal medicine. "2 In order to understand how diabetes may or may not affect hearing, it is necessary to examine two specific diabetic complica- tions. These are diabetic angiopathy and diabetic neuropathy. Diabetic giopathy. --In a discussion on the angiopathy of diabetes, Fagerberg offered a general definition of the term by saying, 1Abraham R. Hollender, Office Practice of Otolaryngology (Philadelphia: F. A. Davis Co., 1965), p. 142. 2M. Balslev J orgensen, "Sudden Loss of Inner Ear Function in the Course of Long -StandingDiabetes Mellitus, " Acta Oto- Laryngolggica, 51 (1960), 579. 22 ". . . in diabetes there occurs a generalized vascular disease which is characteristic of, but not necessarily specific to, the disorder, and this is termed diabetic angiopathy. "1 It is generally believed that the manifestations of diabetic angiopathy are most frequently found in persons who have been diabetic for 10 to 15years. However, in many patients, the appearance of diabetic vascular disease has led to the diagnosis of diabetes. In such/cases, it was generally considered that the metabolic disorder had been present for some time but was not severe enough to cause the usual symptoms. 2’ 3’ 4’ 5 These vascular changes are found in many parts of the body: the renal artery and its branches, the walls of retinal vessels, 6 the glomerular and skin arteries, the muscle arteries of the extremities, the very periphery of the arterial bed, the arterioles, capillaries and venules, the terminal arteriole, andthe arteria auditiva interna. 1S. -E. Fagerberg, "The Angiopathy of Diabetes, " in Diabetes Mellitus, ed. by L. J. P. Duncan (Edinburgh: Edinburgh University Press, 1960), p. 54. 2Ibid. 3Rachmiel Levine, "Diabetes Mellitus, " Clinical Symposia, Vol. 15, No. 4 (Oct., Nov., Dec., 1963), 127. 4Fagerberg, ”The Angiopathy of Diabetes, " p. 55. 5Danowski, Diabetes Mellitus with Emphasis on Children, p. 422. 6Zelenka and Kozak, "Disorder in Blood Supply of the Inner Ear, " p. 314. 23 Vascular changes have also been found in the peripheral nerves of diabetics, in the mucous membrane of the stomach, in the duodenum and in the skin. 1 Sclerotic changes are found in the coronary arteries and arteries of the pelvis. 2 J orgensen3 reported these irreversible changes are represented by swelling and hyaline degen- eration of the ground substance-of the intima (the innermost smooth layer of cells in blood vessels which is the only part of the blood vessel in direct contact with the blood4) and the infiltration of‘lipids and subsequent calcification. Still another difficulty arises from atherosclerosis. This, according to Duncan, 5 is a manifestation of lipid metabolism or, as further explained by Dorland' 8 Medical Dictionary, 6 is a lesion of largeiand medium ~sized arteries, with deposits in the intima of yellowish plaques containing cholesterol, lipoid material, and lipophages. The severity of these disorders can be pointed out by noting that lesions of the vascular system are responsible for death 1Fagerberg, "The Angiopathy of Diabetes, " p. 57. 2 Duncan, Diseases of Metabolism, p. 948. 3Ibid., p. 949. 4Schmitt, Diabetes for Diabetics, p. 186. 5Duncan, Diseases of Metabolism, p. 948. 6Dor1and' s Illustrated Medical Dictionary, p. 159. 24 in more than 50% of patients who have had diabetes for more than 15 years. 1 Rachiel Levine stated that "70 percent of deaths in diabetic patients are due to vascular disease. This is about two- and —one -half times the death rate from the same causes among non- diabetic individuals of all ages. ”2 Recall that earlier in this chapter diabetes mellitus was ranked as the eighth leading cause of deaths in the United States. Also, diabetes is the third leading cause of blind- ness. The above discussion shows that there is diffuse vascular disease in diabetics, in some cases before the actual diagnosis of diabetes mellitus but mostly in persons who have been diabetic for at least 10 to 15 years. That capillaries are also affected has been mentioned; and Levine explained that when they are studied, ". . . the basement membrane may be found thickened with material that is stained red by periodic acid Schiff reagent (PAS) and thus said to be PAS positive. "4 Before leaving this topic, it should be noted that some controversy has been evident concerning the role of insulin in 1Duncan, Diseases of Metabolism, p. 949. 2Levine, Diabetes Mellitus, p. 131. 3Diabetes Source Book, p. 35. 4Levine, Diabetes Mellitus, p. 130. 25 angiopathy. In a paper on insulin action on arterial tissue, R. F. Mahler presented the position that ". . . insulin may be the link between diabetes and atheroma, and that the determining factor is not the lack of insulin but rather its presence, at inappro- priate times and perhaps in excessive amounts. "1 Regardless of its cause, diabetic angiopathy is of interest to this investigation because of its possible presence in the temporal bone and the resulting effect it may have onaudition. Diabetic neuropathy. - — This complication of diabetes mellitus is characterized by an affection of the peripheral nerves which involves both sensory and motor fibers, but chiefly the former. A somewhat more inclusive definition calls it an "acute or chronic degenerative condition of the peripheral nerves, autonomic nervous system or central nervous system peculiar to diabetes. "2 Hoffman3 traced the history of diabetic neuropathy back to 1857 when Billiard theorized a relationship between neuropathy and diabetes and even earlier to 1848 when Marchal de Calvi and also 1R. F. Mahler, "Insulin Action on Arterial Tissue in Rela- tion to Diabetes andAtheroma, " in Diabetes Mellitus, ed. by L. J. P. Duncan (Edinburgh: Edinburgh University Press, 1960), p. 44. ' 2 Joslin et a1. , Treatment of Diabetes, p. 485. 3 Hoffman, "Peripheral Neuropathy in Children, " p. 9. 26 Bumenil talked of diabetic peripheral neuropathy and multiple neuritis in diabetes. The degenerative changes of the diabetic neuropathies—-the characteristically chronic neurologic complications --occur in the peripheral nerves anywhere in the entire nervous system. Thomp- son1 said that any of the cranial nerves can be affected; yet, with the exception of the third, fourth, sixth, and seventh nerves, descriptions of paralyses of cranial nerves are few. As will be seen later in this chapter, Axelsson and Fagerberg feel that the eighth nerve may be affected, as evidenced by the results obtained when they tested some diabetics. Just as it was suggested earlier that long durations of uncontrolled diabetes may be a contributing factor in diabetic angio- pathy, so it is for diabetic neuropathy. In the Treatment of Diabetes Mellitus it is stated that "the longer‘the duration of imper- fectly controlled diabetes, the more frequent and diverse are the 2 . complications involving the nervous system. " It was also pomted out in this book that diabetic neuritis, in almost all instances, follows a period of uncontrolled diabetes. 3 lWillard 0. Thompson, ed., The Diabetic Neuropathies (Springfield: Charles C. Thomas, 1953). p. 42. 2Joslin et al. , Treatment of Diabetes, p. 485. 3Ibid., p. 497. 27 It is possible though that a relationship does exist between the angiopathies and neuropathies because it was suggested by Hoffman1 that arteriosclerosis is a cause of neuropathy and by J oslin et a1. 2 that intraneural vessels are thickened in peripheral nerves that show pathological changes. This is not completely sub - stantiated though, and Hoffman made it clear that the "pathogenesis of diabetic peripheral neuropathy is quite unknown. "3 This complication, diabetic neuropathy, can occur at any time from childhood through very old age. However, it occurs mainly in late diabetics of advanced age. It should not be forgotten, though, that many younger diabetics are affected by this disorder. 4 It is now established that diabetes..does have a chronic complication called diabetic neuropathy, but it has not been actually stated what the pathology really is. That is, what actually happens to the nerve and how could this affect audition? Basically, there are two types of nerve changes relevant to this study. One involves 1Hoffman, "Peripheral Neuropathy in Children," pp. 28 —29. 2Joslin et a1. , Treatment of Diabetes, p. 497. 3Hoffman, "Peripheral Neuropathy in Children, " p. 30. 4Ibid., p. 20. 28 demyelination of the nerve; the other is concerned with a cycle of degeneration and regeneration of the nerves. According to Gilliat, 1 there are two distinct and different pathological processes which can occur in peripheral nerves. One is called Wallerian degeneration, and the other is called segmental demyelination. In the first, Wallerian degeneration, both the axon and myelin sheath are involved. It involves a cycle whereby degen- eration first occurs either by mechanical injury or by certain types of toxic neuropathy; and then, at a rate of a few millimeters per day, regeneration occurs by the sprouting of divided axons at the level of injury. This regeneration involves both the axon and the myelin sheath. Segmental demyelination is a quite different pathological process in which there is a patchy loss of myelin from short segments of the nerve fiber without interruption of the con- tinuity of the axis cylinder itself. . . . however, the axon itself survives so that on either side of the demyelinated seg- ment of segments the nerve is normal. Both of these neural pathologies could have the effect of slowing down the conduction velocity of the nerves. Gilliat explained that in the Wallerian degeneration 1R. W. Gilliat, "Diabetic Neuropathy, " in Diabetes Mellitus, ed. by L. J. P. Duncan (Edinburgh: EdinBurgH University Press, 1960), p. 64. ‘ . 2Ibid . 29 provided repair keeps pace with the myelin breakdown, this process would not affect muscle power or sensation, but the altered myelin thickness and internodal spacing would be expected to slow conduction velocity. It would also produce loss of those functions which require rapid synchronous con- duction in different fibers of the nerve trunk: namely, the tendon reflexes and vibration sensation. 1 In the case of segmental demyelination, "the widespread slowing of nerve conduction velocity is easily explained, as this is a character- istic accompaniment of segmental demyelination in the experimental animal. ”2 If indeed reduced conduction velocity is present in some diabetic patients and if it is present in the eighth cranial nerve, it would be most interesting to determine its audiological manifesta- tions with a typical clinical battery of tests. This test battery could include both vestibular and auditory tests. Juvenile —Onset Diabetes Thus far, this chapter has dealt mainly with the broad topic of diabetes with little or no regard to age. It has been mentioned that maturity -onset diabetes far exceeds juvenile -onset diabetes by about a tenfold difference, depending upon at which age you divide the two. Because this study dealt onlywith diabetics whose medical 1Ihid. 2 Ibid . 30 diagnosis came prior to the age of 25 years, it is necessary to discuss further their disease. As stated by Danowski, "Juvenile diabetics make up some five percent of the total diabetic population of this country, but, compared to adults, present problems out of proportion to their small numbers. "1 The disease is usually detected within a matter of weeks of onset, whereas it is possible for adults to go months or years without clinically disabling symptoms. After the onset, children also present other problems that make them quite different from adults. The disease is usually more severe in children and complicated by the fact that children' s nutritional requirements are relatively greater and constantly changing. Also, infections are often more frequent and severe. Their exercise is more erratic and their emotional pattern is less predictable. In children' 5 diabetes the term "insulin deficient" is applied since there is usually little or no insulin in the pancreas. Because of this, there is little place for the oral anti -diabetic drugs and they have to take insulin. In childhood diabetes, "vascular damage is by far the most important complication. "2 In a long term study of childhood diabetes 1Danowski, Diabetes Mellitus with Emphasis on Children, p. 135. 2 Joslin et al. , Treatment of Diabetes, p. 681. 31 . . . 1 . and vascular compl1cations J oslin et al. observed no vascular les1ons in the first five years of the disease and rarely found them in the first ten years; but by fifteen years such things as retinopathy, calcified arteries, and proteinuria were found. They reported that calcifica- tion of arteries is the second most common vascular lesion in juve- nile diabetes, occurring in 25 to 30% of their sample between the fifteenth and nineteenth years of the disease. They did not observe it in children under the age of ten or prior to the fifth year of diabetes. Farquhar, 2 in discussing the child of the diabetic mother, reported the inner ear has been found disrupted by hemorrhage in a special study involving a small number of early fetuses. He also reported an encounter with one child of a diabetic mother who was found to have unilateral nerve deafness some years after going to school. . However, he offered no support that the hearing loss was due to the mother' 5 diabetes or how, severe the hearing loss was. Before ending this section on juvenile complications, a word needs to be said about neuropathy in the juvenile diabetic. It can be summed up by a statement made by Danowski and reported in an article by Hoffman: 1Ihid. ZJames W. Farquhar, "Child of the Diabetic Mother, " in Diabetes Mellitus, ed. by L. J. P. Duncan (Edinburgh: Edinburgh University Press, 1960), p. 132. 32 Neuropathy . . . fortunately spares juvenile diabetics, at least during the early years of the childhood phase of the disorder. . . . Moreover, in at least one study, it has been reported that vibratory sense which is a functional elaboration of touch, position, and pressure. is intact in juvenile diabetics under the age of 10.1 In summary, it can be stated that juvenile diabetics, after having diabetes for a period of 10 to 15 years, possibly begin developing the diabetic complications of angiopathy and neuropathy. One of the purposes of this investigation was to determine whether these complications affect the sense of audition. A few studies already completed have looked at the effect of diabetes mellitus on hearing. These studies are reviewed in the next section of this chapte r. Previous Audiometric Studies with Diabetics Axelsson and Fagerbergz reported a study involving 99 controlled diabetics. divided equally into three age groups with 33 subjects in each group. Group A contained subjects between the ages of 16 and 30 years, Group B between the ages of 30 and 39, and Group C the ages from 40 to 50 years. There were 59 males and 40 females. Duration of diabetes and severity of complications 1Hoffman, "Peripheral Neuropathy in Children, " p. 21. 2 . Axelsson and Fagerberg, ”Auditory Function, " pp. 49 -64. 33 were well documented. The battery of audiological tests they administered to their subjects was the most complete battery reported in all the studies. The test battery included the following: 1. pure tone audiometry--air conduction 250-4000 Hz. 2. speech discrimination—-Swedish monosyllabic phonetically balanced word lists. 3. distorted speech-—using band —pass filters 640 and 2000 Hz in series at a 40 dB sensation level and testing monaural discrimination. 4. stapedius reflex test--recording the stapedius reflex threshold at 250, 500, 2000 and 4000 Hz in 72 of the 99 subjects. 5. Bekesy audiometry--both continuous and interrupted stimuli on 97 subjects. 6. directional audiometry—-ability to localize the sound source on 92 subjects. In all, 68 subjects (68. 6%) were affected with at least one abnormal result which the authors attributed to diabetes. The most abnormalities were noted in distorted speech audiometry (40 subjects), and the second most abnormalities were found in speech discrimina- tion (34 subjects). On the routine pure -tone audiometric test only six subjects showed abnormal results. .Group B (age range 30 to 39 years) had the largest meanduration of diabetes (16 years) and also the largest percentage of patients with complications (78. 7%). How- ever, this same group had the least number of patients (60. 6%) affected with abnormal hearing on the test battery. Group C showed 34 81. 8% with abnormal hearing and Group A had 63. 6% with abnormal hearing. In studying the results from Bekesy audiometry, it was particularly noted by the authors that a high proportion of the diabetics had greater excursions than a corresponding nondiabetic series. This was interpreted to mean that the diabetics reacted slower than normal to differences in sound intensity. It cannot be settled whether this was due to "slow cerebration, ” vascular diabetic changes in the brain, unfavorable blood sugar-levels during the tests or to any other cause, but it is considered an interesting aspect worth further investigation if the high excur- sions in the Bekesy audiometry could be correlated to a retro- cochlear type hearing impairment. In summary, these authors found that sex, duration, and severity of sugar diabetes had no apparent role in any form of interference with hearing function in their population. Their main conclusion was that the test results point to a retro -cochlear lesion, suggesting that the cochlea is less prone to damage-than brain tissue, "perhaps because at the capillary level brain tissue is more susceptible than the cochlea to anoxia resulting from diabetic angiopathy. "2 11bid., p. 61. 2 Ibid. , p. 62. 35 In 1961 Jorgensen and Buch1 examined 69 diabetic patients in Denmark for inner -ear function and cranial nerve problems. The subjects ranged in age from 16 years to 73 years. The found 39 cases with a hearing loss (this term is not defined in their article) but ruled 11 out because of the possibility that the hearing loss could have been caused by other physiological reasons. This left 28 cases (40. 5%) whose hearing loss they believed was caused by diabetes. They found no distinct correlation between the duration of diabetes and the impairment of hearing, as did Axelssonand Fagerberg, but did, find hearing loss twice as common in patients with severe proliferative retinopathy as in patients without retino- pathy. Further, they reported that in patients under 40 years of age there was an unmistakable correlation between nephropathy and hearing loss. This is in contradiction to the Axelsson and Fagerberg study which reported no correlation between "any type of hearing loss and such diabetic manifestations as angiopathy, retinopathy, nephropathy or neuropathy. "2 They reported no correlation between nephropathywand hearing loss in patients over 40 years of age. In the majority of their hearing loss cases, the authors reported that the loss developed slowly without the patients' noticing, 1Jorgensen and Buch, ”Studies on Inner Ear Function in Diabetics, " pp. 350-364. 2 Axelsson and Fagerberg, "Auditory Function, " p. 60. 36 and in only three cases was there a sudden onset accompanied by transitory tinnitus. Further, the authors reported that only one case complained of impaired vision and hearing on one side during periods of hypoglycemia. Vestibular function of all 69 cases was assessed by investi- gating the patients for spontaneous nystagmus and postural nystagmus as well as by a differential caloric test by the Hallpike method. Abnormal reactions were demonstrated in only two cases, neither of which showed a hearing loss. In their summary the authors stated: The hearing loss appears to develop in two ways. In most cases it comes on slowly and is often so moderate as to escape the patients' attention. As a rule it is bilateral, affecting all frequencies, but with a certain preponderance for high tones. This hearing loss, which is similar to the physiological hearing loss of advancing age, is most common among elderly diabetics, but may be encountered in all age groups. Zelenka and Kozak2 from Czechoslovakia in 1965 reported a study of 17 young diabetics between the ages of 8 and 40 years. They decided to use this age range because, "The atherosclerotic origin of the disorders in the blood supply in general, hence also within the inner ear, is very improbable. "3 1Jorgensen and Buch, "Studies on Inner Ear Function in Diabetics," p. 362. 2 Zelenka and Kozak, "Disorder in Blood Supply of the Inner Ear, " pp. 314-319. 3Ibid. , p. 314. 37 None of their 17 subjects complained of a hearing loss, tinnitus, or vertigo; but audiological examination detected some high frequency loss in "the majority of the investigated individuals"1 with the maximal loss being at 3000 Hz. "This reduction of hearing was roughly proportional to the duration of the diabetes. "2 These authors considered diabetic angiopathy as the cause of the hearing losses they detected. "This disease, specific for diabetes . . . is located, unlike atherosclerosis, in the tiniest arteries at the extreme periphery of the vascular system. "3 In the inner ear they feel the lesion was in the bed of the arteria auditiva interna, which is a relatively long and very fine arteriole. . . . Moreover, it has the character of terminole, so that its obliteration cannot be compensated any more by collateral circulation. Jorgensen, 5 in 1961, reported the histological findings of a study he did on 32 temporal bones from diabetic subjects. The subjects ranged in age from 32 to 84 years; 17 of them.were female. 11bid., p. 316. 2 Ibid. 3Ibid. , p.318. 41bid . 5M. Balseve J orgensen, "Changes of Aging in the Inner Ear, and the Inner Ear in Diabetes Mellitus. Histological Studies, " Acta Oto-Laryngologica. Supplementum 188, pp. 127-128. 38 Most noticeable was that in a large number of the diabetics, the capillary walls of the stria vascularis showed appreciable PAS- positive precipitates. These stood out as "thickcables having walls 10 to 20 times thickerthan normal. "1 These changes were observed in only a few capillaries in mild cases of diabetes, but in the more severe cases they had spread to practically the entire capillary system of the stria. "The changes resembled those seen at times in severe arteriosclerosis, only far more marked. "2 According to J orgensen, the changes in the inner ear appear to be restricted to the stria vascularis. It was reported that these strial changes bore no relation to age, but a distinct relationwas seen to the duration of diabetes. Similarly there was 'an unmistakable correlation to other diabetic complications such as retinopathy, nephro- pathy, and angiopathy of the lower limbs. "These changes are also seen in severe cases of arteriosclerosis but to a much lesser degree. "3 In 1955 Kelamen4 reported the results of a histological study he performed on the embryo of a pregnancy that was terminated in 11bid., p.127. 2Ibid. 3Ibid. 4G. Kelamen, "Aural Changes in Embryo of Diabetic Mother, " American Medical Association Archives of Otolaryngglggy, 62 (1955), 357 -369. 39 the sixth month. The 26 -year -old mother had been diabetic for 15 years, and the pregnancy was interrupted because of a threaten- ing blindness. The ears of the embryo had reached a normal developmental stage, but hemorrhages were found in the mesenchymal cushion of the middle ear. Other hemorrhages were found in modiolus of the cochlea, in the scala media (with the organ of Corti intact), and also between the bundles of the eighth nerve. Recall earlier, Farquhar1 reported a similar finding. In 1956, Ancona2 examined 27 patients aged 16 to 33 years with a history of diabetes. He tested 3 cases between the ages of 16 and 21 years, 21 cases between the ages of 21 and 30 years, and 3 cases between 31 and 33 years. The median duration of their diabetes was five years. All 27 subjects were on insul and were tested twice, one month apart. Air and bone conduction thresholds were obtained, and 70% of the cases demonstrated alternation of hearing. The majority of the hearing losses were described as being high frequency, bilateral, and symmetrical. With the exception of Kelamen' s, the above discussed studies have been done in Europe. The only American audiometric study 1Farquhar, "Child of the Diabetic Mother. " 2F. Ancona, "Considerations of Hearing Disorders in Juvenile Diabetics, " Arcispedale S. Anna Di Ferrara: Rivista Trimestrale Di Scienze Mediche, 9 (1956), 435-443. 40 with diabetics was done by Marshak and Anderson. 1 They tested 30 subjects ranging in age from 11 to 43 years, with a mean age of 20 years, who met the following criteria: (1) a confirmed medical diagnosis of diabetes mellitus priorto the let birthday, (2) a negative history of otological disease, head trauma, noise exposure, familial hearing loss, or the use of ototoxic truge, and (3) a negative ear, nose, and throat examination by a resident otorhinolaryngologist at the time of testing. The mean duration of diabetes was 11 years. All patients were given conventional air and bone conduction threshold tests at the octave frequencies 250 through 8000 Hz for both ears and fixed frequency Bekesy tracings of one to three minutes on the better ear-at 500, 2000, and 4000 Hz for interrupted and continuous stimuli. None of the 30 subjects demonstrated a hearing loss greater than 20 dB (ISO) nor bone conduction thresholds that were more than 5 dB different from the air conduction thresholds. These authors did not mention the width of the Bekesy excursions as Axelsson and Fagerberg did. Positive responses, defined by a separation of 5 to 20 dB between threshold tracings for continuous and interrupted tones on Bekesy audiometry, were given by 57% of the 30 diabetics. The preponderance of such responses and the knowledge of sub- clinical vascular changes accompanying diabetes mellitus lead 1 . Marshak and Anderson, "Bekesy Audiometry with Juvenile - Onset Diabetics, " pp. 323 -330. 21bid.. pp. 324-325. 41 us to the speculation that such responses are not within the normal range but that they may reflect subclinical pathological changes within the auditory system. In their article, Axelsson and Faberberg presented a table2 that compiled most of the previous works that have studied, or at least attempted to study, the hearing of diabetics. The table is pre- sented here (see Table 2), and to it has been added the information of their own study and the information of the Marshak and Anderson study. It is of interest to note that the first study’was reported as long ago as 1899, whereas the latest study reported was in 1968. Some of the studies utilized very large populations while others used relatively small samples. A variety of hearing tests has been used on the sample populations which included an extremely large age range. Almost as large as the age range is the range of percentages of hearing impairment due to diabetes. This range is puzzling because of the five studies that reported no hearing impairment due to diabetes. It can be noted, however, that most of the higher percent- ages of hearing impairment are reported in studies where the age of the tested diabetics was high enough to include the onset of prebycusis. However, the study by Ancona and the Study by Zelenka and Kozak showed hearing losses in 70 and 76% of the cases respectively, 11bid., p. 329. 2 Axelsson and Fagerberg, "Auditory Function, " p. 50. 42 $6 3 «E o 1635 3. sea in: Seesaw .8 .6 mo bwuo>om ens eosaaan mm 3m :9. E: .32 .8 3.6 863m .8 .v mo bfluo>om musmmoun poofim on no ooH ommfi 0:ng ewe. 8 8A 2: om 8-8 on 82 em; o 8v 2 .8 .e we bwuoewom was somewhat .868 8m 3 8m 3- 5 S 82 83885 E. 68533 .8 .e .8 10.83% mm 26.35 S. 33 use expense .8 .e .8 new .3838 .694 E :e 2 22 games H .e 2: 82 saw as ”new “Coachman: 6cm “coauwmag m .8 .U . mummy. momma marge: smegma mfiumos session 3 26 $5532 5 mo .3on gonna—hes cowamaouuoo oz :oflmamuuov unmafimag mm< 9.3552 wfiumom .836an mo marge: 65 cowvfim 93: “m5 mcofimwflmotwfi msogoum: .N 2an 43 - 03086 6803 6888 - .. 63m .3663 £80m 883 - - PE. .6039, 86.86863?» - - >3 .6039, 868386 - - >m 86368 683068636 - .368 .36 8.3680356 680306.30: .6 .83. 536803086 hm6x6m: .x6m 63368 66368637- .86 .898 6 .xwm COWMQUG< 0 .3m $1 3 cm 33 U86 E68862 8368- 08868 .868 .36 .- 0886 .- 0.88868 . .6 .830 .-0836.H . .86 m0 w .5 . .x6m ,om- ow mm 6.8266 . .8 .6 6 .3 .88 8-8 mm 6.886668 m0 803686 .86m 6w< m .3 .>>>>m om-mH mm mama «.86 806686834 183680386 x660v~ 803685 or 63m 34- m S mama 686 6x86H6N 13:68 -08868 .- 0.3868 .- 0836.8 .8868 193680886 .80 6608 8036.85 .A868v N66 Jam-es mm . 63m 3% ~23 ~6636MQ .88 6003 8568 836808368 808m 686 -0.868 8036.85 .6964 :4 63m mun-ma mm 358 866866.83. 668 S 88 2.- S 3 mm 63m om-Hm w 3H6>686m o 63m 66v 3 $3 886 086.308m .8 .6 .30 8036.86 882.66 E. 3.8 3-8 5 88 8668 o ooow 32 8638M 44 and their oldest subject was only 39 years old. In contrast to these are the studies by Profazio and Baravelli and the study by Marshak and Anderson. Their oldest subjects were under 44 years of age and no hearing loss is reported. These results are conflicting and largely unexplainable . Equally perplexing are the contradictions in the last two columns. For example, for every study that lists a correlation between hearing impairment and severity of diabetes mellitus, there is a study that reports no correlation between hearing impairment and severity of diabetes mellitus. Also, four authors report no correlation between hearing impairment and duration of diabetes mellitus, whereas three authors do report a correlation between these two items. It was with these latter thoughts in mind that the current investigation was undertaken. CHAPTER III EXPERIMENTAL PROCEDURE The purpose of this chapter is to describe in detail the subjects, equipment, tests, and procedures utilized in this study. Subjects All subjects were obtained from the Lansing area, largely with the assistance and cooperation of local physicians and the medical staff of Olin Health Center. The majority were obtained by having the local physicians send a letter of explanation to their patients that fit the criteria for this study requesting them to call the hearing clinic to make an appointment for the hearing evaluation. In order to participate in the investigation, a subject could be either male or female but had to have had a medical diagnosis of diabetes mellitus prior to the age of 25 years. Also, the subject had to be no younger than 12 years of age and no older than 45 years at the time of the testing. One further requirement was that each 45 46 subject had to be on a daily dosage of insulin. The type and amount of insulin was not important for the purpose of this study. Before each two-hour session, a thorough case history was taken in order to rule out hearing losses from known pathologies, to learn of any known diabetic complications such as angiopathy, retino- pathy, neuropathy, etc. , and to determine the degree of control the subjects have had over their diabetes. Other questions concerned the familial history of diabetes, previous or present otologic problems, previous hearing tests, vertigo, tinnitus, past noise exposure, and previous health problems. A copy of the questionnaire can be found in Appendix A. Thirty subjects, 8 male and 22 female, were tested. The mean age was 24. 58 years and the median age was 22. 58 years. The duration, from medical diagnosis, of diabetes mellitus ranged from 21 months to 37. 00 years with a mean of 11. 38 years and a median of 10. 22 years. Two subjects reported having diabetic complications. One subject reported having a neuropathy that paralyzed her stomach, and one subject reported having a severe retinopathy and a just noticeable nephropathy. Seventy-three percent of the sub- jects reported a positive family history of diabetes. 47 Auditory Tests Employed Each of the 30 subjects was given the following test battery in the following order: 1. Routine bilateral pure -tone audiometric tests: a. Air conduction-- 12 5, 250, 500, 1000, 2000, 3000, 4000, 8000 Hertz (Hz). b. Bone conduction-~same as above excluding 125 and 8000 Hz. 1 2. Speech audiometry: a. Speech reception threshold in the poorer ear (test earl b. Speech discrimination for CNC monosyllables in the test ear. 3’ 4 1During the bone -conduction testing, the nontest ear'was stimulated with narrow -band noise centered on the test frequency at an effective masking level of 10 dB. 2The test ear for items 2 through 6 above was determined by averaging the air -conduction pure -tone thresholds at 2000, 4000 and 8000 Hz and using the earwith the poorest threshold. 3NorthwesternAuditory Test No. 6, Form A, Lists II and IV, with a male speaker'were employed, always in the above order, and at sensation levels of 15 and 40 dB above the speech reception threshold. 4Tom W. Tillman and Raymond Carhart, ”An Expanded Test for Speech Discrimination Utilizing CNC Monosyllabic Words" (United States Air Force School of Aerospace Medicine), SAM -TR -66 -55, June, 1966. _ 48 3. Carhart' 3 Tone Decay Test at 500, 1000 and 4000 Hz in the test ear. 1 4. Short Increment Sensitivity Index (SISI) in the test ear at 1000, 2000 and 4000 Hz. 5. Sweep -frequency Bekesy audiometry (pulsed and continuous stimuli) in the test ear. 6. Brief -tone audiometry (20, 200 and 400 msec durations) at 2000 and 4000 Hz in the test ear. 2 Equipment The equipment listed below was used during the course of this study and will be or has been referred to. Clinical Audiometer (Beltone, Model 15C) Masking Generator (Beltone, Model NB -102) Short Increment Sensitivity Index (Gordon Stowe & Asso- ciates SISI Adapter, Model 1295) Tape Recorder (Ampex 601) Magnetic Tape (Scotch Brand 201) Speech Audiometer (Grason -Stadler, Model 162) 1Raymond Carhart, "Clinical Determination of Abnormal Auditory Adaptation," Archives of Otolaryngology, 65 (1957), 32 -39. 2H. N. Wright, "Clinical Measurement of Temporal Auditory Summation, " Journal of Speech and Hearing Research, 11 (March, 1968), 109 -127. 49 Bekesy Audiometer (Grason -Stadler, Model E800) Earphones (Telephonics, TDH -39 housed in MX ~41/AR doughnut type cushions) Bone Vibrator (Radioear B-70-A white dot) Electronic Counter-Timer (Bekman, Model 6148) Electronic Switch (Grason -Stadler, Model 829E) Power Supply (Tektronix, Type 160A) Pulse Generators (Tektronix, Type 161) Wave Form Generator (Tektronix, Type 162) Test Suite (Industrial Acoustics Corporation, Series 1600 -ACT, consisting of a single -walled control room and a double -walled test room) Sound Level Meter (Bruel & Kjaer, Type 2203and 2204) Octave Filter Set (Bruel & Kjaer, Type 1613) Artificial Ear (Bruel 8: Kjaer, Type 4152) Condensor Microphone (Bruel & Kjaer, Type 4144) Artificial Mastoid (Beltone, Model M5A) Voltmeter (Bruel & Kjaer, Type 2603) Oscilloscope (Tektronix Storage Oscilloscope, Model 564B) All testing was carried out in a commercially available, prefabricated two -room test suite. The tester'and his equipment were located in a single -walled control room while the subject was seated in an adjacent, double -walled test room. An acoustic mea- surement of the ambient noise in the double -wa11ed test room was 50 made using a precision sound -level meter. On the linear scale the ambient noise level was 58 dB SPL. Using the external filters on the sound -level meter, the noise level was found to be less than 22 dB SPL at the octave frequencies from 125 to 31, 500 Hz. Instrumentation and Calibration Routine pure -tone audiometry. --All the routine pure -tone air and bone conduction thresholds were obtained using the clinical audiometer. All the pure -tone signals were delivered to the ear- phones or bone vibrator after passing through a 20 dB attenuation pad. The frequency calibration of the audiometer was checked before and after the study utilizing an electronic counter. All the test frequencies were within 3% of the frequency selector setting. This is within the specifications established by the American National Standards Institute (ANSI) in S3. 6-1969, the specification for audiometers. 1 Using the storage oscilloscope, it was determined that when the interrupter switch was depressed, the tone rose to its peak level with no overshoot and within the limits specified by the American National Standard Specifications for Audiometers ANSI S3. 6-1969 (New York: American NatiOnal Standards Institute, 1970), p. 8. 51 ANSI S3. 6-1969. 1 The fall time of the tone from its peak-was also found to be within the proper limits. Calibration of the bone vibrator was checked periodically with the Artificial Mastoid and found to be very stable and remained well within normal limits. SISI testing. -- The SISI unit was found to have the required 50 msec risetime, 200 msec peak time, but only a 25 msec decay time by means of the storage oscilloscope. Previous attempts to correct the decay time to the required 50 msec were unsuccessful, so it was employed with that one limitation. It is felt that this had an insignificant effect on the SISI test results. The 1 dB increment was checked periodically throughout the study using the sound level meter. With the meter switch set in the "hold" position and the SISI adapter set at the prOper setting the 1 dB increments could very easily be visualized. Speech audiometry. -- The speech stimulus originated on a magnetic tape and was reproduced by the tape recorder. The signal was then fed into the speech audiometer and finally into the earphone located in the test room. The calibration of the speech audiometer was monitored over the course of this study by feeding both pure llbid. 52 tones and speech noise into the audiometer and reading the acoustic output of the earphone in the test room with the sound level meter and artificial ear assembly. No difficulties arose with the calibration of the speech audiometer. Bekesy audiometry. -- The external calibration of both the intensity and frequency of the Bekesy audiometernwere checked according to the Operator' 8 Manual prior to each Bekesy audiogram. The attenuation rate of this audiometer was timed with a stop watch and found to be the desired 2. 5 dB per second; and the interruption rate, also timed with a stopwatch, was determined to be 2. 5 per second. The chart Speed was determined to be one full octave per minute. Again utilizing the storage oscilloscope, the rise -decay time of the interrupted stimulus was ascertained to be 25 msec. This is in accordance with the manufacturer' s specifications. Brief -tone audiometry. -- From the above test battery the only test administered that could not be performed with a commer -' cially available audiometer or adapter for the audiometer was the brief -tone audiometry. For this part of the test battery a special testing apparatus was employed. An apparatus similar to this was suggested by H. N. Wright1 and consisted of commercially available lWright, "Clinical Measurement of Temporal Auditory Summation, " pp. 109 -127. 53 electronic equipment in conjunction with the Bekesy audiometer. This equipment was originally assembled by Nerbonne1 and was arranged as shown in the simple block diagram shown in Figure 1. _. Response Key TDH-39 Electronic Switch Audiometer \\€ >\\\\\\\\\\\\ On A Off '\ j Eput Timer Scope Pul. Pul. Gen. Gen. Waveform Generator Figure 1. --Block diagram of the brief -tone audiometry apparatus. The test signal was obtained by feeding the output of the oscillator of the Bekesy audiometer into an electronic switch‘which was driven by two pulse generators and a waveform generator. These genera- tors, along with the electronic switch, controlled the duration, repetition rate, and rise -fall time of the tone while frequency and 1M. A. Nerbonne, "A Comparison of Brief Tone Audiometry with Other Selected Auditory Tests of Cochlear Function" (unpub- lished Ph.D. dissertation, Michigan State University, 1970). ' 54 intensity were controlled via the audiometer. The-duration of the pulse going into the electronic switch was constantly monitored with an EPUT timer. The signal was then returned to the audiometer and from there delivered into the earphone. This signal into the earphone-was periodicallyrchecked for rise -fall time, peak duration, and total duration by placing a storage type oscilloscope across the earphone output of the audiometer. Three equivalent durations were used in this test: 20, 200, and 400 msec. Equivalent duration (t) is defined as 31:- + P = t, 1 3 where‘ris the rise -decay time and P is peak duration. Peak dura- tion is defined as the total time the stimulus is at peak amplitude. In order to obtain the equivalent durations of 20, 200, and 400 msec with the constant rise -decay time of 10 msec, the peak durations were 13.33, 193. 33 and 393. 33 msec. Therefore, the total durations were 33.33, 213. 33 and 413. 33 msec. The stimulus envelope is shown in Figure 2. A single stimuluswas presented each second, and the Bekesy attenuation rate of 2. 5 dB per second was employed. The intensity calibration for-this apparatus was checked periodically by setting the electronicswitch to pass a continuous tone 1Peter J. Dallos and Wayne 0. Olsen, "Integration of Energy at Threshold with Gradual Rise -Fall TonePips, " Journal 91 the Acoustical Society of America, 36 (April, 1964), 744i 55 and by checking the earphone output with the sound level meter and artificial ear. — - Peak Duration— '— decay time— — time-j . l ' ————— Total Duration ————— ' Figure 2. —-A diagram of the brief -tone signal. Procedures Each subject was scheduled for a two -hour appointment at the Michigan State University Hearing Clinic. Prior to the auditory testing, each subject was interviewed as mentioned earlier in this chapter. Following the interview, the subject was seated in the test suite and the following instructions were read to him: You are now going to hear some tones. First, all the tones will be in your (right or left) ear and then in the other ear. When you hear the tone please push this button. Some of the tones will be very faint, but push the button whenever you think you hear. the tone. Are there any questions? 56 The routine air -conduction thresholds were then obtained via the Hughson -Westlake technique. 1 The right ear was always tested first unless the subject indicated that his hearing‘was better in his left ear. The examiner did all the testing from the control room of the two -room test suite. The thresholds were always obtained in the following order: 1000, 2000, 3000, 4000, 8000, 1000, 500, 250, and 125 Hz. Following the routine air -conduction threshold testing, the examiner‘returned to the test room and read the following instruc- tions to the subject: Now you are going to hearthe same kind of tone through this apparatus (showing the bone vibrator); however, you will hear a noise in this ear (pointing to the subject' 3 ear'that would receive the masking agent). Try not to pay attention to the noise, but rather-listen very carefully for the tone. When you hear the tone please push the button. Do you have any questions? For this phase of the test, the bone vibrator'was placed on the mastoid of the test ear, and the opposite ear received 10 dB of effective narrow -band masking that was centered on the test frequency. Thresholdswere obtained via the Hughson -Westlake technique and in the following order: 1000, 2000, 3000, 4000, 500, and 250. Hz. At the completion of the bone conduction testing, the test earwas selected. This was done by averaging the air conduction 1Raymond Carhart and James J erger, "Preferred Method of Determination of Thresholds, " Journal of Speech and Hearing Dis“- orders, 24 (1959), 330-345. a nil ... i! oh . SIP-run! 24 u Fin-7'4 5 ... o 57 thresholds at 2000, 4000, and 8000 Hz and then choosing the poorer ear. If the pure -tone air conduction results were identical, alternate ears were chosen. For example, if the previous subject' 3 right ear had been the test ear, this subject' 3 left ear would become the test ear. If one ear seemed to be getting used more than the other, it was balanced out whenever possible. It was explained to the subject that only one ear-would con- tinue to be tested, and then the following instructions were read to him: This test is divided into two parts. On the first part, you will do just as you did earlier. Push the button each time you hear the little beep, no matter how soft it is. On the second part, you will heara steady tone and you are to hold the button down as long as the tone is on. If the tone goes away and you can no longer hear it, release the button. When the tone comes back, depress the button and keep it down as long as you hear the tone. Do you have any questions? The tone decay test was always administered first at 500 Hz followed by 1000 and 4000 Hz. The test was always begun at threshold and increased in 5 dB steps until the subject was able to hear it for 60 seconds. The tone -decay score was the number of dB the signal had to be raised above threshold for the subject to hear it for 60 seconds. The Short Increment Sensitivity Index (SISI) test was administered next at the frequencies 1000, 2000, and 4000 Hz, in that order. The test was always administered 20 dB above the 58 subject' 3 threshold as determined immediately prior to the SISI test at each of the three frequencies. The tone was then put to its proper sensation level of 20 dB, and two increments each were presented at levels of 5, 4, 3, and 2 dB before the test actually began. The remainder of the test was administered in the routine manner;1 and at each frequency tested, each subject received twenty 1 dB incre- ments,‘ for which he received 5% credit for each correct response. Before the test, these instructions were read to the subject: This test is divided into two parts. On the first part, you will do just as you did earlier: push the button each time you hear the little beep, no matter how soft it is. On the second part, you will hear a steady tone for about two minutes. Every once in a while you may hear a little jump in the loudness. Each time you hear this little jump in loudness, push the button. If you think you heard the jump, but you' re not sure, don't do anything. Are there any questions? Speech audiometry was next in the battery of tests; a speech reception threshold was obtained first, followed by speech discrimi- nation. For the spondee testing, the subject was given a list of the CID Auditory Test W-l words and was asked to read them out loud to the experimenter. When this was completed, the following two sets of instructions were read to the subject: 1James Jerger, Joyce Shedd, and E. R. Harford, "On the Detection of Extremely Small Changes in Sound Intensity, " Archives of Otolaryngology, 60 (Feb., 1959), 200-211. 21bid., p. 203. 59 You are now going to hear these same words, but in a different order. You are to repeat the words which will be getting fainter and fainter. If you are not sure of the word, please guess. You will hear the words only in the (right or left) ear. Do you have any questions? You will then hear some common, everyday, one -syllable words. There will be two lists of words, the first list will be very faint and the second list will be much easier to hear. All you have to do is repeat the one -syllable word. Each test word will be preceded by the phrase, "You will say ." Again, please guess if you are not sure of the word: however, only say the word you hear or think you hear. If you talk or explain that you cannot understand the words, it is likely to cause you to miss the following word. Are there any questions? The speech reception threshold (SRT) was obtained in a descending order with two words being presented with each 2 dB decrement. The 2 dB decrements began after two spondees had been correctly repeated at hearing levels of 40 dB, 30 dB, 20 dB; then the test began at 10 dB (attenuator setting). Each correct response was credited as 1 dB, and the test ended when five out of six words were missed. The total number correct minus two was then subtracted from the 10 dB starting point and that was called the SRT. Speech discrimination was measured in the test ear first at a sensation level (SL) of 14 or 15 dB and then at a level of 40 or 41 dB. 1 This test was scored in the typical clinical manner of determining the percentage of correct responses. 1Because the Grason -Stadler speech audiometer' s attenuator is marked in 2 dB steps, it was not possible to always use a sensa- tion level of 15 and 40 dB. 60 Following completion of the speech audiometry, these instructions were given for the Bekesy audiometry portion of the test battery: This test is also divided into two parts. . In the first part, the tone you are to listen to is pulsed (beep, beep, beep, beep) and in the second part the tone is continuous, not pulsed (beeeeeeeeeeeeeep). Press this switch as soon as you hear the tone. Release the switch as soon as it disappears. . Never‘let the tone get loud and never let it disappear for long. Press as soon as you hear it and release as soon as it disappears. Do you have any questions? The Bekesy test was then administered. in the conventional way, beginning at 100 Hz and continuously sweeping to 10, 000 Hz with an attenuation rate of 2. 5 dB per second. The interrupted stimulus was always presented first. In the pulsed segment of the test, the pulses occurred at a rate of 2. 5 per second. When the continuous Bekesy tracing was completed, the subject heard these instructions for the final test, brief—tone audiometry: In this test you will hear a series of short beeps, and you have to make them go away and come back, just as you did on the last test. Again all the tones will be in the (right orleft) ear. Do you have any questions? For this test, 2000 Hz was always tested first followed by 4000 Hz. Thresholds were obtained for the shortest time first and the longest time last. Thresholds were determined by taking the average of five consecutive high peaks and five consecutive low peaks. The 61 variable of interest in this test was the difference in thresholds between the 20 and 200 msec equivalent durations. The 400 msec equivalent duration can be used to compare with the pure -tone threshold obtained in the conventional manner and also to check the linearity of the temporal integration function, but in this study the main interest was in the threshold differences between 20 and 200 msec. The order of presentation for each of the tests just described was held constant for each of the subjects. No attempt was made to randomize the test order because it was felt that if a clinical sympto- matology was to be found, it would be best to find it by presenting the above tests in the approximate manner and order that would likely be found in an audiological clinic. CHAPTER IV RESULTS AND DISCUSSION The general purpose of this chapter is to present the data that were obtained by utilizing the procedures described in Chapter III. The specific purpose of this chapter is to answer the questions posed. in Chapter I. This was accomplished by examining the raw data from the thirty subjects in terms of the means, standard deviations and ranges. Thesevwere obtained by transfer- ring the individual raw scores forxeach subject from each test (Appendix B) onto punch cards and putting these through a Control Data Corporation 3600 digital computer employing the MDSTAT1 routine. This routine was chosen because there were missing data from two subjects who were adopted and consequently were not familiar with their families' history of diabetes. The resulting descriptive examination of the data from all tests follows. 1Michigan State University, Agricultural Experiment Station, "Calculation of Basic Statistics When MissingData Is Involved (The MDSTAT Routine), " STAT Series Description No. 6 (December, 1966). 62 63 Hearing Threshold Level in dB ISO 0 10* I ‘ , ‘SCD ” \i \ C) - l \ 30 i a I \ \ \ 40 T If p I 1 50 T l 60 i I 70 O—() Right ear mean for 30 subjects 30 — X—X Left ear mean for 30 subjects 90. O--O Highest right ear threshold X—— ->( Highest: left ear threshold 100 ‘ 7 I ~ - L l , 125 2.50 500 1000 2000 4000 8000 Frequency in Hz Figure 3. --Audiogram showing mean air -conduction thresholds for thirty subjects and the highest thresholds obtained at each frequency. 64 Test Results Conventional pure -tone audiometry. --Figure 3 illustrates the results from the conventional pure -tone testing by showing the mean air -conduction thresholds from the 30 subjects plotted to the nearest 5 dB. Also shown are the poorest thresholds obtained at each frequency. The bone -conduction results were not plotted because they were within 5 dB of the air -conduction thresholds. The recorded mean thresholds for the two ears were within normal limits and were similar, being 5 dB apart at two frequencies, 250 and 4000 Hz. Two subjects had thresholds poorer than 25 dB. These thresholds were 30 and 35 dB in the left ear at 8000 Hz. The other pure -tone thresholds for these two subjects were typically just below the group mean at each frequency. The actual mean thresholds, expressed in hearing level (HL) and the standard deviations from the conventional pure -tone air -conduction testing are shown in Table 3. Inspection of the threshold values reveals the highest threshold to be 10. 3 dB at 8000 Hz in the left ear. This was also the same frequency and ear that demonstrated the largest standard deviation (8. 8 dB). The standard deviations ranged from this 8. 8 dB to 3. 8 dB at 250 Hz in the right ear. 65 Table 3. --Mean pure -tone air -conduction thresholds in dB HL and standard deviations for thirty subjects. Right ear Left ear Frequency Mean Standard Mean Standard Thre shold D eviation Thre shold Deviation 125 6.8 5.6 5.0 4.7 250 8.2 6.8 6.8 3.8 500 2.7 5.7 4.7 5.6 1000 2.2 5.7 2.3 5.0 2000 -1.8 5.9 -2.2 5.5 3000 2.5 5.4 2.0 6.8 4000 3.5 6.6 7.7 7.0 8000 8.3 7.9 10.3 8.8 Speech audiometry. —- The mean speech reception threshold (SRT) for the 30 subjects was 4. 27 dB HL (25. 27 dB sound pressure level). This is within normal limits. The range of SRT' 3 was from -1 dB to +14 dB HL and the standard deviation was 3. 4 dB. The speechdiscrimination scores at the +15 dB sensation level (SL) ranged from 64 to 88% with a mean of 75. 4% and a standard deviation of 5. 4%. This compared favorably with the results obtained by Rintelrnann1 using the same list of words and testing 12 normal- hearing young adults. They were tested at a +16 dB SL and had a mean score of 74% with a standard deviation of 8. 5%. 1William F. Rintelmann, in an unpublished study conducted at Michigan State University. 66 When tested at a +40 dB SL the mean speech discrimination score for this sample of juvenile -onset diabetics was 98. 3% with a range from 92 to 100% and a standard deviation of 1. 9%. Table 4 summarizes the speech discrimination results. Table 4. -- The means, ranges and standard deviations of the speech discrimination scores. Sensation M Ra e Standard level ean ng deviation +15 (:13 75.4% 64- 88% 5. 4% +40 dB 98.3% 92 - 100% 1. 9% Special pure -tone audiometry. -- The first special test administered to these subjects was tone decay. This test was given at 500, 1000, and 4000 Hz. The mean scores were 5. 2 dB of decay at 500 Hz, 4. 7 dB of decay at 10004 Hz and 5. 3 dB of decay at 4000 Hz. These mean scores can be considered clinically negative. The largest tone -decay score was 20 dB, and that occurred at 1000 Hz. Table 5 presents the meanresults of the tone -decay test and standard devia- tions. The second special test the subjects responded to was the SISI test. The mean scores for the SISI test were 17.0, 17.3 and 15. 3% respectively for the three frequencies 1000, 2000, and 4000 Hz. 67 Table 5. --Mean tone decay scores and standard deviations. Frequency Mean amount of decay Standard deviation 500 Hz 5.2 dB 4. 8 dB 1000 Hz 4. 7 dB 4. 9 dB 4000 Hz 5. 3 dB 4. 7 dB A total of 90 SISI tests were presented; each of the 30 subjectswas tested at three frequencies. Out of these 90 tests there were 8 positive scores (60% or above) which can be accounted for by 5 sub- jects. Table 6 summarizes the results of the SISI test. Table 6. --Mean SISI scores, standard deviations and highest score at each frequency. Standard Highest Frequency Mean (1 . . ev1ation score 1000 Hz 17.0% 23. 9% 85% 2000 HZ 17. 3% 26. 8% 90% 4000 Hz 15.3% 20.2% 70% Statistically the eight positive scores contributed to the large standard deviations shown in Table 6, whereas clinically they are difficult to explain. In two cases these scores appeared to be 68 spurious. Subject No. 2 had SISI scores of 65, 15, and 0% respectively at the three frequencies 1000, 2000, and 4000 Hz; and Subject No. 3 had scores of 10, 85, and 20% at the same three frequencies. The other scores to all tests for four of these five subjects are unremarkable, whereas Subject No. 21, who had two of the three highest scores on the SISI test, did exhibit some other unusual scores which will be discussed later. Over half (53. 3%) of the SISI scores were 5% or lower, and 78. 8% of the scores were 30% or lower. Sweep -frequency Bekesy audiometry was the third special test administered. All 30 subjects were quick to learn the Bekesy test procedure, and a careful study of the audiograms revealed all 30 to be Type I according to the J erger1 classification. Without an actual determination of the Bekesy thresholds it could be seen that these threshold tracings very closely approximated the thresholds obtained via conventional audiometric testing. The widths of the tracings did not appear to be abnormally large, as suggested by Axelssonand Fagerberg, 2 nor were they abnormally small tracings. The large majority of the tracings were around 8 to 10 dB in width. 1Jerger, "Bekesy Audiometry in Analysis of Auditory Dis - orders, " p. 278. 2Axelsson and Fagerberg, "Auditory Function, " pp. 60 -61. 69 Brief -tone audiometry was the last of the special tests, and the purpose of administering this test was to determine the difference in thresholds between the equivalent durations of 20 and 200 msec. Olsen and Carhart1 and Nerbonne2 used these durations on normal hearers and found differences of 6. 9 and 7. 4 dB respectively at 4000 Hz. In the present study differences of 7. 19 and 5. 94dB respectively were found at 2000 and 4000 Hz. The standard devia- tions were 2. 9 dB at 2000 Hz and 2. 4 dB at 4000 Hz. Table 7 shows the mean threshold. differences for the 30 subjects between the equivalent durations of 20 and 200 msec. Also included are the standard deviations for these differences. Table 7. --Mean threshold differences, and their standard deviations, between the equivalent durations of 20 and 200 msec. , Frequency , Threshold difference Standard deviation 2000 Hz 7.17 dB 2.9dB 4000 Hz 5.94dB 2.4 dB A 1Wayne 0. Olsen and Raymond Carhart, "Integration of Acoustic Power at Threshold by Normal Hearers, " Journal of the Acoustical Society of America, 10 (1966), 591 -599. 2M. A. Nerbonne, "A Comparison of Brief Tone Audiometry with Other Selected Auditory Tests of Cochlear Function” (unpub- lished Ph.D. dissertation, Michigan State University, 1970). 70 To determine whether the 5. 9 dB difference obtained in this study at 4000 Hz betweenthe equivalent durations of 20 and 200 msec was significantly different from that obtained at the same frequency and with the same equivalent durations in Nerbonne' 5 study with 20 normal -hearing subjects, a 1 test was performed. In order to perform this _t_ test a pooled standard deviation was used because of the unequal N' s. The hypothesis under test was that the two means were equal and a significance level of .05 was chosen. A_t_ of 2. 5 was obtained which, with 48 degrees of freedom, was significant. Therefore, the hypothesis of equal means had to be rejected. Statistically, this can be explained largely on the basis that the variance of this study' s sample population was more than three times that of Nerbonne' 8 sample population. This may have been due to the fact that the two sample populations were not matched. Nerbonne' 3 population was restricted in age to a college population, whereas the juvenile -onset diabetics ranged in age from 12 to 44 years and were not as homogeneous a population as college students. Clinically the explanation could be that this population actually responded differently to the test because of a difference in the auditory mechanism. If this were true, brief -tone audiometry may have located a sub -clinical pathology that was undetected by the other tests. 71 Subject No. 21 was mentioned earlier as having high SISI scores and some other unusual results. It would seem appropriate at this time to discuss these unusual results. On the SISI test Subject No. 21 had scores of 85, 60 and 70% respectively at 1000, 2000 and 4000 Hz. His pure -tone audiogram, however, showed no abnormalities nor did his SRT or speech discrimination scores. The other results that were abnormal were a 2. 6 dB difference between the 20 and 200 msec duration thresholds at 4000 Hz and his Bekesy audiogram. On the latter, the continuous tracing was typically better than the interrupted tracing throughout the frequency range tested. From 125 to 3000 Hz the continuous tracingaveraged 6 dB better than the interrupted tracing, whereas at 4000 and 8000 Hz the difference was 28 and 20 dB respectively. The thresh- olds were close to zero from 125 to 4000 Hz where they dropped sharply by 40 dB. Theresulting Bekesy threshold at 8000 Hz was much poorer than the threshold obtained routinely. These responses are unusual and cannot be considered as reliable and valid responses from a normal hearing individual. There are two explanations for these atypical responses. The first, and perhaps the most tenable, is that they simply are not reliable test results. If this subject were to be retested, it is possible his responses to these same tests would more closely resemble those of the other 29 subjects. The 72 second possibility is that these unusual responses were due to a sub -clinical cochlear pathology that did not manifest itself on the other tests. Statistical Results In this section a closer examination of the data will focus on answering three questions from Chapter I. These questions asked what effect age at onset of diabetes mellitus, current age of the diabetic, and duration of diabetes had on the responses of this sample population to the auditory tests administered to them. To test for significance of these three effects, it was decided to choose four variables that would be representative of all the tests rather than testing for significance with all the possible variables (these totaled 36) since it was shown that the data were within normal limits. Taken into account was the fact that previous research'with diabetics has shown a high -frequency hearing loss (Ancona, 1 J orgensen and 2 Buch, Zelenka and Kozak3); therefore, two of the variables chosen 1Ancona, "Hearing Disorders in Juvenile Diabetics, " pp. 435-443. 2 J orgensen and Buch, "Studies on Inner Ear Function in Diabetics," p. 362. 3Zelenka and Kozak, ”Disorder in Blood Supply of the Inner Ear, " p. 316. Ehrlil"; Rigs 73 to be studied were SISI and brief -tone audiometry at 4000 Hz. The two other variables chosen to be studied were speech discrimination at +15 dB SL and the speech reception thresholds. Speechdiscrimi- nation at +40 dB SL and tone decay were not chosen because their range of scores was small. In order to test for the effect of duration, age at onset, and current age on the four chosen variables, one -way analyses of variance tests were performed following the model outlined by Dixon and Massey. 1 When the effect of age at onset was being tested, the 30 subjects were divided into 3 groups of 10 according to their age at onset. When the effect of duration was being tested, the 30 sub- jects were divided into 3 equal groups according to how long they had been diabetic. Finally, when current age was being tested for an effect, the sample population was again divided into three equal groups. Table 8 shows how the 30 subjects were divided for each 0f the above three effects. In the one -way analysis -of-variance tables these group divisions became the column headings, and the appropriate 10 individual raw scores were row entries. It was decided to use a significance level of .05 to test the significance of the obtained F ratios. 1Wilfred J. Dixon and Frank J. Massey, Introduction to Statistical Analysis (2d ed. ; New York: McGraw -Hill Book Co. , Inc. , 1957). pp. 149-152. 74 Table 8. --Division of the thirty subjects into the appropriate groups for the one -way analyses -of -variance tests. 4. Groups A B C Age at onset 12 - 123 months 124- 181 months 182 -299 months Duration 20 - 86 months 87 - 137 months 138- 444 months Current age 146-245 months 246 - 320 months 321 - 550 months A_ge at onset. -- To determine the effect age at onset of diabetes had on hearing, as measured by the four variables, it was necessary to have a null hypothesis which could be tested. The null hypotheses tested in this section were as follows: 1. H01 = The age at the onset of diabetes mellitus had no 'significant effect on SISI scores at 4000 Hz. 2. H02 = The age at the onset of diabetes mellitus had no significant effect on the speech reception threshold. 3. - H03 = The age at the onset of diabetes mellitus had no significant effect on the speech discrimination score at a sensation level of +15 dB. 4. H04 = The age at the onset of diabetes mellitus had no significant effect on the threshold difference between the equivalent durations of 20 and 200 msec at 4000 Hz. 75 .Table 9 displays the three groups' mean scores for each of the four variables under consideration. Table 9. --Group means according to "age at onset" for the SISI test at 4000 Hz, SRT, speech discrimination at a +15 dB SL and the difference between the 20 and 200 msec duration thresholds at 4000 Hz. Mean scores Variable Group A Group B Group C (15-123 (124-181 (182-299 months) months) months) SISI at 4000 Hz 16. 0% 12. 5% 17. 5% SRT 4.2 dB 5.0 dB 3.6dB Discrimination at +15 dB SL 72. 0% 78.0% 76. 2% 20-200 msec difference at 4000 Hz 5. 4 dB 6. 4 dB 6.0 dB Table 10 lists the results of the analyses -of-variance tests by showing the obtained F ratios. Table 10. -- F ratios obtained from a one -way analysis -of—variance test between the variable "age at onset" and selected auditory variables. Variables F ratio F. 05 SISI at 4000 Hz . 0.151 3. 35 SRT 0. 441 3. 35 Discriminationat +15 dB SL 4. 027* 3. 35 20-200 msec difference at 400 Hz 0. 406 3. 35 *Significant at the 0.05 level of significance. 76 From this table it can be seen that H01, H02, and H04 failed to be rejected at the 0.05 level of confidence, indicating that the age at onset had no effect on these three variables. Clinically, it can be seen from studying Table 9 that the differences between the means for each variable are insignificant. There is, however, statistical evidence that H03 should be rejected, demonstrating that the age at onset did indeed have an effect on the speech discrimination scores at the +15 dB SL. Again referring back to Table 9, it can be seen that Group A, with the earliest onset, did have the lowest mean discrimination score, whereas Group B had the highest mean discrimination score. The difference between these two means is 6%. Whether this amount is clinically significant or not is questionable. To fully answer this question it might have been helpful to test speech discrimination at more sensation levels. Duration. -- To study the effect of duration of diabetes mellitus on the four variables being considered as representative of hearing, it was again necessary to develop four null hypotheses. They were as follows: 1. H = Duration of diabetes mellitus had no significant 01 effect on SISI scores at 4000 Hz. 77 2. H02 = Duration of diabetes mellitus had no significant effect on the speech reception threshold. 3. H03 = Duration of diabetes mellitus had no significant effect on speech discrimination scores at a sensa- tion level of +15 dB. 4. H04 = Duration of diabetes mellitus had no significant effect on the threshold difference between the equivalent durations 20 and 200 msec at 4000 Hz. Table 11 shows the grouping of the thirty subjects into groups of ten according to ”duration of diabetes" and lists the mean scores on the four variables. Table 11. --Group means according to "duration of diabetes" for the SISI test at 4000 Hz, SRT, speech discrimination at a +15 dB SL and the difference between the 20 and 200 msec duration thresholds at 4000 Hz. Mean scores Variables Group A Group B Group C (20-86 (87-127 (138-144 months) months) months) SISI at 4000 Hz 15. 5% 22. 0% 8. 5% SRT 5.8dB 3.8dB 3.2 dB Discrimination at +15 dB SL 78. 6% 74. 2% 73. 4% 20-200 msec difference at 4000 Hz 5. 9 dB 6.1 dB 5. 8 dB 78 The results from the one —way analyses -of -variance tests between the duration of diabetes and the above four variables are exhibited in Table 12. None of the F ratioswere statistically sig- nificant, indicating that all four null hypotheses failed to be rejected. This suggests that the duration of the diabetes had no effect on the hearing of the subjects in this study. Table 12. -- F ratios obtained from a one -—way analysis -of -variance test between the variable "duration of diabetes" and selected auditory variables. Variables F ratio F 05 SISI at 4000 Hz 1.125 3. 35 SRT 1 . 658 3. 35 Discrimination at +15 dB SL 3. 167 3. 35 20-200 msec difference at 4000 Hz 0.050 3. 35 Current age. -- The last question to be asked by this investi- gation dealt with the effect the current age of the diabetic had upon hearing. Again it was necessary to put this question into four null hypotheses in order for it to be studied statistically. The resulting four null hypotheses were as follows: 1. H = The current age of the juvenile -onset diabetic had no 01 significant effect on the SISI scores at 4000 Hz. 79 — The current age of the juvenile -onset diabetic had no 02 significant effect on the speech reception threshold. 3 H03 = The current age of the juvenile -onset diabetic had no significant effect on the speech discrimination score at a sensation level of +15 dB. 4 H04 = The current age of the diabetic had no significant effect on the threshold difference between the equivalent durations of 20 and 200 msec at 4000 Hz. In Table 13 the subjects are grouped according to age and listed are each group' 3 mean scores on the variables under study. Table 13. --Group means according to "current age" for the SISI test at 4000 Hz, SRT, speech discrimination at a +15 dB SL, and the difference between the 20 and 200 msec duration thresholds at 4000 Hz. Mean scores Variables Group A Group B Group C (146-245 (246-320 (321-550 months) months) months) SISI at 4000 Hz 21. 5% 13.5% 11. 0% SRT in hearing level 4. 8 dB 4. 3 dB 3. 7 dB Discrimination at +15 dB SL 74. 8% 77. 0% 74. 4% 20-200 msec difference at 4000 Hz 5. 6 dB 6. 6 dB 5. 6 dB 80 The results of the one -way analyses -of-variance tests are displayed in Table 14. Every obtained F ratio was less than unity, meaning that each null hypothesis failed to be rejected. This sug- gests that the present age of the diabetic had no effect on his hearing as represented by these four'variables. Table 14. --F ratios obtained from a one -way analysis -of-variance test between the variable "current age” and selected auditory variables. Variables F ratio F 05 SISI at 4000 Hz 0.721 3. 35 SRT 0. 271 3. 35 Discrimination at +15 dB SL 0. 650 3. 35 20-200 msec difference at 4000 Hz 0. 531 3. 35 Discussion The above descriptive and statistical presentation of the data has demonstrated that the responses of the 30 juvenile -onset diabetics to all tests employed were within normal limits. The pure -tone audiogram and the speech reception threshold, according to the definition established. in Chapter ,1, reveal no hearing loss. Also, the responses to the other tests reveal no abnormalities. Therefore, the answer to the first question asked in Chapter I is 81 negative. That question was, "Do juvenile -onset diabetics demonstrate a hearing loss not accounted for by other causes?" During the initial interview, none of the 30 subjects reported ever having a hearing loss. Because of the interest shown by the subjects to the auditory testing and because of the consistency of the subjects' responses, the obtained test results are believed to be a valid and reliable indication of their auditory function. Referring back to Table 2 in Chapter II, it can be seen that the results of this study agree with the results of Camisasca, Profazio and Baravelli, and Marshak and Anderson, who reported no hearing impairment because of diabetes mellitus in subjects under the age of 44 years. It can be seen, however, that this study' 8 results disagree with some of the other studies. The reason for this discrepancy is unknown. Because those studies were done in Europe, it is a possibility that diabetes mellitus is in some manner treated differently there. The quality of the medical care, the quality of insulin, the type of diet and the populations studied could be a part of the explanation of the difference. It is also a possibility that the tests utilized in this study were not sufficiently sensitive to detect sub -clinical auditory disorders. Such disorders, though, could not account for the large percentages of hearing losses reported from Europe because most of those studies reported using only pure -tone audiometry. 82 Two statistically significant results were found. The first was with the effect "age at onset" had on the speech discrimination scores at the +15 dB SL. The group'with the earliest "age at onset" had the poorest discrimination scores at that sensation level. The second statistically significant result was found with brief -tone audiometry at 4000 Hz. When compared to a group of 20 normal hearers, the juvenile -onset diabetics had a statistically significant lower score. The clinical significance of these cannot be evaluated presently because all other test results were essentially normal; however, it is possible that the discrimination scores at the +15 dB SL and the brief -tone audiometry scores at 4000 Hz are indicative of subtle auditory dysfunction. CHAPTER V SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Summary Previous audiometric studies with diabetics have reported hearing losses in up to 81% of their sample populations. In the few studies reported that examined juvenile -onset diabetics the reported incidence of hearing impairment ranged from 0 to 70%. The major purpose of this study was to determine the audiological manifestations of 30 juvenile -onset diabetics. These 30 subjects all had a medical diagnosis of diabetes mellitus prior to the age of 25 years, and at the time of testing they were between the ages of 12 and 45 years. Seventy -three percent of these diabetics reported a positive family history of diabetes, whereas two of the subjects reported having diabetic complications. These 30 subjects responded to a battery of audiological tests using both pure -tone and speech stimuli. The tests admin- istered included conventional pure -tone audiometry, tone decay, 83 84 SISI, speech reception thresholds, speech discrimination, sweep- frequency Bekesy audiometry and brief-tone audiometry. Mean responses to all the tests were within normal limits. The poorest mean pure -tone threshold was 10. 3 dB hearing level, and there were no air -bone gaps greater than 5 dB. The mean speech reception threshold was 4. 3 dB hearing level, whereas the mean speech discrimination scores were 75.4% at the +15 dB sen- sation level and 98. 3% at the +40 dB sensation level. The mean responses to the special pure -tone tests could be classified as negative. The mean tone decay scores were 5. 2, 4. 7, and 5. 3 dB respectively at 500, 1000, and 4000 Hz, whereas the mean SISI scores were 17.0, 17.3, and 15. 3% at 1000, 2000, and 4000 Hz respectively. All 30 Bekesy audiograms were classified as Type I and the Bekesy thresholds closely approximated the pure —tone thresholds obtained via conventional audiometry. In brief -tone audiometry the difference between the thresholds at the equivalent durations of 20 and 200 msec at 2000 Hz was 7. 2 dB, whereas the difference at 4000 Hz was 5. 9 dB. Conclusions Within the limits imposed by the design of this study and on the basis of the analysis of the data, the following conclusions appear warranted. For this sample of 30 juvenile -onset diabetics: 85 1. There-was no hearing loss discernible by conventional pure -tone audiometry at the octave frequencies from 125 to 8000 Hz including 3000 Hz, nor by speech discrimi- nation testing, nor by any of the other special tests administered. 2. The "age at onset" had a significant effect on the discrimi-s nation of faint speech. 3. Juvenile -onset diabetics behave differently on brief-tone audiometry at 4000 Hz than normals. 4. There were no other clinically significant auditory mani— festations because of the effects of (a) the duration of the diabetes from the time of the medical diagnosis, (b) the age at the onset of the diabetes, or (c) the current age of the diabetic. A word of caution regarding these conclusions is appropriate here. The subjects in this investigation developed their'diabetes prior to the age of 25 years, whereas the large majority of diabetics develop their diabetes much later in life. As was pointed out in Chapter II, the two types of diabetesare not the same. Therefore, it is important that the conclusions from this study not be generalized to different diabetic populations until an adequate amount of research permits such generalizations. 86 Recommendations for Further Research It has been demonstrated by this investigation that these 30 juvenile -onset diabetics have suffered no apparent loss of auditory function. This is not in agreement with many of the European studies that have found a high incidence of auditory impairment in both juvenile - and maturity -onset diabetics. This disagreement irnme - diately raises the question of whether or not diabetes mellitus may be treated differently in Europe, whether there are differences in EurOpean insulin or other methods of controlling and regulating the diabetes, or-whether differences in dietary habits could cause dif- ferent or more severe complications, etc. This matter is worth further investigation. Another question that is raised concerns theaudiological testing from the European studies. In the majority of the studies it was not reported what type of audiometric equipment was used, what type of test environment was present, nor what their definition of hearing loss was. The maximum age of the diabetics in this study‘was 44 years. Before any conclusive statements can be made about the effect diabetes has on hearing, this age limit must be increased in a study of both juvenile -onset and maturity -onset diabetes. Because of the arterio- and atherosclerosis caused by diabetes, it is possible that III I]! III III..." Ill). 87 an older population would demonstrate a hearing impairment, possibly akin to presbycusis but on an earlier and more severe scale. A study involving older diabetics should be undertaken to support this speculation. In this investigation there was no control or measurement of blood -sugar levels. Further investigation into the effect blood- sugar levels have on hearing should be undertaken. This could be accomplished by having subjects respond to auditory stimuli during various degrees of hypo- and hyper -glycemia. None of the subjects in this study (with the possible (excep- tion of the two who were adopted and were lacking the information) was conceived, carried and delivered by a mother who was diabetic. It has been reported in the literature that such a child could suffer from a hearing loss, but thus far there have been no reported audiometric studies on a child or adult who fits this description. To study and report on such a population‘would add important informa- tion to the information available on the relationship between diabetes and audition. Another possible study would investigate the possiblity that insulin, or the lack of it, may be a contributing factor in the effects diabetes has been reported to have on hearing. All the subjects in this study were using insulin daily. This is generally true in the 88 case of juvenile -onset diabetics. However, a study involving two older groups of diabetics, one group not using insulin and the other group using insulin, could offer some further evidence to help answer this question. One further study that should be undertaken would involve testing at least the same type of population studied in this investiga - tion and possibly some older diabetics as well. Recall that in the present investigation there was a statistically significant effect between "age at onset" and discrimination at the +15 dB sensation level. In this new study it would be beneficial to test speech dis- crimination at lower sensation levels and also in noise. By further testing the speech discrimination under these different conditions, it could be determined whether the statistical significance found in this study was clinically significant. It would also be interesting in this new study to present the SISI increments at a higher sensation level. Harford1 suggests that with mild hearing losses of less than 35 to 40 dB, a sensation level of 25 or 30 dB may show positive SISI scores in cochlear-lesions where otherwise a negative score may result when the customary 20 dB level is employed. The subjects from the present investigation 1 , Earl R. Harford, "Clinical Application and Significance of the SISI Test, " in Sensorineural Hearing Processes and Disorders, , ed. by A. Bruce Graham (Boston: Little, Brown and Co. , 1967), p. 231. 89 fit this criterion of a mild hearing loss less than 35 dB, and there is the possibility the presence of a subclinical cochlear lesion could be proved or disproved. BIBLIOGRAPHY BIBLIOGRAPHY Books Clarke, C. -A. "Genetic Aspects of Diabetes. " Diabetes Mellitus. Edited by L. J. P. Duncan. Edinburgh: Edinburgh Uni- versity Press, 1966. Dally, Ann G. A Readers Guide to Modern Medicine. New York & ,Evanston: Harper 8: Row Publishers, 1966. Danowski, T. S. Diabetes Mellitus with Emphasis on Children and Young Adults. The Williams & Wilkins Co. , 1957. Dixon, Wilfred J. , and Massey, Frank J. Introduction to Statistical Analysis. 2d ed. New York: McGraw -Hill Book Co. , Inc. , 1957. Dolger, Henry, and Seeman, Bernard. How to Live with Diabetes. Revised ed. New York: Pyramid Books, 1965. Dorland' s Illustrated Medical Dictionary. 24th ed. Philadelphia & London: W. B. Saunders Co., 1965. Duncan, Garfield G. Diseases of Metabolism. 5th ed. Philadelphia 8: London: W. B. Saunders Co., 1964. Dunlop, Sir Derrick. "Opening Address. " Diabetes Mellitus. Edited by L. J. P. Duncan. Edinburgh: Edinburgh Uni- versity Press, 1966. Fagerberg, S. -E. "The Angiopathy of Diabetes. " Diabetes Mellitus. Edited by L. J. P. Duncan. Edinburgh: Edinburgh Uni- versity Press, 1966. 90 91 Farquhar, James W. "Child of the Diabetic Mother. " Diabetes Mellitus. Edited by L. J. P. Duncan. Edinburgh: Edin- burgh University Press, 1966. Gardner, Ernest. Fundamentals of Neurology; 4th ed. Philadelphia 8: London: W. B. Saunders Co., 1963. Harford, Earl R. "Clinical Application and Significance of the SISI Test. " Sensorineural Hearing, Processes and Disorders. Edited by A. Bruce Graham. Boston: Little, Brown and Co. , 1967. Hollender, Abraham R. Office Practice of Otolarynglogy; Phila- delphia: F. A. Davis Co., 1965. J oslin, Elliot; Root, Howard: White, Priscilla; and Marble, Alexander. Treatment of Diabetes Mellitus. 10th ed. ; rev. Philadelphia: Lea & Febiger, 1959. Mahler, R. F. "Insulin Action on Arterial Tissue in Relation to Diabetes and Atheroma. " Diabetes Mellitus. Edited by L. J. P. Duncan. Edinburgh: Edinburgh University Press, 1966. Papaspyros, N. S. The History of Diabetes Mellitus. Stuttgart: Georg Thieme Verlag, 1964. Physicians' Desk Reference. 23d ed. Oradell, New Jersey: Medical Economics, Inc., 1969. Schmitt, George F. Diabetes for Diabetics. 2d ed. Miami: Diabetes Press of America, 1968. Sheppard, L. Benjamin. Current Concepts of Diabetes Mellitus with Special Reference to Ocular Changes. Springfield: Charles C. Thomas, 1955. Thompson, Willard 0., ed. The Diabetic Neuropathies. Springfield: Charles C. Thomas, 1953. Wilder, Russell M. A Primer for Diabetic Patients. 9th‘ed. Phila- delphia: W, B. Saunders Co., 1950. 92 Periodicals Ancona, F. "Considerations of Hearing Disorders in Juvenile Diabetics. " Arcispediale S. Anna Di Ferrara; Revista Trimestrale Di Scienze Mediche, 9 (1956), 435-443. Axelsson, A. , and Faberberg, S. -E. "Auditory Function in Diabetics." Acta Oto-Laryngologica, 66 (1968), 49 -64. Carhart, Raymond. "Clinical Determination of Abnormal Summa- tion. " Archives of Otolaryngology, 65 (1957), 32 -39. , and J erger, James. "Preferred Method of Determination of Thresholds. " Journal of Speech and Hearinfig Disorders, 24 (1959), 330-345. Dallos, Peter J. , and Olsen, Wayne O. "Integration of Energy at Threshold with Gradual Rise -Fall Tone Pips. ” Journal of the Acoustical Society of America, 26 (1964), 743 -751. Goodner, Charles J. "Newer Concepts in Diabetes Mellitus, Includ- ing Management. " Disease -a -Month. Chicago: Yearbook Medical Publishers, Inc. , September, 1965. Hoffman, Julius. "Peripheral Neuropathy in Children with Diabetes Mellitus. " Acta Neurologica Scandinavica, Supplementum 8, Volume 40, 1964. Hognestad, S. "Hereditary Nerve Deafness Associated with‘Diabetes. " Acta Oto-Laryngglogica, 64 (1967), 219-225. J erger, James F. "Bekesy Audiometry in Analysis of Auditory Disorders. " Journal of Speech and Hearing Research, 3 (1960), 275-287. ; Shedd, Joyce Lassman; and Harford, Earl. "On the Detection of Extremely Small Changes in Sound Intensity. " Archives of Otolaryngology, 69 (1959), 200 -211. J orgensen, M. Balslev. ”Changes of Aging in the Inner Ear, and the Inner Ear {in Diabetes Mellitus. Histological Studies. " Acta Oto -Laryngologica, Supplementum 188. 93 . "Sudden Loss of Inner Ear Function in the Course of Long- Standing Diabetes Mellitus. " Acta Oto-Laryngologica, 51 (1960), 579-584. . "Influence of Maternal Diabetes on the Inner Ear of the Fetus. " Acta Oto -I.aryngologica, 53 (1961), 49 -54. , and Buch, N. H. "Function of the Inner Ear and Cranial Nerves in Pregnant Diabetics. " Practica Oto -Rhino Laryn- gologica, 24 (1962), 111-116. , and . "Studies on Inner -Ear Function and Cranial Nerves in Diabetics. " Acta Oto-Laryngolggica, 53 (1961), 350-364. Joslin, Elliott P. "Diabetes for the Diabetics." Diabetes, 2 (1956), 137 -146. Kelmanen, G. "Aural Changes in Embryo of Diabetic Mothers." AMA Archives of Otolaryngology, 62 (1955), 357 -369. Koide, Y.; Tajima, S.; Yoshida, M.; and Konno, M. "Biochemical Changes in the Inner Ear Induced by Insulin, in Relation to the Cochlear Microphonics. " Annals of Otology, Rhinology and Laryngo‘logy. 69 (1960), 1083-1097. r ’— Levine, Rachmiel. "Diabetes Mellitus." Clinical Symposia, 15 (1968), 103-132. Marshak, Gabriel, and Anderson, Charles V. "Bekesy Audiometry with Juvenile ~Onset Diabetics. " Journal of Auditory Research, 8 (1968), 323 -330. Olsen, Wayne 0. , and Carhart, Raymond. "Integration of Acoustic Power at Threshold by Normal Hearers. " Journal of the Acoustical Society of America, 40 (1966), 591-599. .Wright, H. N. "Clinical Measurement of Temporal Auditory Sum- mation. " Journal of Speech and Hearing Research, 11 (1968). 109-127. Zelenka, J. , and Kozak, P. "Disorder in Blood Supply of the Inner Ear as Early Symptom of Diabetic Angiopathy. " Journal of Laryngolog and Otology, 79 (1965), 314-319. 94 . Unpublished Material Nerbonne, M. A. "A Comparison of Brief Tone Audiometry with Other Selected Auditory Tests of Cochlear Function." Unpublished Ph. D. dissertation, Michigan State University, 1970. Rintelmann, W. F. An unpublished study of speech discrimination utilizing N. U. Auditory Test No. 6, conducted at Michigan State University, 1968. Other Sources ANSI S3. 6- 1969. American National Standard Specifications for Audiometers. American National Standards Institute, Inc. , 143 Broadway, New York, New York 10018. Diabetes Source Book, U. S. Department of Health, Education, and Welfare. Public Health Service Publication No. 1168 (1964). Michigan State University, Agricultural Experiment Station. "Cal- culation of Basic Statistics When Missing Data Is Involved (The MDSTAT Routine), " STAT Series Description No. 6 .(1966). Tillman, Tom W. , and Carhart, Raymond. "An Expanded Test for Speech Discrimination Utilizing CNC Monosyllabic Words. " United States Air Force School of Aerospace Medicine, SAM-TR-66-55, June, 1966. APPENDICES APPENDIX A SUBJ E C T INFORMATION SHE E T 95 APPENDIX A: SUBJECT INFORMATION SHEET Subject No. Name Birthdate Sex_ Age— Address ’ Phone Month and year at medical diagnosis of diabetes Diabetes mellitus in family: Mother___ Father__ Brothers— Sisters Other Was your mother diabetic while carrying you Birthweight Duration of insulin treatment Frequency and amount of injec- tions Type of insulin How easy is your diabetes regulated 5 4 3 2 1 Extremely Difficult Trouble Once Easy Extremely Difficult in a While Easy Explanation (if necessary) What diabetic complications are present and how would they be rated? 5 4 3 2 1 Severe Definite Mild Just Non - A Problem Noticeable existent Retinopathy - Neuropathy (where) Nephropathy Vascular compli- cations (where) Other llll llil llll llll Other health problems (illnesses, surgery, allergy, etc.) History of vertigo History of tinnitus 96 Appendix A. - — Continued. Has any hearing loss or change in hearing been noted When did it begin Which ear is better Does it fluctuate Is there a history of hearing loss in the family Any personal history of middle ear infection Which of the following diseases did you have and were they severe: __ measles __ mumps __ scarlet fever __ diphtheria __ menningitis __ chicken pox encephalitis whooping cough Pregnancy or birth problems (rubella, anoxia, etc.) Ototoxic drugs (quinine, streptomycin, neomycin, kanamycin, aspirin) Head trauma or mechanical injury to ear Kernicterus (Rh incompatibility) Noise exposure (vocational, armed services, gunshot, farming, R & R music) Any previous hearing tests Where Any previous otologic exams or surgery Present medications (besides insulin) Notes: APPENDIX B SUMMARY OF RAW DATA 97 APPENDIX B: SUMMARY OF RAW DATA Air conduction thresholds ,_, Right ear Left ear .32.: 00000 00900 LGOOOOOOOLOOOOOOOO NLQOOOOOONLOOOOOOO Hmm—«vaoo—cmm—«vaoo 1 5‘ o 5 o o - 5 o 10 5 o 5 5 5 o o 5 - 5 2 5‘ 20 25 15 20 1o - 5 5 15 1o 15 1o 15 5 5 o 35 3 5‘ 1o 5 o o o - 5 - 5 - 5 o o - 5 0 —1o -10 o o 4 5‘ o 5 5 5 o - 5 - 5 - 5 5 5 5 o - 5 o 5 5 5 5‘ 5 1o - 5 — 5 o 5 o 15 5 5 o o - 5 10 5 10 6 5‘ o 5 o — 5 o 0 1o 5 - 5 5 o o o 5 1o 15 7 11. 5 5 o o o - 5 10 1o 5 10 10 o o 10 5 25 8 5‘ 15 15 - 5 - 5 o 5 5 20 1o 10 o o - 5 0 1o 15 9 54 1o 0 - 5 - 5 - 5 5 5 o 5 5 o - 5 o 15 1o 5 10 5‘ o o o 5 -1o 5 o 15 o o o o 0 -1o 5 1o 11 5‘ 1o 15 5 o - 5 - 5 o 5 5 10 10 o - 5 0 1o 5 12 ;5 5 5 o - 5 -1o 0 — 5 5 5 1o 5 0 -1o 0 0 1o 13 5‘ 5 5 o - 5 -1o 0 - 5 1o 10 1o 5 - 5 o o 5 5 14 14 o 5 o 5 -1o 5 10 1o 0 5 o - 5 -1o 0 1o 10 15 5‘ 5 5 o 5 o 15 o 10 o 5 1o 5 o 5 1o 5 16 5? 20 25 20 1o 10 10 1o 15 5 o 5 o o o o_ 10 17 5‘ 15 20 10 1o 5 5 o o 15 15 15 1o 5 5 1o 5 18 54 5 5 5 5 o o 5 5 5 5 o 5 1o 0 15 5 19 5‘ 5 5 o o — 5 - 5 - 5 1o 0 5 o - 5 - 5 -1o 5 5 20 54 5 5 o 5 o o 15 1o 10 5 5 5 - 5 - 5 5 20 21 14 o 5 o 5 -1o 5 10 o 5 5 o 5 o o 15 10 22 IP' 10 5 5 5 5 15 15 25 15 1o 5 5 o 10 25 30 23 54 5 1o 5 o o 5 o 5 5 1o 5 5 o o - 5 5 24 5‘ 5 5 5 o o 5 -1o 5 o 5 o - 5 - 5 5 o 5 25 5‘ o 5 - 5 o - 5 o 5 o 0 10 o o -15 o 5 10 26 5‘ 1o 15 5 0 -1o 5 5 20 5 5 2o 5 -1o - 5 5 5 27 14 10 1o 5 o - 5 o o o 5 5 1o 5 - 5 o 5 5 28 5? 1o 10 o o o 5 1o 25 1o 10 o 5 o 5 1o 25 29 5‘ 1o 10 5 1o 10 5 10 1o 5 5 10 1o 5 5 20 1o 30 5‘ 5 10 1o 5 - 5 5 o 5 10 1o 10 1o 5 20 25 1o 98 Appendix B. - - Continued. 00 055505055005505055555005055 NEOOOfi 21 1 .1. 1 11... 12. 11 112 05 000000500500505500005505055 NmOOOm 1 . 1. 1 . 1 1 r 05 505005555050500050505500005 a unooom 2 1 1 21 1 11 1 1 e . ... . .... ._ ... . fl 8 00 500005555550000550005005000, L NmoooH l 2.1 ..1 1 . 1.. _ 11. 3 . . ....m 55 555505055005055055550555005 5m chom 1 1. 11. 1 11“. 111..1 S e . . r 00 000550550005005050050500050. um Nmomm 1 1 1 1 1 1... 1 n 0 fi . c 00 500000555505505000055005050 .W NHHOOOV 1 . 11 11 112. .11 11 n m . , 55 500050050555005550005005055 m Nmooom .. .141... 11.1. 111 2 1 O B r a 55 000005555055050000500550050 e swooom 1 11111 21 1 11 1 1 ............. .. ... t .m. ..1 00 550000050000000005555050505 R NmOOOH J. .. 1141 1.. 111 1. . .1 55 00.5055500050500050550000050 Nmoom 1 .11 1 11 1.11“. 1... .55 000500050005005055050555050 NmomN . .I.. 11 1 1... 1 wommnfim 12 456789012345678901234567890 111111111122222222223 Appendix B. - - Continued. 99 Z“ Tone decay SISI 0 +4 E 3. ‘8 g m E .223 G n 8'3 .._, o 0) N N N N N o 3? 8 5 ~13 3 43 '3 N m :I: a: I: It: ‘3» '5‘ 9 e 58 .. :5 5 o o o o o "Q as s. :3 f“ 66 8 8 8 8 8 8 c?) In 0-: S Q '8 <5 ‘8 to H v H N v 1 - 245 229 15 10 0 0 45 25 40 2 + 256 79 177 0 10 5 65 15 0 3 + 161 88 73 O 5 5 10 85 20 4 + 274 139 135 0 0 10 55 30 0 5 - 146 72 74 0 5 0 0 0 45 6 + 314 33 281 5 0 0 0 5 15 7 + 386 126 260 5 5 5 60 90 45 8 - 375 157 217 5 5 10 0 0 0 9 + 418 223 195 5 0 10 15 45 15 10 + 344 86 258 0 5 5 5 0 0 11 + 414 268 146 0 5 10 5 0 10 12 + 328 164 164 5 10 10 0 0 0 13 - 286 134 152 10 5 5 10 0 0 14 + 173 50 123 10 0 0 0 50 25 15 + 348 120 228 10 5 5 10 0 20 16 + 162 23 139 5 0 0 10 5 0 17 + 289 84 205 15 0 10 0 0 0 18 + 242 98 144 5 0 0 55 5 65 19 + 235 120 115 15 20 15 25 50 5 20 + 233 131 102 0 10 10 20 0 15 21 + 249 47 203 10 15 10 85 60 70 22 + 550 444 106 15 5 10 0 0 10 23 - 226 45 181 0 0 5 10 5 0 24 ? 248 35 213 5 10 0 5 0 0 25 + 248 198 50 5 5 0 5 0 0 26 ? 260 116 144 5 5 0 15 45 40 27 - 241 123 118 0 0 0 O 0 0 28 + 320 192 128 0 0 10 0 0 10 29 + 436 137 299 5 5 0 0 0 10 30 + 441 338 103 5 5 10 0 5 0 1 00 Appendix B. - - Continued. 63339 5 61 146 96 5984 owmfioov.....om.0fi..0.... . .. 681308088342461 1 38 1185 Nmooov 1 . . 1 22 11 450506 735 7513 4. 7 1 622 oomfioom.....2...2....... .2. 22 286308089240483 14 46 353396 NHHOOCV 2 . . 11.. 22 11.. 11.. 06380 866253971214313606 30 740609 N 411563539263137299 5921964541 m chcov 2111 1 1111 11 211 2211121121 0 . .... . f .m 8.58.. 322757528525544222. .7.o.o.2...o....o.1.... . 81632401412019.7477 2619605356 B NMOOON 1 1. 1 —. _ .._._ ._ _ . 1 56 1 11 5 ommfioom A7.78.82.7. .9.5.7.90.4H5.2.7.5. . . . .8. .9. . .7. . 394343016109106797 328425354. NMOOON 1 1 _. . _ __ ._ _ . 00mgom 674030141745639925 8707 890 NMOOON 242533358561001113 50.3.3.L022. 111 11 1 1. 11 . 1 11 n Oghmmxmm IIIIIIIIIIIIIIIIIIII .11 .I. O h 800086608868068280000080086808 m Ammvov+ 900099909999099990000090099909 .1 111 1. 1.. 11111 11 1. .m 4 mH+ 8202406486266688684460.80008886 0 mm“. 677878776877776777767788787777 s .1 D 8953784225426204042225 10 03 9mm , 1.1 . 1 wommnsm 123456789012345678901234567890 111111111122222222223 u3111(sungmm1114111112111le1ijm