The aim of pediatricians and parents is to guide children with chronic illness safely through childhood and adolescence to become healthy, young adults who achieve their maximum potential.1 Where diabetes is concerned, the long- and shortterm goals may conflict. The doctor, looking into the future, will have as his main concern that the child should achieve near-perfect glycemic control to avert blindness, renal failure, and the 20- to 35-fold increased relative mortality in young adults.2 In contrast, parents will wish to avoid the demoralizing hypoglycemia and may find it difficult to resolve the conflict between the discipline demanded by diabetes and their desire to encourage normal living and learning. The child must come to terms with something that never goes away, is difficult to conceptualize, does not distinguish you outwardly from your fellows, and takes constant vigilance to balance between hyper- and hypoglycemia.
Few children and even fewer adolescents achieve satisfactory diabetic control. There are many potential explanations (Table) and most doctors concentrate first on technical factors with which we are familiar. However, it is the diesis of this paper that the most usual explanation of poor control is normal life stress or a family disturbance. The doctor prescribes treatment but, it is the parents and child who carry it out from day to day, monitor its effects, and make minor adjustments as required. Blood glucose control is affected by anything that distracts attention from this process. Hence, the diabetologist needs to understand normal and abnormal responses of young people to the constraints imposed by diabetes, as well as normal and abnormal family dynamics and how children and adolescents respond to stress.
IMPACT OF DIAGNOSIS
People enter parenthood with high expectations and the unheralded discovery that the child has diabetes is a shock, loss, and disappointment. Initially, most parents react with a combination of fear, guilt, sadness, anger, and denial.
In the days and weeks after diagnosis, parents have much to learn about a strange subject. Reading about die possible blindness or kidney failure increases the parents' fears. Above all, they worry that they will be unable to cope. Searching for a cause for their misfortune, many feel guilty at having transmitted a tainted heredity to their child or having in some other way caused diabetes. Exaggerated importance may be given to events preceding the onset; parents may remain irrationally convinced that "It was the shock of her granny's death which gave her diabetes" or "If only I hadn't had so many rows with her, this would never have happened." Most parents mourn the loss of their "normal" child which often includes anger displaced toward the child or medical staff in the form of complaints about delay in diagnosis and errors in treatment. A stage of denial is common; hopes may rest on the possibility of spontaneous recovery which is often cruelly reinforced by the false dawn of the honeymoon period. Alternatively, they may believe in the imminence of a scientific breakthrough as promised by premature and over-optimistic reports in the press about trials of cyclosporin or other agents.
POSSIBLE CAUSES OF POOR DIABETIC CONTROL
There are conflicting reports about the degree and duration of psychological distress in parents caused by the development of diabetes in their child. In one study of 74 consecutive families in Pittsburgh3 the most striking finding was "the relative ease with which parents weathered this life stress. " One in four mothers mourned - typically feeling sad, tired, or worried and crying and being irritable. However, severe anxiety was not common and the mild depression resolved within six months. Fathers were even less affected. This well-conducted study paints a more optimistic picture than many, of the mental health of parents of diabetic children and has strengths and weaknesses which should be considered when assessing similar studies:
* A strength was that the emotional wellbeing of parents was assessed with standardized symptom scales which focused on how the parents fek in themselves, not on how they felt about the child and the illness. However, it is possible that questionnaires cannot bypass even the most superficial psychological defenses.
* The families were predominantly intact, white, and middle-class which may have helped them to withstand a stressful experience.
* The study concentrated on only the first year after diagnosis when management is relatively easy. Maintaining good control is more difficult after three to five years when endogenous insulin secretion has been lost.
Parents may survive the first year with ease but there is no doubt that looking after the child in the long term imposes demands and discipline which strain all family members. Anxiety may remain manageable until the child's first serious hypoglycemic reaction, a situation vividly described by David and Penny Thomas.4 A bad reaction may cause the parents to take turns getting up through the night to check on the child or, in extreme cases, mother may move into the child's bed. With time, previously latent marital problems may be exposed by the strain. Husbands may feel jealous and angry with their wives who are absorbed in the ill child and one hears comments such as "If we go out, she can't relax and spends all the time talking about diabetes or wondering of the child is okay."
It is also widely believed that nondiabetic siblings suffer by being "handicapped" in the race for growth, attention, and affection and may defend themselves by becoming demanding or taking it out on the diabetic sibling. Siblings of diabetic children have been reported by their mothers to complain of headaches, stomach cramps, and other somatic complaints four times as often as siblings of healthy children.5 Mothers report that nondiabetic siblings may ask for an insulin injection or mimic the behavior of insulin shock to get attention.
Maladaptive parental (especially maternal) attitudes have been described frequently as determinants of the child's adaptation to diabetes. Some stereotypes and their effect on the child's reaction include:
* Overprotectiveness. The parents' wish to protect a disadvantaged child is so normal that it may be difficult to hold in check and may progress to the point where the child is overwhelmed. If this happens, the child may demand more and more attention, resign from doing even the simplest things for himself and end up by dominating family life.
* Domination. These parents are often perfectionists who hold the child to a strict schedule and expect Spartan bravery with little reward. ' The end result is usually a crushed or rebellious child and a doctor who dreads the appearance of the parent who gives him a rough time, questioning most aspects of treatment and shifting blame for the child's misdeeds (diabetic and otherwise) to the doctor.
* Self-pity. This stereotypic mother, often from a one-parent family, is preoccupied with her own problems and draws attention and sympathy to herself by recounting the difficulties of having a diabetic child. Often the child reacts by becoming "difficult" to gain more attention which merely reinforces the mother's martyrdom.
* Rejection. There are hints that the child with diabetes (or other chronic illness) is more likely to be abused either physically or psychologically.6 This abuse may be overt in undisciplined families with parents who lack basic skills in handling their children. Alternatively, some parents may manifest a "sustained negative affect" by continually criticizing the child's self-management, making him or her feel guilty because the diabetes is poorly controlled, and not praising the child in any way. Cases of "brittleness" have been described where the mother was deliberately under- or overdosing the child with insulin.
It should be stressed that these are extreme examples of maladaptation and most parents (especially those described as tolerant, consistent, and flexible) have children who are well adjusted and in satisfactory, if not perfect, metabolic control. The child is part of a dynamic unit, the family, and any study which regards the child as a passive entity acted on by the environment tells only part of the story.7 Correlation between the child's metabolic control and individual psychological variables, such as anxiety in the mother and low self-esteem in the child, are relatively poor; whereas investigations which concentrate on the family as a whole generally show a strong relationship between poor control in the child and defective family integration and functioning. Thus, it has generally been found that the best controlled children come from families with few conflicts, a low level of stress in the parent-diabetic child relationship, satisfactory home adjustment by the child, and no money worries. A longitudinal investigation by Koski8 found the best control where there was good cooperation between all members of the family and shared responsibility for seeing that the child followed the regimen. At the five year follow-up, Koski found that the families of eight children whose control had deteriorated were characterized by increased family conflict, increased illness in general, or loss of a family member. Where the child's control had improved, family relationships were healthy and the families less stressed than previously. Koski concluded that "good diabetic control is but one aspect of a healthy, well-functioning family."
How the Child Feels
In the first year after diagnosis Kovacs et al9 examined children's views of themselves and their strategies for coping. This was a predominantly middle-class cohort from intact families in which self-assessment by children aged 8 to 13 years suggested they were just as confident and emotionally at ease as normative groups with fewer symptoms of depression and a similar degree of self-esteem. Most had some feelings of sadness and wishful thinking ("If I do everything right, maybe it will go away" ) shortly after diagnosis but few blamed themselves or worried unduly about the future. One interesting finding was that noncompliance with blood and urine tests was already evident in the first few weeks and became common by the end of the first year. There have been many cross-sectional surveys of the emotional health of diabetic children later in the course of the disease. Overall, where objective measures and unbiased designs are used, there is no substantial difference between diabetic children and controls. Some have suggested that diabetic children suffer more anxiety, inadequate self-image, lower selfesteem, increased parental dependence, less satisfactory adjustment at home, and worse academic performance. However, children are remarkably resilient and in most children in most studies, the degree of emotional upset actually found is relatively small.
Many potential (and predictable) problems depend on the age at which the child develops diabetes. It is helpful to focus separately on the preschool years, primary school, and adolescence.
There have been few systematic surveys on the problems of managing diabetes in very young children or how it affects their emotional development. Parents necessarily shoulder total responsibility for care and are often overwhelmed at first and worried that the baby is going to die. Bonding with the mother may be upset because the child interprets injections, blood tests, and denial of certain tasty foods as punishment. Young children attribute accidents or illness to recent family events; their understanding of the cause of diabetes may be "I stole some candy and got diabetes," or "I shouted at daddy and he gave me an injection; I don't shout at him any more and I still have the injections."
The stress on parents is often great. They may find it difficult to distinguish hypoglycemic reactions from temper tantrums, may be loath to leave the child in the care of others, and may come to dread mealtimes in case, after receiving an injection, the child refuses to eat.
Primary School Years
By this age children can think through problems and have increasing motor skills. Thus, they may be able to inject themselves, understand the rudiments of their diet, and assist with blood or urine testing. It may please a child to "help mommy" in these areas of diabetes management but it is important for the parent to be realistic and not expect the child to do more than is appropriate for his or her age and developmental stage.
At this age children often begin to realize that they are different from their peers and may feel damaged. This is obvious if one looks at their drawings where they may represent themselves (and others such as the doctor) in fantastic guises.
With or without diabetes, adolescence is often turbulent.10 The extra difficulties facing the diabetic teenager can only be understood in the context of normal adolescence. The emphasis is on changes in physique, sexual development, thoughts, and selfimage. Diabetes interferes at every stage.
The growth spurt and sexual maturation require rapid changes of self-perception to cope with the changing body image. Normal adolescents feel physically awkward but at the same time have a narcissistic interest in their own bodies and need to contemplate changes in privacy. The diabetic teenager may be subjected to frequent physical examinations which, although well-intentioned, are seen as an invasion of privacy and modesty. Delayed sexual maturation, common with poor diabetic control, may lead to a sense of inferiority or reinforce previous feelings of being defective.
The adolescent strives for independence but diabetes may make it more difficult for mothers to relinquish their hold. The adolescent has an absolute need to be like and be liked by his or her peer group. Diabetes makes this hard by enforcing regular mealtimes, forbidding junk foods, and carrying with it the everpresent risk of hypoglycemia which will give the teenager unwanted prominence and expose him as being an unreliable and alien being. Parents worry that their adolescent offspring are not taking their diabetes seriously, doing the requisite number of tests or carrying glucose. However, they may be reluctant to voice these concerns for fear of being accused of nagging. Friction between teenagers and their parents is normal, but the difficulty with diabetes is in restricting the conflict to matters such as hair length, fashion, and overloud music rather than allowing it to spill over into arguments about urine or blood testing.
It is difficult to draw firm conclusions about the mental health of diabetic adolescents and many studies ignore the background of "normal" adolescence to exaggerate the problems of the diabetic. It is worth noting that large surveys of normal teenagers have shown that up to half have isolated neurotic symptoms and transient feelings of depression. Around 10% of "normal" 14 to 15 year olds have a definable psychiatric disorder, but more admit unexpressed feelings of misery and self-doubt. It is probably inevitable that teenagers with a chronic disorder will be more miserable and have a deeper feeling of inferiority although most develop strong defense mechanisms to preserve their personality. Any relative weakness of earlier psychological development will be highlighted by adolescence. Patients who had the most trouble with diabetes are usually those who have shown signs of personal difficulties earlier. It is also certain that the normal turbulence of adolescence will be expressed through upsets in diabetic control. This is particularly noticeable in girls who in their teenage years have higher mean glycosylated hemoglobin levels, higher hospital admission rates, and provide diabetic clinics with most of their "problem" patients. This pattern is not confined to diabetes but is seen with asthma and overdoses. The evidence is overwhelming that the fragile health record of adolescent girls is predominantly of psychosocial origin.11
PREVENTING PSYCHOLOGICAL PROBLEMS
Much could be done if the average diabetologist was trained to be aware of potential problems and recognize them early rather than after months or years of fruitless juggling with insulin doses, changing to the newest and most highly purified insulin formulation, prescribing yet more sessions of education and dietary instruction, etc. A few suggestions for preventing psychosocial problems in diabetes patients are listed below.
* At diagnosis the physician must strike a balance between being realistic and avoiding unnecessary anxiety. A good start is to find out what the parents know already, ask whether they have relatives or friends with diabetes and, if so, what happened to them. It is surprising how often one gets the answer "Yes, my uncle was a diabetic who went blind and lost both legs." This tells you what their fears are and indicates the need for a series of gentle but frank discussions.
* A member of the team must get as much background information as possible about the family and how it works. When first seen, families often give the impression of warm affectionate relationships and stability, whereas a more detailed inquiry may reveal preexisting strains and potential problems. If these are foreseen, it makes them easier to deal with in the future.
* The team dealing with a particular child should be as small as possible so that consistent information and advice is given.
* Continuing support should be provided which means more than just asking "How are you ?" or "Are you managing the injections O.K. ?" In the first few months, all the parents and child may want is a sympathetic nurse to whom they can unburden themselves. Later they may need an opportunity to talk about life problems rather than insulin doses.
Cheating and Noncompliance
Failure to follow the doctor's advice to the letter is a common human characteristic and well known in diseases such as hypertension with long-term treatment goals. Factors known to encourage compliance include11:
1 . that the patient appreciates the seriousness of the illness and believes that it can be treated effectively;
2. a reward for compliance or a disincentive for noncompliance;
3. an easy regimen which requires little behavioral change;
4. close supervision and positive feedback from doctor, nurses, and parents;
5. a supportive and stable family background and good doctor-patient relationship.
The reasons for noncompliance vary with developmental stage. Young children do not understand the seriousness of diabetes and as part of their learning process want to test the limits. Adolescents do not think about their own mortality and, even if they did, the promise of freedom from complications 40 years hence carries little weight in relation to the major short-term crises provoked by hypoglycemia. The diabetic regimen is so demanding and frustrating that many young people simply give up the unequal struggle.
Most youngsters do not non-comply out of sheer devilry, but more often resort to "tricks of the trade" to deceive parents and physician, thereby avoiding disapproval and displeasure. Particularly common is the faking of urine or blood tests. This may simply involve writing down spurious results or may be a more complex deception such as presenting a nondiabetic sibling's urine to their mother for testing. Before castigating children for these misdemeanors it is worth reminding oneself of a study in which surreptitiously modified memory meters showed that of 19 adults, three quarters reported fictitiously low values at least once, 10% omitted readings from their log book, and 40% added imaginary values.12
Using illness to one's advantage is another common trick pulled by diabetic children to avoid school or obtain some other secondary gain. Most children try their parents from time to time with imaginary or exaggerated symptoms such as head or stomach ache. The parent of the well child can challenge this by sending the child to school in the knowledge that sequelae are improbable. In contrast, the mother of the diabetic child, especially one with a history of ketoacidosis, may keep the child at home to be on the safe side and hence set in motion a long sequence of school absences.
Children may also discover by experience that a complaint of hypoglycemia will be uncritically accepted and "rewarded" with candy. Epidemics of pseudohypoglycemia occur regularly at diabetic children's camps and in one study only 56% of reported episodes were confirmed by a low blood glucose value. Most parents would probably have a similar striking rate although one must be careful not to ascribe all discrepancies to manipulative behavior. Children often mistake the symptoms of anxiety or tiredness for those of hypoglycemia and their feelings may lead to an honest error of judgment.
When children are found to be cheating, the way they are challenged is important. The underlying mechanism must be investigated by means of one or more careful interviews emphasizing self-understanding rather than confession and judgment.13
Many attempts have been made to define brittle diabetes in terms of blood glucose swings or other physiological characteristics such as the presence or absence of endogenous insulin secretion, but I think it is more useful to reserve the term for a small but conspicuous minority of patients whose lives are "constantly disrupted by episodes of hypo- of hyperglycemia whatever the cause."14
Hypoglycemia is the commonest complication of insulin treatment, thus it is difficult to decide how frequently it should be permitted before being labelled brittle diabetes. Because there is at least one hypoglycemic coma in 25% of stable insulin-dependent diabetics per year, two or more severe hypoglycemic attacks per year would be abnormal and require investigation. Frequent causes include missed meals, loss of warning symptoms, unaccustomed exercise, etc, which should be curable by simple maneuvers. However, some attacks are nonacc idental.
Hypoglycemia or the threat of it is a devastating weapon with which to manipulate one's environment. Situations in which a nondiabetic child might provoke abdominal pain or a half-hearted overdose may be dealt with by playing the trump card of self- induced hypoglycemia. The diabetic adolescent can compel attention or escape distress by not eating - a form of blackmail ignored by those at whom it is directed only at the risk of terrifying unconsciousness in the perpetrator. I have reported several cases of this type, and Or et al15 described six adolescents found to be secretly taking extra insulin either with suicidal intent or as a form of symptom substitution in an escalation of self-destructive behavior. The underlying psychopathology varied but depression and severe personality problems were common.
Recurrent ketoacidosis (DKA) is a commoner manifestation of brittle diabetes. These patients are frequently overweight adolescent girls who take relatively large doses of insulin and have enormous bundles of case notes. Some have had 50 or more admissions which in the early days are attributed to intercurrent infection and later to more arcane conditions such as insulin resistance or erratic insulin absorption. It is generally accepted that the main (and maybe the only) cause of recurrent DKA is a deliberate act or omission by the patient.11,16,17 The reasons why patients repeatedly allow themselves to go into DKA are varied, but ultimately it enables the patient to temporarily escape from the family or some other problem. The ability of the diabetic child to use DKA as a refuge from unpleasantness at home or school was noted in 1944 by Loughlin and Mosenthal,18 a third of whose children with recurrent DKA came from homes broken by divorce, separation, or widowhood, or homes where the mother was out most of the day and meals were haphazard. Most of these children frankly stated that they liked the hospital better than their homes.
Management of brittle diabetes is beyond the scope of this review, but I submit that the essential first step is to be aware that the cause probably lies in the child's life situation rather than in the pharmacodynamics of insulin.
Sticking to a rigid eating plan (ie, controlling one's food intake) is the sine qua non of diabetes treatment and weight gain the almost inevitable consequence of intensive insulin therapy. Therefore anorexia nervosa and other eating disorders are common. Steel et al19 found a 7.2% prevalence of eating disorders among diabetic women aged 16 to 25 years; this was probably an underestimate because it depended on clinical impression. Wing et al20 found that young diabetic patients had more eating pathology than their nondiabetic peers. Self-reported bulimic behavior was significantly associated with decreased glycémie control.20 There have been conflicting reports about the effect of anorexia nervosa on diabetic control. Some early reports were of young women who rigidly controlled both their insulin and food intake so as to lose weight and maintain normoglycemia. However, in the later study from the same clinic19 poor control and acute polyneuropathy were common in those with both bulimia and anorexia nervosa. 1 have also seen several adolescent girls in whom it eventually became apparent that the cause of their poor glycémie control was deliberate underdosing with insulin or omission of injections. Such behavior may be relatively common. It is worth noting that anorexia nervosa and brittle diabetes have in common a female preponderance and the fact that both may be undiagnosable without a history from a family member.
For the patient the long-term treatment of diabetes is onerous, often boring, and may generate much anxiety. Psychosocial problems are part and parcel of the course of diabetes and clinicians need to be on the lookout for them. Many of these problems may seem trivial to the doctor, but unless tackled early may escalate until they dominate the life of the child and the family. If this happens the diabetic child, or more commonly adolescent, may deploy its most powerful weapon - the taking of excessive or inadequate quantities of insulin to produce either hypoglycemia or DKA which results in admission to a safe haven staffed by health care professionals who do not ask too many awkward questions.
1. Hughes J: Trie emotional impact of chronic disease: The paediatrician's responsibilities. Am J Dis ChM 1976; 130:1199-1203.
2. Oreen A, Borch- Johnsen K, Andersen PK. et al: Relative mortality of type I (insulindependent) diabetes in Denmark: 1933-1981. Diabetofagia 198$; 28:339-342.
3. Kovacs M. Rnkelstein R. Feinberg TX. et al: Initial psychological responses of parents to the diagnosis of insulin-dependent diabetes mellitus in their children. Diabetes Care 1985; 8:568-575.
4- Thomas O, Thomas P: Living with a diabetic child, in Baum JD. Kinmonth AL (eds): Care of the Child with Diabetes. New York. Churchill Livingstone. 1985, pp 3-11.
5. Ferrari M: Chronic illness: Psychosocial effects on siblings - L Chronically ill boys. I ChM Psychol Psychiatry 1984; 25:459-476.
6. Horan P, Gwyn F, Renzi D: Insulin dependent diabetes mellitus and child abuse: Is there a relationship? Diabetes Care 1986: 9:302-307.
7. Tattersall RB: Diabetes: The young person, their family and the doctor- International Diabetes Federation Bufletrn 1983; 27:1-5.
8. Koski ML, Kumento A: The interrelationship between diabetic control and family life, tediarne and Adolescent Endocrinology 1977; 3:41-45.
9. Kovacs M, Brent D, Steinberg TF. et al: Children's self-reports of psychologic adjustment and coping strategies during first year of insulin-dependent diabetes mellitus. Diabetes Care 1986; 9:472-479.
10. Tattersall RB, Lowe J: Diabetes in adolescence. Diobetologia 1981; 20:517-523.
11. TattersaURB. WalfordS: Brittle diabetes, a spectrum ofillness in response to life stress: The place of cheating and manipulation, in Pickup JC (ed): Brittle Diabetes. Oxford, Blackwell. 1985, pp 76-102.
12. Mazze RS. Shamoon H, rVsmantier R, et al: Reliability of home blood glucose monitoring by patients with diabetes mellitus. Am J Med 1984; 77:211-215
13. Citrin W, Ellis G), Skyler JS: Glycosylaredhcnugkibin: Tool in iaenblymg psychological problems. Diabetes Coze 1980; 3:563-564.
14. Tattersall RB: Brittle diabetes. Om Endocrinol Metab 1977: 6:403-419.
15. Ott DP, Ecctes T. Lawlor R, et al: Surreptitious insulin administration in adolescents wrrh insulin dependent diabetes mellitus. JAMA 1966; 256:3227-3230.
16. Schade DS. DrummDA, Duckworth WC, et al: The etiology of incapacitating bnrrle diabetes. Diabetes Care 1985; 8:12-20.
17. White K, Kolman ML, Vfcxler P. et al: Unstable diabetes and unstable families: A psychological evaluación of diabetic children with recurrent ketoacidosis. Atienes 1984; 73:749-755.
18 Loughlin WC Mosemhal HO: Study of the personalities of children with diabetes. Am J Db CMd 1944; 68:13-15.
19 Steel JM. Young RJ, Lloyd CS, et al: Clinically apparent eating disorders in young diabeoc watneiu Associations with painful neuropathy and other corapUcanoos- Br Med; 1987; 294:859-862.
20. Wing RR, Norwalk MP, Marcus MD, et al: Subclinical eatmg disorders and gfycenuc control in adolescents wich type 1 diabetes. Diabetes Care 1986: 9.162-167.
POSSIBLE CAUSES OF POOR DIABETIC CONTROL