It is accepted wisdom that childhood diabetes is rare in India. Most published series quote a prevalence of less than five percent of all persons with diabetes (1). Majority of the reports are based on hospital and clinic based data:
Gender differences in young/ childhood diabetes
| Area | Diagnosis | Age group | Male | : | Female | Ref |
|---|---|---|---|---|---|---|
| Orissa, Cuttack | PDDM* | <30 y | 2.7 | : | 1 | 2 |
| Orissa, Cuttack | FCPD** | <30 y | 3 | : | 1 | 2 |
| Orissa, Cuttack | IDDM | <30 y | 2 | : | 1 | 3 |
| New Delhi, AIIMS | IRD*** | <30 | 0.64 | : | 1 | 1 |
| New Delhi, AIIMS | KP**** | <30 | 0.45 | : | 1 | 1 |
| New Delhi, AIIMS | KR@ | <30 | 0.67 | : | 1 | 1 |
| New Delhi, AIIMS | NIRD@@ | <30 | 0.69 | : | 1 | 1 |
| New Delhi, AIIMS | CCP@@@ | <30 | 0.6 | : | 1 | 1 |
| Kerala, Cochin | IDDM | <20 y | 20 | : | 19 | 4 |
| Kerala, Cochin | NIDDM | <20 y | 3 | : | 5 | 4 |
| Kerala, Cochin | FCPD | <20 y | 2 | : | 9 | 4 |
| T Nadu, Chennai | All groups | <20 y | 3 | : | 3 | 5 |
| T Nadu, Chennai | FCPD | <20 y | 17 | : | 11 | 6 |
| T Nadu, Chennai | NIDDY | <30 y | 188 | : | 126 | 7 |
| T Nadu, Chennai | IDDM | <30 y | 73 | : | 46 | 7 |
| T Nadu, Chennai | FPD | <30 y | 17 | : | 4 | 7 |
| T Nadu, Chennai | PDPD | <30 y | 6 | : | 1 | 7 |
| Andhra P, Vizag | IDD,MRDM | <20 y | 2 | : | 8 | 8 |
| T Nadu,Kudremukh | <25 y | 0 | : | 0 | 9 | |
| Fiji | <25 y | 0 | : | 0 | 10 | |
| Tanzania | <25 y | 0 | : | 0 | 10 | |
| Africa | MRDM | M > F | 12 | |||
| Africa, Uganda | PDDM | 16 | : | 6 | 13 | |
| Africa, Ethiopia | <15 Y | 32 | : | 48 | 14 | |
| Malta | <32 y | 52 | : | 65 | 15 | |
| Africa, Sudan | IDDM | 7-14 Y | More in girls | 16 | ||
| United Kingdom | IDDM | <14 y | 26.8% more in males | 17 | ||
| United Kingdom | IDDM | <30 y | 65 | : | 45 | 18 |
| Europe, Norway | IDDM | <14 y | 12% more in boys | 19 | ||
* protein deficient diabetes mellitus
**fibro calculous pancreatic diabetes
***insulin requiring diabetes
****ketosis prone
@ketosis resistant
@@non insulin requiring diabetes
@@@chronic calcific pancreatitis, with diabetes
Not unexpectedly, a wide range of age groups, diagnostic criteria and differences in gender were reported. Diabetes was more prevalent in males, with few exceptions in New Delhi (1), Kerala (4), Andhra Pradesh (8) Ethiopia (14), Maltese islands (15) and Sudan (16).
Data from our Centre, on children with diabetes starting at or below the age of 15 years is given in Table 2. The male female ratio continued to favour females (ie, more females), consistent with our earlier publication ((8).
Childhood diabetes: EDC, Visakhapatnam (1988-1996)
| Characteristic | Boys (n:24) | Girls (n:34) | P* |
|---|---|---|---|
| Sex ratio(M:F) | 24 | 34 | |
| Onset of diabetes(years) | 9.62+/- 3.82 | 8.5+/- 3.99 | NS |
| Duration of diabetes (years) | 2.58+/- 3.45 | 3.38+/- 5.7 | NS |
| Newly diagnosed DM | 5(20.83%) | 6(17.65%) | NS |
| Duration DM=<6 mo. | 19(79.17%) | 19(55.88%) | NS |
| | |||
| Geographic origin | |||
| Andhra Pradesh: rural | 5(20.84%) | 6(17.65%) | NS |
| Andhra Pradesh: urban | 16(66.67%) | 27(79.41%) | NS |
| Other states | 3(12.5%) | 1(2.94%) | NS |
| | |||
| Height (cm) | 125.85+/- 43.68 (n: 20) | 122.3+/- 23.28 (n:20) | NS |
| Weight (kg) | 28.93+/- 13.85 (n: 23) | 26.62+/- 12.43 (n:32) | NS |
| Body mass index | 15.5+/- 2.65 (n:20) | 15.12+/- 2.42 (n: 20) | NS |
*NS: not significant
There was no statistically significant difference in the various parameters studied, although there was a trend towards a greater proportion of urban boys from rural areas, and boys from other states being brought for attention (Table 2). Similarly, there was a a greater proportion of boys with duration of diabetes less than six months.
Community based prevalence studies in India:There are few community based prevalence studies of childhood diabetes in India, that employed well established criteria for diagnosis. Ramachandran at al estimated the prevalence of childhood diabetes in an urban population in South India (the city presently known as Chennai). The overall prevalence was 0.26/1000. Among 30 IDDM patients with onset of diabetes below 15 so identified, there were 13 boys and 17 girls 20).
The other major prevalence survey in the community was conducted all over India by Dr. P.V. Rao, under the supervision of Prof. M.M.S. Ahuja. The prevalence of diabetes mellitus in the age group 15-19 years was 0.34% in males (n:3) and 0.19% among both males and females (3 boys, no girls). (21).
Is childhood diabetes truly common in boys ?
Most studies show that boys are more prone to develop diabetes mellitus compared
to girls. The difference exists in both community based surveys as well as
hospital based reports. Given the interaction of genetics, environment and the
immune system in the pathogenesis of IDDM, the male preponderance has been well
established (17). Exposure to possible environmental toxins, such as N-nitroso
compounds has been instigated as partly contributing to this.
However, a closer examination of children who presented below the age of five
from United Kingdom shows that more girls than boys developed diabetes in this
age group (18). Perhaps a difference in the genetic susceptibility and
environmental insult could explain the difference in gender difference in the
very young.
Age at presentation: gender differences in relation to puberty
An interesting sex difference in the age at presentation was shown by
Ramachandran et al (22). The peak age of diagnosis was 11 years in girls (n:293),
while boys (n:321) showed multiple peaks between 11 and 18 years. In urban
patients the maximum occurrence of the disease was at 11 years in contrast to a
delayed peak at 18 years in the rural group. This demonstrates that the peak
was delayed in children from the rural areas and more markedly in those from the
lower income group.
The peak occurrence of IDDM at puberty could partly result from high
susceptibility at that time (23). It is probably caused by decreased insulin
sensitivity at that age, which also explains why girls develop diabetes early,
as they undergo pubertal development earlier than boys (24).
Islet cell antibody (ICA) positivity, a surrogate of smouldering autoimmune
pancreatic beta cell destruction was seen in girls at a younger age, suggesting
that beta cell destruction is faster, and total destruction occurs earlier in
girls (25). On the other hand, ICA positivity in boys with longer duration of
diabetes might be associated with slow progression of diabetes mellitus (26).
The precise reason for the difference is not known, although influence of sex
hormones can be one factor responsible for diabetes occuring in girls at a
younger age (25). Estrogens and progesterone, the principal sex hormones in
girls which rapidly increase during puberty, are known to decrease insulin
levels, and decrease glucose tolerance (27). However they do not appear to
produce clinical glucose intolerance.
When autoimmune diseases are common in women, why isn't childhood diabetes so ?
It is well recognised that autoimmune diseases are far more common in women than
in men (28). The logical cause for this difference would be the sex hormones.
Women might be supposed to respond more to conventional antigens, due to sex
hormones (28). The interaction of gonadal steroids and the immune system is
another area for further study. Genetic control of autoantigen presentation to T
lymphocytes by specialised cells might be affected by sex hormones (29,30).
The other reason why childhood diabetes is not uniformly more common in girls is,
it is not all autoimmune. Contributions from genetic, environmental and
immunological factors ensure that. Besides, a variety of clinical syndromes are
seen in young diabetics, especially from India and other developing countries.
Whereas childhood onset diabetes below 15 years of age is more often classical
type 1 autoimmune mediated, younth onset, ie 16-30 years appears to be more
heterogenous (31).
Studies on residual beta cell function in young diabetics showed a spectrum of
changes, from insulin requiring diabetics (n:77) who were ketosis prone (40%) to
ketosis resistant insulin requiring (60%) patients; twenty three were non
-insulin requiring to those with calcification of the pancreas (32), the last
category included in the generic term of malnutrition related diabetes mellitus
(33).
Clinicians are aware of unusual variants, including those simulating non-insulin
dependent diabetes mellitus (34). A variety of types including insulin dependent
diabetes mellitus, malnutrition related diabetes mellitus, non insulin dependent
diabetes of young, maturity onset diabetes of young (34a) are all described,
which could account for the skewed gender prevalence of childhood diabetes.
Why does diabetes favour boys ?
Given that IDDM is autoimmune mediated, why is it that girls are relatively
spared, at least when compared to other autoimmune diseases such as rheumatoid
artiritis ? Delayed diagnoses, missed diagnoses, not being brought to medical
attention, all as part of social deprivation (35) are obvious reasons. But could
there be something else ? Some of the other reasons, particularly in our country
can be speculated upon.
The pathogenesis of IDDM can be summarised as follows: (a) IDDM is a polygenic
disorder, in which genetic susceptibility is conferred by an unfavourable
combination of common alleles of normal genes (b) each gene involved in IDDM
susceptibility quantitatively controls a part of the pathogenetic process, which
is initiated by (c) release of islet beta cell antigen, processed and presented
byMo to Th lymphocytes. There is a self-perpetuating and self-limiting circuit
of cytokine production of which IL-1 is islet beta cell cytotoxic, an effect
potentiated by tumour necrosis factor alfa (36).
In other words, there is a role for environmental agents to initiate the
autoimmune process. Efforts are on to identify the environmental agent. Putative
mediators include dietary toxins (17,33), consumption of caffeine by the mother
during pregnancy (37), nutrients and food additives (38).
Recent interest has focussed on the role of breast-feeding in preventing IDDM. A
Swedish study has shown that breast-feeding could decrease the risk of IDDM in
certain subgroups (39). The reason seems to be early exposure to foreign
proteins, such as cow's milk, which can trigger the autoimmune process of IDDM
(40). Another reason for high risk in non-breast-fed infants could be that
breast-fed infants gain less weight compared with non-breast-fed children, which
may protect against type-1 diabetes later on (41).
The protective effect of breast-feeding seems to be lost in Indian children
differently, depending on the gender of the child. Recent studies have shown
that girls are breast-fed for a shorter length of time (42,43). The
differential was shown to occur across socio-economic disparity, where breast
-feeding continued for a longer period among land-owners than among landless
labourers (44).
Nutritional state of the female child is poor compared to the male (42). Besides,
boys are more likely to be given nutritionally rich food such as milk (43).
Lastly, boys were attended by the doctor on the very first day of illness,
compared to girls. A combination of all these environmental factors on genetic
susceptibility, along with increased mortality of Indian girls in the first five
years of life (44) make more diabetic boys present for attention.
A recent study from Bangladesh also showed that the duration of breast feeding
was decreasing in younger mothers, compared to older women (45).
Associations by themselves do not imply causality; risk factor stratification
has shown that increased risk of childhood IDDM in Northern Ireland and Scotland
was seen with delivery by Cesarean section (46).
Diabetes begins in the womb: Barker's hypothesis
Following a series of painstaking and elegant epidemiological studies, Barker
and Hales proposed that diabetes mellitus in later life, had its genesis from a
combination of undernutrition early in life and overnutrition later on (47).
Peak pancreatic beta cell mass may be determined early in life, even during
gestation and the factors that influence it are likely important in the
development of diabetes (48). Infants with poor intrauterine nutrition were
shown to have fewer beta cells, a finding in line with Barker's hypothesis which
was put forth later. Overnutrition or obesity in later life leads to insulin
resistance, and the functional beta cell mass, programmed in leaner times, may
then be unable to meet the rising demand for insulin (48).
Given that the etiology of IDDM is multifaceted, early undernutrition can
contribute to its pathogenesis, although not as significantly as in adult onset
or type 2 diabetes (NIDDM).
It seems possible therefore that susceptibility to type 1 diabetes could be
determined during gestation or infancy in response to nutrition. the stress that
leads to type 1 diabetes is likely insulitis, or a progressively diminishing
insulin reserve. The outcome in type 1 diabetes can be logically influenced by
the peak beta cell mass, which in turn could depend on early maternal nutrition
(48).
A combination of poor antenatal nutrition, commonly seen in women in developing
countries, combined with relative post natal overfeeding of boys, ... by social
factors, could be the right mix for Barker's hypothesis to take effect in these
boys.
In collaboration with Dr. Barker, Dr. C.S. Yajnik from Pune, has been
undertaking an ambitious project in which women of childbearing age from the
community are screened periodically, through to the time of their delivery and
later. The ongoing project should generate invaluable data that throws light on
early life nutrition as a cause of insulin resistance and insulin deficiency.
An earlier study on infants aged 3.5 to 4.5 years was carried out to see whether
there was a difference in metabolic profile depending on birth weight (49). At
the age of four, girls had smaller head circumference and greater skinfold
thickness than boys. Plasma glucose was not different between boys and girls,
but plasma insulin insulin concentrations in girls were consistently higher,
implying insulin resistance. Pooling the data, the highest 30 minute plasma
glucose concentration were in children who were light at birth and heavy at four
years (49).
Considering that IDDM consists of insulinopenia, with insulin resistance partly
contributing, fetal undernutrition could be partly responsible for diabetes.
Correction of maternal undernutrition in pregnancy, and appropriate nutrition
for both boys and girls are necessary goals.
There is a difference in the microvascular diabetic complication rate and
morality, depending on whether IDDM develops before puberty or not. Studies from
the Children's Hospital of Pittsburgh Insulin-Dependent Diabetes Mellitus
Registry has shown that postpubertal duration of IDDM may be a more accurate
determinant of the development of microvascuar complications and diabetes
-related mortality than total duration (50).
By implication, girls, due to their earlier puberty would be expected to be
exposed to a longer duration of diabetic complication risk, and could have
greater microvascular complications. The same group has shown that girls have a
more prominent association of glycemic control with albumin excretion rate,
systolic blood pressure, presence of microaneurysms, serum triglyceride and
serum cholesterol concentrations during the first five years of IDDM (51).
Similarly, elevation of lipoprotein(a) have been shown to occur during puberty
in IDDM children (52).
Differences were also found in the metabolism of triglyceride-rich subfractions
in IDDM males and females (53). The composition of subfractions 100 to 400 and
20 to 100 liopoproteins was abnormal in IDDM men, due to enrichment in total
cholesterol as shown by higher total cholesterol/ triglyceride, total
cholesterol/protein, and triglyceride/phospholipid ratios both fasting and
postprandially. A similar enrichment in total cholesterol was observed in
diabetic versus normal women following fat ingestion in subfractions greater
than 400 only (53).
Late teenage period is associated with improved insulin sensitivity, caused partly by changes in growth hormone (54).
Given that diabetes mellitus is the most complex of all common metabolic
disorders to manage (55), in terms of physical discomfort, mental strain and
economic factors, it is not surprising there could be differences in the way
boys and girls have access to modern management.
Among 614 IDDM patients with onset of diabetes at or before 20 years of age
registered at the Diabetes Research Centre, 234 belonged to the high income
group (family income Rs. >2000/- per month) and 380 belonged to the low
income group (monthly family income Rs.<2000/-). Interestingly, the
proportion of females in the high income group was more (115/119; 0.967),
compared to the low income group (178/202; 0.88). Whether the difference is due
to difference in prevalence or due to difference in access to medical care
between sexes cannot be ascertained; the latter possibility however is likely
(43).
Diabetes being so complex to manage, compliance to treatment can be expected to be difficult. In a study perfomed at All India Institute of Medical Sciences, New Delhi, more males than females were likely to be irregular in their contact and follow-up with diabetes education programme (56). Childhood and early adolescence is the period of intense change and necessity to cope up, even without the additional burden of managing diabetes mellitus. Given that manifest needs differ in the genders (57), it is likely that diabetes and its treatment could adversely affect meeting these needs in the adolescent period.
IDDM affects not only the child but the immediate family as well. Management depends on adult intervention, especially in the pre-school and early school going age group(58). Often, the mother carries a disproportionate share of the burden of diabetes care (59). When fathers do not participate in the initial period of diabetes management, they feel out of touch with the complexities of its management. Therefore care should be taken to distribute the responsibility between the two parents.
Gender differences in IDDM parents as transmitters of their disease
IDDM being partly genetic, the risk of transmission to further generations is
real. Studies have shown that there could be gender differences in the risk of
transmission between the mother with IDDM and father with IDDM. Warram et al
suggested that offspring of type I diabetic fathers were at higher risk for
developing autoimmmune diabetes than offspring of type I diabetic mothers (60).
This difference appeared to be confined to fathers with a DR4 allele, and was
not due to selective loss of diabetes susceptible fetuses in the perinatal
period (61).
In conclusion
A variety of differences were reported, both in the presence and absence of
gender differences between boys and girls with diabetes. Such information gives
insights into possible differences in pathogenesis, intervention strategies, as
well as ways to overcome obstacles in delivering health care between the sexes.