  |
|
|
||||
|
||||||
| Journal of the New Zealand Medical Association,
02-June-2006, Vol 119 No 1235
|
||||||
|
First
national audit of the outcomes of care in young people with diabetes in New
Zealand: high prevalence of nephropathy in Māori and Pacific
Islanders
Adrian Scott, Robyn Toomath, David Bouchier, Raymond Bruce,
Nic Crook, David Carroll, Rick Cutfield, Paul Dixon, John Doran, Peter
Dunn, Cheri Hotu, Maunt Khant, Maureen Lonsdale, Helen Lunt, Esko Wiltshire,
Denise Wu
New
Zealand (NZ) has a population of 4 million people with approximately 20%
identifying themselves as Māori, and 5% as Pacific Islanders. In 2000,
there were an estimated 115,000 people with known diabetes (NZ Ministry
of Health data)—predominantly Type 2 diabetes mellitus (T2DM).
Under the age of 26 years, most people with diabetes have
Type 1 diabetes mellitus (T1DM)—although with the rising tide of obesity,
more teenagers and young adults are found to have T2DM. The estimated number of
people with diabetes under age 26 in NZ is uncertain, but Christchurch data
indicate an increase in T1DM over the last 30 years.1,2
Children diagnosed with diabetes have greater morbidity and
mortality at all ages compared to their non-diabetic counterparts.3–5
Moreover, compared with adult-onset diabetes patients, the risk of developing
renal or retinal complications is greater if diabetes is diagnosed under 15
years.6
It is over 10 years since the Diabetes Control and
Complications Study reported the beneficial effects of tight glycaemic control
in adults and adolescents.7 The care of young people with diabetes is
challenging, and recent studies from Europe and Japan have illustrated the
difficulties of achieving and maintaining good glycaemic control. The studies
also highlight the high prevalence of complications and the wide range of
glycaemic control between centres (unrelated to patient selection or choice of
insulin regimen).8–13 Nevertheless, some centres consistently have mean
HbA1cs as good if not better than the intensive arm of DCCT, without the
increased risk of hypoglycaemia.14–16
Little is known about the prevalence or progression of Type
2 diabetes in children and young adults in New Zealand, but with the increase in
obesity, the picture is likely to mirror that of the rest of the World.17
MethodsStudy
design—This audit had approval from the local ethics committee of
each participating district health board (DHB).
The all-NZ young person’s diabetes audit was
begun in July 2003 when 14 centres (adult and paediatric diabetes services in
each) covering 2.3 million of the NZ population were invited to participate in
an audit of the process and outcomes of care of young people with diabetes up to
the age of 26. By April 2004, 1 centre in South Island and 8 in the North (8
adult and 6 paediatric) had submitted data. Data from the Waikato has been
published in detail elsewhere.18
If they had attended a diabetes centre at least once in
the previous 3 years, any person with diabetes born after 1 January 1978 was
eligible for inclusion in the study. Up to 45 data items (including date of
birth, duration of diabetes, last weight, height and BMI, lowest HbA1c during
first year after diagnosis, last HbA1c (and mean of last 3), presence of
microvascular complications, and details of treatment regimen were collected
from either paper health records or electronic diabetes registers.
Pathology laboratory databases were searched for
missing test results. The hospital number was used to eliminate or combine
duplicates where individuals had attended more than one centre over the last 3
years.
HbA1c was measured by a variety of methods but all were
Diabetes Control & Complications Trial (DCCT)-aligned. Microalbuminuria was
defined as an early morning urine with an albumin:creatinine ratio of >2.5 in
males (>3.5 females) on more than one occasion. Where available the number of
abnormal urines was recorded.
Statistics—Statistical
analysis was performed using SPSS for Windows (version 12, SPSS Inc., Chicago,
IL) software. Univariate ANOVA was used to compare centre differences (using
HbA1c as the dependent variable), with age and duration of diabetes as
covariates.
Chi-squared was used to compare ethnic differences and
type of diabetes. Pearson correlation coefficient was used to explore the
relationship between early glycaemic control and recent HbA1c. The significance
value was set to 5%
ResultsThere were 1282 (1251 after duplicates removed) people with
diabetes under the age of 26 as of 1 July 2003. The number, gender, ethnicity,
and type of diabetes by centre can be seen in Table 1. Of
Europeans with diabetes, 90.4% had T1DM compared to 66% of non-Europeans
(p<0.0001).
The mean Hba1c was 9.1±0.1%. After correcting for
gender and ethnicity, there were significant differences in mean HbA1c between
the 2 centres, with the highest values (9.6±0.2%) and lowest
(8.5±0.12%) values (p=0.000). There were significant differences in HbA1c
between age groups (p=0.034) (Figure 1).
When type of diabetes was included as a variable, age group
remained significant (p=0.06). For T1DM alone, age group (p=0.017) and centre
(p<0.0001) were highly significant.
There were no ethnic differences in glycaemic control either
during the first year after diagnosis, or the latest or the mean,
HbA1c. An HbA1c measured during the
first year after diagnosis was available in 274 (27%) of the subjects. The mean
of the lowest recorded HbA1c during that year was 8.2±0.1%.
Of those who had a recorded HbA1c during this period, 12.5%
were in the normal range. There was a positive correlation between the lowest
HbA1c during the first year after diagnosis and future glycaemic control (Figure
2).
There were 662 subjects with Type 1 diabetes between the
ages of 16–25 years. Over 50% were on multiple injection therapy (4 or
more injections per day); 23 patients were on pumps (predominantly from 1
centre). There was no correlation between number of injections and glycaemic
control. The prevalence of microalbuminuria and retinopathy varied considerably
from centre to centre, and increased with increasing duration of diabetes
(Figure 3 and Figure 4).
Figure 2: Relationship between lowest HbA1c during
first year after diagnosis and latest HbA1c (mean of last 3) in a sub-sample of
274 who had HbA1c results available from the first year after diagnosis.
Pearsons correlation is significant (p<0.01) (2-tailed)
 There were 105 subjects with T2DM; 63% were non-European.
The mean age was 20±0.4y with duration of 3±0.3 y. All were overweight
(mean±SE BMI 35±0.8). The majority were managed with diet alone, 19%
insulin treated and 8% on Metformin or Acarbose with or without insulin.
The mean HbA1c was 8.5±0.2%; 20% were hypertensive
(blood pressure [BP] > 130/80 mmHg), 72% had microalbuminuria (of whom 19%
were treated with ACE-inhibitors), 4% had background retinopathy, and 4% had
sight-threatening retinopathy.
The mean total cholesterol was 5.5±0.1 mmol/L, HDL
cholesterol 1.2±0.05 mmol/L, triglycerides 3.5±0.5 mmol/L. Amongst
those with T2DM, hyperlipidaemia was common with 61.5% having a total
cholesterol >5.0mmol/L, 36% an HDL < 1 mmol/L, and 52.6% with
triglycerides > 2.0 mmol/L. Only 3% were on lipid-lowering drugs.
Figure 3. Prevalence of microalbuminuria in 16–25
year olds with T1DM (left bars), including those with diabetes for 10 years (y)
(right bars)
 Over the age of 12 years, microalbuminuria were more common
amongst Māori and
Pacific Islanders (43.8%) compared to Europeans (17%) or Others (17.8%). This
was independent of the type of diabetes.
Figure 4. Prevalence of retinopathy in 16–25 year
olds with T1DM and effect of duration of diabetes (Centre 6 reported no
retinopathy)
 DiscussionThis is the first study to look at the outcomes of care of
children and young adults with diabetes across NZ and involved 8 adult and 6
paediatric diabetes services in 9 centres. The results demonstrate a
disappointing picture of poor glycaemic control and moderately high rates of
microvascular complications, as seen in other studies. Despite widespread use of
multiple injection therapy, at all ages, few people achieved satisfactory
control (only 22% had a recent HbA1c < 8%).
Although glycaemic control in each centre was poor, it is
similar to other published studies in Europe of unselected young people with
diabetes.8–12
Little data on young people with diabetes in NZ is available. In 2002, the
Christchurch group reported a mean HbA1c for females aged between 13 and 20
years was 10.2% and 9.5% for males.19
The type of insulin regimen (including use of pumps) did not
appear to have much impact on glycaemic control. Our finding of a relationship
between the HbA1c during the first year after diagnosis and future glycaemic
control is consistent with data from Australia which suggested that poor control
in childhood led to poor control in adolescence and beyond.20
Other studies have suggested that poor early control is
associated with a four-fold increase in the subsequent prevalence of
nephropathy.21 An intriguing observation of the DCCT collaborators was that
tight control initiated a year after diagnosis was associated with preservation
of islet cell function for a greater period than the group randomised to
conventional (poor) control.22
There have been a few small studies (but no long-term
randomised studies) looking at the impact (on beta cell function) of intensive
normalisation of glycaemic control from diagnosis, with conflicting
results.23,24
Recent twin studies may offer an alternative explanation for
the association between early and long term glycaemic control, which suggest
that 62% of the population variance in HbA1c levels is genetically determined
and independent of the genes influencing fasting blood glucose.25
This is unlikely to be the explanation, since both in our
study and in a UK study26 there were marked differences between centres in the
number of children with a normal HbA1c during the first year after diagnosis.
This implies differences in both expectation and training of the person with
diabetes rather than differences in genetics.
Access to health is not always equitable and socioeconomic
factors may explain some of the differences between centres, though in the
Scottish study, age, sex, insulin regimen, BMI, season, social circumstances,
and family history were all associated with glycaemic control but not with
deprivation score based on post code.8
The DCCT trial also suggested a period of improved control
during adolescence is associated with long-term improvements in risk of
complications—although HbA1c became similar in the intensive management
and control groups soon after the end of the DCCT, the benefits of intensive
management on microvascular complications persisted.27 This suggests adolescence
is a critical period for future risk of complications.
The success of the DCCT, and the difficulties in obtaining
similarly improved control outside the clinical trial setting (together with the
association we have observed between metabolic control soon after diagnosis and
future metabolic control) suggest that more intensive effort in diabetes
education, support, and motivation (as occurred in the intensive arm of the DCCT
along with intensive insulin management) may be particularly important during
the first year after diagnosis as well as during adolescence.
The prevalence of retinopathy amongst those screened was
similar to published series,9,10,28,29 but the difference in screening methods
between centres (from direct ophthalmoscopy without pupillary dilatation to
retinal photography with mydriasis) makes comparisons between centres difficult.
Microalbuminuria rates showed less variation which makes the
finding of no retinopathy in one centre (where screening was with direct
ophthalmoscopy) suspect, and emphasises the need for standardised screening
methods across the country.
Some centres did not begin retinal screening until 16
whereas the International Society for Paediatric and Adolescent diabetes (ISPAD)
guidelines are to start screening for microvascular complications from either 5
years after diagnosis or age 11 (whichever is earlier) with pre-pubertal
diagnosis or from 2 years from diagnosis with pubertal onset.30The extremely
high prevalence of
microalbuminuria in
those with T2DM, and the known high incidence of renal failure and increased
mortality in Māori and Pacific Islanders with the metabolic
syndrome,31–33 make early diagnosis and intervention essential.
This ethnic predisposition to nephropathy was apparent in both T1DM and T2DM.
Although the cause of nephropathy was not confirmed by renal
biopsy in our study, adult studies in obese patients with T2DM suggest that
persistent microalbuminuria is associated with either diabetic nephropathy, or
obesity related focal and segmental sclerosis.
Use of ACE inhibitors is reasonable in those with confirmed
nephropathy, although in part this depends on the criteria for diagnosis (in one
centre, of 13 young people with 3 or more abnormal results, 11 were on ACE
inhibitors. None of those with just 2 abnormal results were treated with
ACE-inhibitors).
A recent publication found that up to 60% of people with
T1DM have spontaneous resolution unrelated to ACE inhibitor use.34 This finding,
and concerns about using ACE inhibitors in young women of child bearing age, may
be the reasons why they are not frequently used.
Sub-optimal lipid profiles were very common especially in
those with type 2 diabetes, yet only 3% were receiving any lipid lowering
therapy. Cardiovascular Risk charts underestimate risk and are inappropriate for
this age group.35 Only the American Diabetes Association has published specific
guidelines for young people with T1DM.36
With the knowledge that most will die prematurely from a
vascular accident, and that vascular disease is even more common in T2DM
(especially in
Māori and Pacific
Islanders), earlier use of statins may be appropriate. As with use of ACE
inhibitors, however, consideration has to be given to the risk to the developing
foetus in the event of conception occurring whilst taking them.
Nearly 36% of those with T1DM over 16 yrs, and 100% of those
with T2DM, are overweight. This reflects obesity in the community as in 1997
approximately 25% of 15 to 18 year olds and one in three 19 to 24 year olds were
overweight or obese.37 In addition, weight gain (especially in girls) on
intensive insulin therapy can sometimes be spectacular,38,and likely to be a
disincentive to better glycaemic control.
A limitation of our study is that the exact number of people
under 26 years of age with diabetes in NZ is unknown. However, a careful
prevalence study from Christchurch, New Zealand39 estimates that there are
approximately 2540 with T1DM in this age group in the country.
Using their estimates of prevalence for the participating
centres we appear to have identified 92% of predicted. Of those young people
‘lost to follow-up’, the published literature suggests they have
worse control and greater risk of complications.40
Another limitation is that not all centres in New Zealand
participated in the study. Of concern, some were unable to gather the data. On
the basis of studies from countries with similar socioeconomic circumstances as
New Zealand, it is unlikely that the addition of data from these remaining
centres would have altered our conclusions, but it is vital that all centres
(caring for children and young adults with diabetes) audit their services and we
urge them to do so.
In summary, this multicentre study of nearly 50% of the
children and young people with diabetes in NZ has revealed large numbers with
poor glycaemic control and a disturbing prevalence of early microvascular
disease, despite introduction of intensive insulin therapy. Nevertheless, there
are highly significant differences between centres independent of other factors
suggesting opportunities for improvement.
The factors influencing success or otherwise in achieving
good glycaemic control need to be investigated further. There needs to be
greater adherence to management guidelines in screening for complications. T2DM
in young people is becoming a major problem and is associated with early onset
nephropathy and dyslipidaemia (almost from diagnosis), thus suggesting the need
for earlier diagnosis, which is likely only going to be achieved by targeted
screening of high risk children and young adults.
Author information:
Adrian R Scott, Endocrinologist;1 Robyn Toomath, Endocrinologist;2 David
Bouchier, Paediatrician;1 Raymond Bruce, Endocrinologist;3 Nic Crook,
Endocrinologist;4 David Carroll, Endocrinologist;3 Rick Cutfield,
Endocrinologist;5 Paul Dixon; Endocrinologist;6 John Doran, Paediatrician;7
Peter Dunn, Endocrinologist;1 Cheri Hotu, Specialist Registrar;5 Maunt
Khant, Endocrinologist;8 Maureen Lonsdale, Diabetes Educator;7 Helen Lunt,
Endocrinologist;9 Esko Wiltshire, Paediatrician;10 Denise Wu, Medical
Student9
Acknowledgements:
We thank Professor Michael Greenacre (Department of Economics, University Pompeu
Fabra, Barcelona, Spain) for statistical help and advice as well as Robyn
Toomath for the original idea for this study.
Correspondence: Dr
Adrian Scott, Diabetes Centre, Northern General Hospital, Herries Road,
Sheffield S5 7AU, UK. Fax: +44 114 2266924. email: adrian.scott@sth.nhs.uk
References:
|
|
||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|