DIABETICMedicine

DOI: 10.1111/dme.12252

Research: Complications
Factors associated with diabetic ketoacidosis at onset of
Type 1 diabetes in children and adolescents

L. de Vries1,2, L. Oren1,2, L. Lazar1,2, Y. Lebenthal1,2, S. Shalitin1,2 and M. Phillip1,2

1
The Jesse Z and Sara Lea Shafer Institute for Endocrinology and Diabetes, National Center for Childhood Diabetes, Schneider Children’s Medical Center of Israel,
Petah-Tikva and 2Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

Accepted 6 June 2013

Abstract
Aims To identify risk factors for diabetic ketoacidosis at diagnosis of Type 1 diabetes in children and adolescents.
Methods In three time periods (1986–1987, 1996–1997 and 2006–2007) 75, 86 and 245 patients, respectively, aged
< 20 years were newly diagnosed with Type 1 diabetes in one tertiary care centre. In this retrospective comparative
study, data of clinical characteristics, laboratory evaluation at diagnosis, as well as demographic data were retrieved
from the patients’ files. Comparative analyses were performed between patients presenting with or without diabetic
ketoacidosis and between the three time periods.
Results Patients presenting with diabetic ketoacidosis were younger (9.2  4.7 vs. 10.4  4.7 years; P < 0.02),
thinner (weight standard deviation score –0.59  1.2 vs. –0.25  1.1; P = 0.002) and less frequently had a first- and/or
second-degree relative with Type 1 diabetes compared with those without diabetic ketoacidosis at presentation (16.0 vs.
31.2%, respectively; P = 0.001). Children with diabetic ketoacidosis were less likely to have had relevant testing before
diagnosis than children without diabetic ketoacidosis. Children aged < 2 years presented more often with diabetic
ketoacidosis than the older children (85 vs. 32%; P < 0.001). Children of Ethiopian origin had a higher rate of diabetic
ketoacidosis at diagnosis than the rest of the cohort (57.8 vs. 33%; P = 0.04).
Conclusions Factors affecting the risk of developing diabetic ketoacidosis at diagnosis of Type 1 diabetes may be
related to the degree of awareness of symptoms of diabetes among parents and primary care physicians. Prevention
programmes should aim at increasing awareness and consider the application of special measures to avoid diabetic
ketoacidosis in children aged < 2 years and high-risk ethnic groups.
Diabet. Med. 00, 000–000 (2013)

at diagnosis of Type 1 diabetes and to analyse possible

Introduction
Diabetic ketoacidosis is associated with significant morbidity
and non-negligible mortality and constitutes the most com-
mon cause of hospitalization and death in children with
diabetes [1]. Moreover, diabetic ketoacidosis at presentation
has been associated with decreased residual b-cell function
[2,3]. Diabetic ketoacidosis at disease onset in the paediatric
population has been reported to range from 12% to more
than 80% of cases at various sites in the world [4,5].
Although a recent study at our tertiary care centre found that
in the past decade there had been a decline in the rate of
diabetic ketoacidosis in paediatric Type 1 diabetes onset [6],
the current rate is still 29.4%. We therefore undertook to
identify the risk factors associated with diabetic ketoacidosis
Correspondence to: Liat de Vries. E-mail: liatd@clalit.org.il
changes in these factors over the past two decades, with the
aim to further lower the rate and severity of diabetic
ketoacidosis.
Patients and methods
Patients
The study sample comprised 406 patients aged < 20 years
diagnosed with new-onset Type 1 diabetes at a single tertiary
centre in three time periods: 75 diagnosed in 1986–1987
(period 1), 86 in 1996–1997 (period 2) and 245 in 2006–
2007 (period 3). Of the 462 files retrieved, 56 cases were
excluded because of any of the following: incomplete clinical
data, Type 2 diabetes mellitus, genetic defects of b-cell
function, drug- or chemical-induced diabetes, cystic fibro-

ª 2013 The Authors.

Diabetic Medicine ª 2013 Diabetes UK 1
DIABETICMedicine Factors associated with diabetic ketoacidosis at diagnosis  L. de Vries et al.
What’s new?
● Risk factors for diabetic ketoacidosis at diagnosis did
not change over two decades.
● Children aged < 2 years and children of Ethiopian
Jewish origin are at high risk for diabetic ketoacidosis
at diagnosis.
● Factors associated with lower risk were Type 1 diabe-
tes in a first- and/or second-degree family member,
screening tests for hyperglycaemia at the community
clinic prior to admission, higher weight standard
deviation score and older age at diagnosis.
● Recognizing the risk factors for diabetic ketoacidosis at
presentation would be the basis for a targeted preven-
tion programme aimed at increasing awareness and
knowledge of the clinical symptoms.
sis-related diabetes, genetic syndromes associated with dia-
betes and with insulin resistance or insulin deficiency.
The study protocol was approved by our Institutional
Review Board.
Methods
Our institutional diabetes registry consecutively records
every patient with new-onset diabetes referred to our clinic,
registering date of diagnosis, date of birth and gender.
Demographic data extracted from the medical files of each
patient included family history, ethnicity and documentation
of Type 1 diabetes in first- and second-degree relatives, as
well as presenting symptoms, duration of symptoms and
estimated weight loss. Ethnic groups included Israeli Arabs,
Ashkeanzi Jews, Ethiopian Jews, Yemenite Jews and Middle
Eastern Jews. Clinical data included age, weight, height and
pubertal stage (prepubertal: Tanner 1; in puberty: Tanner 2–
4; fully pubertal: Tanner 5) at diagnosis. Weight was
measured following full rehydration to avoid misinterpreta-
tion of fluid loss as weight loss. Weight standard deviation
score (SDS) at presentation was calculated. Data of BMI and
BMI SDS are not shown as BMI was only determined in
children > 2 years of age.
Tests performed at the community clinic (urine dipstick,
capillary glucose by portable glucose meter, laboratory
urinalysis and/or serum glucose) before referral to the
hospital were documented.
Laboratory analysis
Laboratory data included blood glucose level, pH, bicarbon-
ate and HbA1c at diagnosis. Diabetic ketoacidosis was
2 Diabetic Medicine ª 2013 Diabetes UK
defined as a pH value of < 7.3 or bicarbonate < 15 mmol/l
[4]. Severity of diabetic ketoacidosis was defined as follows—
mild: pH 7.2–7.3, serum bicarbonate 10–15 mmol/l; mod-
erate: pH 7.1–7.2, serum bicarbonate 5–10 mmol/l; severe:
pH < 7.1, serum bicarbonate < 5 mmol/l.
Standard laboratory methods in the clinical laboratory of
the Schneider Children’s Medical Center of Israel were used
to measure glucose and blood gases.
Statistical methods
All analyses were carried out using the BMDP program [7]
and the results are expressed as mean  SD or number
(percentage). Analyses were carried out for patients pre-
senting with or without diabetic ketoacidosis for the entire
cohort and for the three time periods. For continuous
variables, comparisons between groups were made using
analysis of variance (ANOVA) with Bonferroni’s adjust-
ment for multiple comparisons. Discrete variables were
compared using Pearson’s v2-test or Fisher’s exact test, as
appropriate. For those variables that did not have Gaussian
distributions, or had a relatively small sample size, com-
parisons between groups were performed using the Mann–
Whitney U-test (when comparing two groups) or the
Kruskal–Wallis test (when comparing more than two
groups). For variables that were found significantly differ-
ent between the study groups, a comparison was also made
between patients presenting with and those without
diabetic ketoacidosis. Using the variables found to be
significant on univariate analysis, we applied a stepwise
logistic regression to determine those variables significantly
associated with diabetic ketoacidosis. Included in the
logistic regression was also the period of diagnosis
(group 1, 2 or 3). A P-value of < 0.05 was considered
significant.
Results
Rate of diabetic ketoacidosis by age and ethnic origin
Overall 33.7% of patients presented with diabetic ketoaci-
dosis at diagnosis: 36.1% of the boys and 32.7% of the girls
(P = 0.47). For the entire cohort, the percentage of children
aged 0–4, 5–9, 10–14 and 15–20 years who presented with
diabetic ketoacidosis was 40.5, 35.7, 38.4 and 23.6%,
respectively, showing a significantly lower rate for the older
age group (P < 0.02) compared with the rest of the cohort.
These rates were similar when analysing by period of
diagnosis.
For all time periods, children < 2 years had a significantly
higher rate of diabetic ketoacidosis at presentation than the
older children (85 vs. 32%, P < 0.001); 34.6% had severe
diabetic ketoacidosis, compared with 5.1% of the older
patients (P < 0.001). These differences were observed in the
three time periods.
ª 2013 The Authors.
 Research article DIABETICMedicine

Patients of Ethiopian Jewish origin had a higher rate of
diabetic ketoacidosis at diagnosis than the rest of the cohort
(57.8 vs.33%; P = 0.04). Such a difference was not observed
for the other ethnic groups. However, the rate of patients of
Ethiopian origin presenting with severe diabetic ketoacidosis
was similar to that of the other groups (5.2%).
Characteristics of patients presenting with diabetic
ketoacidosis at diagnosis (Table 1)
Patient characteristics
Patients presenting with diabetic ketoacidosis were younger
(9.2  4.7 vs. 10.4  4.7 years; P < 0.02), weighed less
(weight SDS –0.59  1.2 vs. –0.25  1.1; P = 0.002) and
less frequently had a first- and/or second-degree relative with
Type 1 diabetes than those without diabetic ketoacidosis at
presentation (16.0 vs. 31.2%, respectively; P = 0.001). In the
group with diabetic ketoacidosis, the proportion of patients
with a family history of an autoimmune disorder other than
Type 1 diabetes tended to be lower than for the group
without diabetic ketoacidosis (16.7 vs 24.6%; P = 0.07).
The distribution of patients by gender and by pubertal
status was similar in both groups. Upon analysis by monthly
rate as well as by seasons, the diabetic ketoacidosis rate was
found to be similar all year round. Neither the number of
children in the family nor the order of the child in the family
(whether first, second, third, etc.) were associated with
diabetic ketoacidosis at diagnosis.
Presenting signs and symptoms
Duration of symptoms before diagnosis was similar for the two
groups—those with and without diabetic ketoacidosis at
presentation—and no correlation was found between symp-
tom duration and pH. The prevalence of the classical symp-
toms of polydipsia, polyuria and nocturia was also similar.
As expected, the rate of patients who reported weight loss
before diagnosis (78.8 vs. 66.8%; P = 0.01), as well as the
extent of weight loss (8.5  6.6 vs. 5.3  5.5% of body
weight; P < 0.001), were significantly higher among patients
with diabetic ketoacidosis than for the rest of the cohort.
Significant differences between groups were observed for all
symptoms related to diabetic ketoacidosis, such as dyspnoea,
abdominal pain, vomiting, etc.
The differences between those who presented with diabetic
ketoacidosis and those who did not were similar in the three
study periods (Table 2).

Table 1 Characteristics of patients presenting with and without diabetic ketoacidosis at diagnosis

Diabetic ketoacidosis No diabetic ketoacidosis

(n = 137) (n = 269) P-value

Age at admission (years) 9.2  4.7 10.4  4.7 < 0.02

Height SDS 0.15  1.0 0.15  1.0 0.9
Weight SDS 0.59  1.3 0.25  1.1 0.002
Pubertal stage (pre-/in puberty/post-) (%) 59/29/18 52/31/18 0.9
Symptoms
Polyuria (%) 94.7 90 0.1
Polydipsia (%) 95.5 90.8 0.1
Appetite loss (%) 26 8.5 < 0.001
Weight loss (%) 78.8 66.8 < 0.02
Estimated weight loss (% of body weight) 8.5  6.6 5.3  5.5 < 0.001
Nocturnal enuresis (%) 39.1 42.9 0.5
Constipation (%) 2.3 0.4 0.1
Fatigue (%) 37.6 17.8 < 0.001
Weakness (%) 57.1 20.1 < 0.001
Dyspnoea (%) 15 0.8 < 0.001
Fever (%) 27.6 18.5 0.05
Abdominal pain (%) 29.3 13.1 < 0.001
Vomiting (%) 33.1 7.3 < 0.001
Genital irritation/fungal infection (%) 4.5 5.8 0.6
Mental alteration (%) 19.4 0.8 < 0.001
Hospitalized (%) 98.5 74.2 < 0.001
Hospitalization duration (days) 4.6  3.1 3.3  2.5 < 0.001
Laboratory findings
Serum glucose (mmol/l) 29.1  9.7 24.9  11.0 < 0.001
HbA1c (mmol/mol) 109  7 100  8 0.02
(%) (12.1  2.8) (11.3  2.9)
pH 7.17  0.11 7.36  0.04 < 0.001
Bicarbonate (mmol/l) 11.5  5.4 22.4  3 < 0.001
Creatinine (mmol/l) 0.82  0.26 0.76  0.15 < 0.03
Background
First- and/or second-degree relatives with Type 1 diabetes (%) 16.0 31.2 < 0.001
Glucose testing at primary care centre before referral (%) 57.1 78.3 < 0.001

*Data are mean  SD unless otherwise indicated.

ª 2013 The Authors.

Diabetic Medicine ª 2013 Diabetes UK 3
4
DIABETICMedicine

Table 2 Characteristics of patients presenting with and without diabetic ketoacidosis at diagnosis in three time periods*

Diabetic ketoacidosis No diabetic ketoacidosis

1986–1987 1996–1997 2006–2007 1986–1987 1996–1997 2006–2007 P for diabetic

(n = 30) (n = 36) (n = 72) (n = 45) (n = 50) (n = 173) ketoacidosis† P for group‡

Age at admission (years) 9.7  5.1 9.4  4.6 8.4  4.7 11.6  4.8 10.8  4.6 10.0  4.6 0.009 0.18
Weight SDS 0.91  1.2 0.37  1.2 0.57  1.3 0.67  0.98 0.27  1.3 0.14  1.1 0.05 0.02
Puberal stage
Pre-/puberty/post- 53/32/15 59/34/7 61/26/13 43/36/21 53/32/15 53/29/18 0.3 0.6
puberty (%)
Estimated weight 7.6  6.4 7.2  6.8 9.4  6.5 6.7  6.2 5.3  4.8 5  5.4 < 0.001 0.43
loss (% of
body weight)
Polyuria (%) 96 82 100 95 90 89 0.08 0.11
Polydipsia (%) 96 85 100 95 90 90 0.07 0.15
First- and/or second- 10.7 22.2 15.1 33.3 32.7 31.6 < 0.001 0.77
degree relatives with
Type 1 diabetes (%)
Glucose testing at primary 50.0 46.7 67.6 78.6 73.2 78.8 < 0.001 0.15
care centre
before referral (%)
Patients < 2 years (%) 14.3 8.8 12.3 0.0 2.0 1.2 < 0.001 1.2

*Data are mean  SD unless otherwise indicated.

†P-value for the difference between those with and those without diabetic ketoacidosis.
‡P-value for the difference between the three time periods.
No interaction was found for any of the variables.
Factors associated with diabetic ketoacidosis at diagnosis  L. de Vries et al.

Diabetic Medicine ª 2013 Diabetes UK

ª 2013 The Authors.
 Research article
Characteristics of groups with high risk for diabetic
ketoacidosis at diabetes onset
Compared with older children, children < 2 years of age
(n = 20) presented more frequently with vomiting (58 vs
14%; P < 0.001), dyspnoea (21 vs. 5%; P < 0.003) and
fever (47 vs. 21%; P < 0.006), and less frequently with
weight loss (47 vs 72%; P < 0.02), with no difference for
rates of polyuria, polydipsia and fatigue. Nocturnal enuresis
was reported in 43% of older children and in the one child
younger than 2 years who had been toilet-trained before
diagnosis (P = 0.001). Duration of symptoms before diag-
nosis was significantly shorter in children younger than
2 years (12.8  13.9 vs. 28.1  43.5 days; P = 0.003) and
they less frequently had screening tests for hyperglycaemia
before referral (25 vs. 69.9%; P < 0.001).
Children of Ethiopian Jewish origin (n = 19) were char-
acterized by a significantly higher proportion of female
patients (78.9 vs. 50.7%; P = 0.01), and no family history of
Type 1 diabetes in either a first- or second-degree relative in
any of the patients, compared with 32.3% in the rest of the
cohort (P = 0.001). They were similar to all patients from
other ethnic groups in all other clinical and biochemical
variables, except for the higher rate of diabetic ketoacidosis.
Using stepwise logistic regression, we found that factors
associated with diabetic ketoacidosis at presentation were
age < 2 years [for < 2 years old vs. older, odds ratio 18.0,
95% CI 3.7–28.7; P < 0.001] and weight loss before diag-
nosis [odds ratio 1.8 (95% CI 0.99–3.16); P = 0.05], while
factors associated with lower risk were Type 1 diabetes in a
first- and/or second-degree family member [odds ratio 0.28
(95% CI 0.15–1.0); P < 0.001], screening tests for hyper-
glycaemia at the community clinic prior to admission [odds
ratio 0.35 (95% CI 0.21–0.59); P < 0.001], higher weight
SDS [odds ratio 0.82 (95% CI 0.67–0.99); P = 0.05] and
older age at diagnosis [odds ratio 0.94 (95% CI 0.90–0.99);
P = 0.03]. The area under the receiver operating character-
istics curve for these variables was 0.73. Similar results were
found when performing logistic regression for each time
period separately. Although the rate of diabetic ketoacidosis
was significantly higher among children of Ethiopian origin,
we did not include ethnicity in the logistic regression, as our
cohort included patients diagnosed during 1986–1987 and
most Ethiopian Jews immigrated to Israel only after 1990.
The influence of the various factors studied on risk of
diabetic ketoacidosis at diagnosis of Type 1 diabetes is
summarized in Table 3.
Discussion
This study sought to determine which factors are associated
with diabetic ketoacidosis in children and adolescents with
new onset of Type 1 diabetes. Found to be at highest risk
were children younger than 2 years, those of Ethiopian
origin, those with a greater weight loss before diagnosis and
ª 2013 The Authors.
Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
Table 3 Risk factors for diabetic ketoacidosis at diagnosis
Increased risk Not affecting risk Reduced risk
Age < 2 years Gender Age > 15 years
Ethnic Ethiopian Year of diagnosis Family history of
origin Type 1 diabetes
Weight loss Time of year Higher weight
Pubertal stage Screening tests for
hyperglycaemia
Number of siblings
a lower weight at diagnosis. In contrast, children with a first-
or second-degree relative with Type 1 diabetes and those
who had a preceding screening test for hyperglycaemia at the
community centre seemed less likely to present with diabetic
ketoacidosis. Gender, pubertal stage and time of the year did
not affect the risk for diabetic ketoacidosis.
Our finding in the present study, that diabetic ketoacidosis
at diabetes onset was more common and more severe in
children younger than 2 years of age, is similar to observa-
tions reported by many others [8–13]. The short duration of
symptoms and high frequency and severity of diabetic
ketoacidosis in this age group has been attributed to a more
aggressive form of disease and a more rapid decline in b-cell
function, further aggravated by the higher incidence of
intercurrent infection precipitating acidosis [14,15], higher
risk of dehydration in the young and a less developed
mechanism of metabolic decompensation [16].
The higher rate of diabetic ketoacidosis may possibly also
be related to any of the following: (1) a delay in diagnosis
because of low awareness, as onset of Type 1 diabetes before
the age 2 years is uncommon; (2) failure to recognize typical
signs such as polyuria in an infant still using a nappy; (3) the
high rate of fever and/or acute infection that may have
masked the symptoms of diabetes [17]. Indeed, the rate of
reported nocturia and/or nocturnal enuresis in infants less
than 2 years of age in our cohort was significantly lower,
simply because all except one had not yet been toilet-trained.
The possibility of delayed diagnosis is supported by a
retrospective study of toddlers with diabetes, which found
that parents were more likely to recognize symptomatic
hyperglycaemia in children older than 2 years [11].
In a previous study, we found that patients presenting with
diabetic ketoacidosis were less likely to have had prior
glucose testing in the community before referral than the
patients without diabetic ketoacidosis (57.1 vs. 78.3%;
P < 0.001) [6].
The low rate of relevant laboratory tests performed in the
community before diagnosis in this young age group
supports the assumption of a lower awareness of diabetes
by either caregivers or physicians. A study from Ontario
recently reported that, although children presenting with
diabetic ketoacidosis more frequently had a medical encoun-
ter before diagnosis than children with diabetes without
5
DIABETICMedicine Factors associated with diabetic ketoacidosis at diagnosis  L. de Vries et al.
diabetic ketoacidosis, they were less likely to have had
relevant laboratory testing before diagnosis [8]. Unfortu-
nately, we were unable to collect comprehensive data on
medical encounters prior to diagnosis as data were not
computerized in the earlier years. It is therefore not possible
to know if the parents were unaware and did not seek help or
if the physician was unaware and did not perform relevant
laboratory tests.
Other than in the subgroup of children younger than
2 years, there was no significant difference in the diabetic
ketoacidosis rate in the children aged 0–4, 5–9 or 10–
14 years. These findings do not support those recently
reported from Finland showing that adolescents aged 10–
14 years were at increased risk of diabetic ketoacidosis [17].
The increased prevalence of diabetic ketoacidosis among
Ethiopian Jews could be attributable to different genetics
[18] or to a lower socio-economic status and parental
education. Although the latter variables were not assessed in
our study, most of our Ethiopian patients are of low
education and socio-economic status, factors which have
been shown to be risk factors for diabetic ketoacidosis [19].
Interestingly, none of our patients of Ethiopian origin
reported a family history of Type 1 diabetes in first- or
second-degree relatives. This may have been a contributing
factor to the lower awareness, resulting in a delay in
diagnosis.
Higher awareness and earlier recognition of the signs and
symptoms of hyperglycaemia could also explain our finding
of a higher rate of family history of Type 1 diabetes among
the patients without diabetic ketoacidosis at diagnosis. These
findings are in line with a previous report on familial Type 1
diabetes, showing that the diabetic ketoacidosis rate was
lower in the second affected family member than in sporadic
cases [20].
Weight loss before diagnosis is a typical sign of metabolic
derangement, with a higher risk of onset of diabetic
ketoacidosis. Therefore, the highest prevalence of weight
loss before diagnosis in children presenting with diabetic
ketoacidosis in our cohort is not surprising. This seems to be
the obvious explanation for the association between higher
weight SDS and lower risk for diabetic ketoacidosis. Yet,
higher weight SDS could be a protective factor by itself:
overweight in a child with evolving Type 1 diabetes could
result in insulin resistance [21] and hyperglycaemia, with an
earlier appearance of symptoms while insulin secretion is still
sufficient to suppress production of ketone bodies.
Children followed by the Diabetes Autoimmunity Study in
the Young, an observational study following children at
high risk for Type 1 diabetes by periodic testing in the USA,
had a milder clinical course at diagnosis: they did not
develop diabetic ketoacidosis, were rarely hospitalized, had
nearly normal HbA1c values at diagnosis and significantly
lower requirements for insulin in the first year of illness [22].
This milder clinical course at diagnosis may have very
important long-term implications, as near-euglycaemic
6 Diabetic Medicine ª 2013 Diabetes UK
control at onset occurring spontaneously [23] or achieved
by intensive insulin treatment has been shown to preserve
the secretion of insulin. The Diabetes Control and Compli-
cations Trial found that residual endogenous insulin secre-
tion predicted a 65% lower risk of severe hypoglycaemia
[24] and a 50% lower risk of the progression of diabetic
retinopathy [25].
Thus, the ultimate aim should be to decrease the rate and
severity of diabetic ketoacidosis, as well as to shorten the
duration of symptoms. This would reduce morbidity and
mortality, as well as healthcare costs, and also result in a
higher residual b-cell reserve. While a general educational
programme in Parma (Italy) has been reported to have
resulted in a major decrease in prevalence of diabetic
ketoacidosis at presentation [26], the literature contains no
other reports of such dramatic reductions in ketoacidosis
rates following publicity. In Wales, a publicity campaign did
not have any effect on ketoacidosis rates at diagnosis of
Type 1 diabetes [27]. Recognizing the risk factors for
diabetic ketoacidosis at presentation would be the basis for
a targeted prevention programme aimed at increasing
awareness and knowledge of the clinical symptoms. Such
programmes should consider the application of special
measures to avoid diabetic ketoacidosis in high-risk groups.
The ‘target audience’ should include caregivers of babies and
young infants, such as team members of baby clinic centres.
In neighbourhoods highly populated with ‘high-risk’ ethnic
groups, the means of education should be in their language
(in the case of Ethiopian Jews, in Amhari) and should be
adapted to their culture and beliefs.
The strength of the current study stems from the long-term
follow-up in a single tertiary centre, as well as the detailed
data. We were able to identify factors associated with
diabetic ketoacidosis at diagnosis of Type 1 diabetes in our
population and to show that these did not change over two
decades. The difference in the size of the three study groups
stems mostly from the increase in the size of the Israeli
population over the study period and the increase in
incidence of Type 1 diabetes.
Unfortunately, we were unable to collect precise data on
socio-economic factors such as family income or parental
education.
In conclusion, as diabetic ketoacidosis at diagnosis of
Type 1 diabetes is associated with a lower rate of partial
remission, the identification of risk factors for diabetic
ketoacidosis is of utmost importance. Children aged
< 2 years, children originating from specific ethnic groups
and those without a family history of Type 1 diabetes are at
high risk for diabetic ketoacidosis at diagnosis. Thus, efforts
should be directed at increasing awareness in the general
population, with specific programmes for high-risk groups.
Funding sources
None.
ª 2013 The Authors.
 Research article
Competing interests
None declared.
Acknowledgments
The authors wish to thank Pearl Lilos for the statistical
analysis and Ruth Fradkin for her editorial assistance. This
work was performed in partial fulfillment of the MD thesis
(LO) requirements of the Sackler Faculty of Medicine, Tel
Aviv University, Tel Aviv, Israel.
References
1 Scibilia J, Finegold D, Dorman J, Becker D, Drash A. Why do
children with diabetes die? Acta Endocrinol Suppl (Copenh) 1986;
279: 326–333.
2 Bowden SA, Duck MM, Hoffman RP. Young children (< 5 years)
and adolescents (> 12 years) with type 1 diabetes mellitus have low
rate of partial remission: diabetic ketoacidosis is an important risk
factor. Pediatr Diabetes 2008; 9: 197–201.
3 Abdul-Rasoul M, Habib H, Al-Khouly M. ‘The honeymoon phase’
in children with Type 1 diabetes mellitus: frequency, duration, and
influential factors. Pediatr Diabetes 2006; 7: 101–107.
4 Wolsdorf J, Craig ME, Daneman D, Dunger D, Edge J, Lee WRW
et al. ISPAD clinical practice guidelines 2009 compendium. Dia-
betic ketoacidosis in children and adolescents with diabetes. Pediatr
Diabetes 2009; 10: 118–123.
5 Usher-Smith JA, Thompson M, Ercole A, Walter FM. Variation
between countries in the frequency of diabetic ketoacidosis at first
presentation of type 1 diabetes in children: a systematic review.
Diabetologia 2012; 55: 2878–2894.
6 de Vries L, Oren L, Lebenthal Y, Shalitin S, Lazar L, Phillip M.
Decrease in frequency of ketoacidosis at diabetes onset over the past
two decades—perspectives of a paediatric tertiary care centre.
Diabet Med 2012; 29: e170–e175.
7 Dixon WJ, ed. BMDP Statistical Software. Los Angeles: University
of California Press, 1993.
8 Bui H, To T, Stein R, Fung K, Daneman D. Is diabetic ketoacidosis
at disease onset a result of missed diagnosis? J Pediatr 2010; 156:
472–477.
9 Neu A, Willasch A, Ehehalt S, Hub R, Ranke MB. Ketoacidosis at
onset of type 1 diabetes mellitus in children—frequency and clinical
presentation. Pediatr Diabetes 2002; 4: 77–81.
10 Abdul-Rasoul M, Al-Mahdi M, Al-Qattan H, Al-Tarkait N,
Alkhouly M, Al-Safi R et al. Ketoacidosis at presentation of type 1
diabetes in children in Kuwait: frequency and clinical characteris-
tics. Pediatr Diabetes 2010; 11: 351–356.
11 Quinn M, Fleischmann A, Rosner B, Nigrin DJ, Wolfsdorf JI.
Characteristics at diagnosis of type 1 diabetes in children younger
than 6 years. J Pediatr 2006; 148: 366–371.
12 Hekkala A, Knip M, Veijola R. Ketoacidosis at diagnosis of type 1
diabetes in children in Northern Finland. Temporal changes over
20 years. Diabetes Care 2007; 30: 861–866.
ª 2013 The Authors.
Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
13 Szypowska A, Sk
orka A. The risk factors of ketoacidosis in children
with newly diagnosed type 1 diabetes mellitus. Pediatr Diabetes
2011; 12: 302–306.
14 Mallore JT, Cordice CC, Ryan BA, Carey DE, Kreitzer PM, Frank GR.
Identifying risk factors for the development of diabetic ketoacidosis in
new-onset type 1 diabetes mellitus. Clin Pediatr 2003; 42: 591–597.
15 Nimri R, Phillip M, Shalitin S. Children diagnosed with diabetes
during infancy have unique clinical characteristics. Horm Res
2007; 67: 263–267.
16 Rosenbauer J, Icks A, Giani G. Clinical characteristics and
predictors of severe ketoacidosi at onset of type 1 diabetes mellitus
in children in a north Rhine-Westphalian region, Germany.
J Pediatr Endo Metab 2002; 15: 1137–1145.
17 Hekkala A, Reunanen A, Koski M, Knip M, Veijola R; Finnish
Pediatric Diabetes Register. Age-related differences in the frequency
of ketoacidosis at diagnosis of type 1 diabetes in children and
adolescents. Diabetes Care 2010; 33: 1500–1502.
18 Zung A, Elizur M, Weintrob N, Bistritzer T, Hanukoglu A, Zadik
Z et al. Type 1 diabetes in Jewish Ethiopian immigrants in Israel:
HLA class II immunogenetics and contribution of new environ-
ment. Human Immunol 2004; 65: 1463–1468.
19 Rewers A, Klingensmith G, Davis C, Petitti DB, Pihoker C,
Rodriguez B et al. Presence of diabetic ketoacidosis at diagnosis of
diabetes mellitus in youth: the Search for Diabetes in Youth Study.
Pediatrics 2008; 121: e1258–e1266.
20 Lebenthal Y, de Vries L, Phillip M, Lazar L. Familial type 1
diabetes mellitus—gender distribution and age at onset of diabetes
distinguish between parent–offspring and sib-pair subgroups.
Pediatr Diabetes 2010; 11: 403–411.
21 Knip M, Reunanen A, Virtanen SM, Nuutinen M, Viikari J,
Akerblom HK. Does the secular increase in body mass in children
contribute to the increasing incidence of type 1 diabetes? Pediatr
Diabetes 2008; 9: 46–49.
22 Barker JM, Goehrig SH, Barriga K, Hoffman M, Slover R,
Eisenbarth GS. Clinical characteristics of children diagnosed with
type 1 diabetes through intensive screening and follow-up. Diabe-
tes Care 2004; 27: 1399–1404.
23 Fernandez-Casta~ ner M, Monta~ na E, Camps I, Biarnes J, Merino JF,
Escriba JM et al. Ketoacidosis at diagnosis is predictive of lower
residual beta-cell function and poor metabolic control in type 1
diabetes. Diabetes Metab 1996; 22: 349–355.
24 Diabetes Control and Complications Trial Research Group. Effect
of intensive therapy on residual beta-cell function in patients with
type 1 diabetes in the Diabetes Control and Complications Trial. A
randomized, controlled trial. Ann Intern Med 1998; 128: 517–523.
25 Diabetes Control and Complications Trial Research Group: The
effect of intensive treatment of diabetes on the development and
progression of long-term complications in insulin-dependent dia-
betes mellitus. N Engl J Med 1993; 329: 977–985.
26 Vanelli M, Chiari G, Costi G, Ghizzoni L, Giacalone T, Chiarelli F.
Effectiveness of a prevention program for diabetic ketoacidosis in
children. Diabetes Care 1999; 22: 7–9.
27 Lansdown AJ, Barton J, Warner J, Williams D, Gregory JW, Harvey
JN et al. Brecon Group. Prevalence of ketoacidosis at diagnosis of
childhood onset Type 1 diabetes in Wales from 1991 to 2009 and
effect of a publicity campaign. Diabet Med. 2012; 29:1506–1509.
7