Table 1.
Baseline characteristics of participants admitted to the clinic with DKA, classified into those on or off insulin 12 months after presentation
Characteristic | All (n=88) |
On insulin (n=48) |
Off insulin (n=40) |
p value |
---|---|---|---|---|
Demographics | ||||
Age, years | 35±10 | 33±10 | 38±10 | 0.014 |
Male sex | 50 (57) | 25 (52) | 25 (63) | 0.326 |
Ethnicity | ||||
Black | 50 (57) | 25 (52) | 25 (63) | 0.390 |
Mixed ethnicity (Coloured)a | 38 (43) | 23 (48) | 15 (38) | |
Clinical history | ||||
Family history of diabetes | 46 (52) | 23 (48) | 23 (58) | 0.370 |
History of other metabolic diseaseb | 16 (18) | 7 (15) | 9 (23) | 0.338 |
HIV diagnosis | 4 (5) | 1 (2) | 3 (8) | 0.326 |
History of smokingc | 33 (38) | 17 (35) | 16 (40) | 0.658 |
History of alcohol usec | 42 (48) | 25 (52) | 17 (43) | 0.399 |
Precipitant identified | 22 (25) | 13 (27) | 9 (23) | 0.621 |
Precipitated by infection | 20 (23) | 12 (25) | 8 (20) | 0.740 |
Examination | ||||
BMI, kg/m2 | 28.5 (23.3–33.4) | 25.1 (22.0–31.0) | 32.0 (28.3–36.2) | <0.001 |
BMI <18.5 kg/m2 | 3 (3) | 3 (5) | 0 | 0.193 |
BMI 18.5–24.9 kg/m2 | 24 (27) | 19 (40) | 5 (13) | 0.004 |
BMI 25–29.9 kg/m2 | 22 (25) | 13 (27) | 9 (23) | 0.404 |
BMI 30–39.9 kg/m2 | 27 (31) | 9 (19) | 18 (45) | 0.007 |
BMI >40 kg/m2 | 12 (14) | 4 (8) | 8 (20) | 0.101 |
Acanthosis nigricans present | 52 (59) | 15 (31) | 37 (93) | <0.001 |
Systolic BP (mmHg) | 127±17 | 125±18 | 131±14 | 0.075 |
Diastolic BP (mmHg) | 79 ±12 | 78 ±10 | 84 ±12 | 0.006 |
Admission investigations | ||||
Glucose, mmol/l | 27.9 (23.6–44.0) | 28.7 (23.3–44.0) | 27.8 (24.0–37.8) | 0.900 |
pH | 7.15 (7.04–7.22) | 7.10 (7.00–7.18) | 7.20 (7.11–7.25) | <0.001 |
Bicarbonate, mmol/l | 10.0 (6.8–12.3) | 8.4 (6.4–11.1) | 12.2 (7.5–14.5) | 0.001 |
Investigations | ||||
HbA1c, mmol/mol | 112±21 | 115±21 | 111±22 | 0.386 |
HbA1c, % | 12.4±1.9 | 12.7±1.9 | 12.3±2.0 | 0.386 |
Creatinine, μmol/l | 62±15 | 60±16 | 64±15 | 0.041 |
Total cholesterol, mmol/l | 4.9±1.2 | 5.0±1.1 | 5.0±1.3 | 0.983 |
Triglycerides, mmol/l | 1.6±0.7 | 1.5±0.8 | 1.8±0.7 | 0.010 |
HDL-cholesterol, mmol/l | 1.3±0.5 | 1.5±0.6 | 1.2±0.4 | 0.024 |
LDL-cholesterol, mmol/l | 2.9±1.0 | 2.9±0.9 | 3.0±1.1 | 0.440 |
Triglyceride/HDL-cholesterol ratio | 1.4±1.0 | 1.3±1.2 | 1.7±0.9 | 0.004 |
Bloods phenotype | ||||
Fasting plasma glucose, mmol/l | 11.2±6.6 | 11.5±9.6 | 7.8±5.4 | 0.007 |
Fasting C-peptide, nmol/l | 0.33 (0.17–0.62) | 0.23 (0.13–0.23) | 0.45 (0.30–0.70) | <0.001 |
C-peptide/glucose, nmol/mmol×100 | 3.43 (1.47–6.83) | 2.34 (0.98–4.32) | 5.95 (3.10–10.48) | <0.001 |
Fasting C-peptide >0.3 nmol/l | 50 (57) | 19 (40) | 31 (78) | <0.001 |
Fasting C-peptide >0.2 nmol/l | 65 (74) | 29 (60) | 36 (90) | <0.001 |
Anti-GAD positive (>10 U/ml) | 20 (23) | 19 (40) | 1 (3) | <0.001 |
Anti-IA-2 positive (>10 U/ml) | 11 (13) | 9 (19) | 2 (5) | 0.061 |
Antibody (GAD or IA-2) positive | 25 (28) | 22 (46) | 3 (8) | <0.001 |
Data are mean ± SD, n (%) or median (IQR) unless otherwise stated
aEthnicity was self-identified based on current official South African population group classification
bHistory of hypertension, peripheral vascular disease, cerebrovascular disease, ischaemic heart disease, dyslipidaemia, polycystic ovary syndrome or gestational diabetes
cAny use at all in the last year
p value is for Mann–Whitney U test for continuous variables and χ2 test or Fisher’s exact for the categorical variables