Page 286 - HIVMED_v21_i1.indb

Page 4 of 8 Original Research

Analysis of electrolytes significantly associated with potassium abnormalities. In
PLWH/DM, for every year increase in the duration of DM,
Sodium
the odds of hypokalaemia increased by 97% (OR: 1.97; 95%
Hyponatraemia was the second most frequent electrolyte CI: 1.13–3.43; p = 0.025). However, in HIV-uninfected patients
abnormality, which occurred in 18 (18.75%) PLWH/DM and with DM, the odds of hyperkalaemia increased by 10% (OR:
26 (13.54%) HIV-uninfected patients with DM (Table 1). 1.10; 95% CI: 1.02–1.19; p = 0.048). Multivariate linear
Serum sodium was the only electrolyte significantly regression also showed significant associations between
negatively correlated with HbA1c in both PLWH/DM serum potassium levels and the duration of DM. For every
(r = -0.34; p = 0.001) and HIV-uninfected patients with DM year increase in the duration of DM, serum potassium
(r = -0.28; p < 0.001) (Table 2). Adjusted multinomial logistic decreased by 0.04 mmol/L amongst PLWH/DM (β = -0.04;
regression analysis amongst PLWH/DM suggests that for p = 0.018) and increased by 0.01 mmol/L amongst HIV-
every per cent increase in HbA1c, the odds of hyponatraemia uninfected patients with DM (β = 0.01; p = 0.042) (Table 4).
significantly increased by 55% (odds ratio [OR]: 1.55; 95%
confidence interval [CI]: 1.19–2.02; p = 0.003), whilst in HIV- Calcium
uninfected patients with DM the odds of hyponatraemia
significantly increased by 26% (OR: 1.26; 95% CI: 1.04–1.54; Serum-corrected calcium was the only electrolyte with
p = 0.009) (Table 3). Multivariate linear regression showed median (IQR) levels significantly lower in PLWH/DM
significant associations between serum sodium and HbA1c. compared with HIV-uninfected patients with DM (2.24 [2.18–
Amongst PLWH/DM, for every per cent increase in HbA1c, 2.30] mmol/L vs. 2.29 [2.20–2.36] mmol/L; p = 0.001).
serum sodium decreased by 0.51 mmol/L (β = -0.51; Furthermore, the most frequent electrolyte abnormality in
p = 0.004), and for every per cent increase in HbA1c amongst PLWH/DM and HIV-uninfected patients with DM was
HIV-uninfected patients with DM, serum sodium decreased hypocalcaemia (31.25% vs. 22.91%) (Table 1). Adjusted
by 0.45 mmol/L (β = -0.45; p < 0.001) (Table 4). multinomial logistic regression in PLWH/DM and HIV-
uninfected patients with DM found no factors significantly
associated with hypocalcaemia or hypercalcaemia. However,
Potassium
multivariate linear regression analysis in HIV-uninfected
The duration of HIV was not significantly associated with patients with DM showed that for every year increase in age,
hypokalaemia on adjusted multinomial logistic regression serum calcium decreased by 0.01 mmol/L (β = -0.01; p = 0.031)
analysis (OR: 0.85; 95% CI: 0.59–1.23; p = 0.645). Furthermore, (Table 4).
the odds of hypokalaemia in PLWH using TDF compared
with non-TDF-based ART were not significant (OR: 0.87; 95% Phosphate
CI: 0.06–13.12; p = 0.766). Adjusted multinomial logistic
regression determined that the duration of DM was Adjusted multinomial logistic regression in PLWH/DM
and HIV-uninfected patients with DM found no
TABLE 2: Correlation of electrolytes and glycated haemoglobin in people living factors significantly associated with hypophosphataemia
with human immunodeficiency virus and diabetes mellitus, and human or hyperphosphataemia. Notably, the use of TDF was
immunodeficiency virus-uninfected patients with diabetes mellitus. not significantly associated with hypophosphataemia in
Electrolyte PLWH/DM HIV-uninfected patients with DM
r p r p PLWH/DM (OR: 1.69; 95% CI: 0.20–14.12; p = 0.800).
Sodium −0.340 0.001 −0.28 < 0.001 Multivariate linear regression in PLWH/DM determined
Potassium 0.090 0.397 −0.08 0.283 that for every year increase in age, serum phosphate
Calcium 0.764 0.475 0.05 0.475 decreased by 0.01 mmol/L (β = -0.01; p = 0.033). Moreover,
Phosphate 0.110 0.302 −0.05 0.473 on average, serum phosphate was 0.09 mmol/L higher in
Source: Edendale Hospital diabetes clinic datasheet women than in men amongst HIV-uninfected patients with
r = Pearson’s correlation coefficient. DM, and all other variables were constant (β = 0.09;
PLWH/DM, people living with HIV and diabetes mellitus; HIV, human immunodeficiency
virus; DM, diabetes mellitus. p = 0.006) (Table 4).
TABLE 3: Adjusted and un-adjusted odds ratio estimates for the associations of electrolyte abnormalities with glycated haemoglobin in people living with human
immunodeficiency virus and diabetes mellitus, and human immunodeficiency virus-uninfected patients with diabetes mellitus.
Parameter PLWH/DM HIV-uninfected patients with DM
Univariate Multivariate Univariate Multivariate
OR 95% CI p aOR 95% CI p OR 95% CI p aOR 95% CI p
Hyponatraemia 1.39 1.13–1.71 0.007 1.55 1.19–2.02 0.003 1.22 1.01–1.46 0.009 1.26 1.04–0.54 0.009
Hypernatraemia 0.99 0.57–1.73 - 0.16 0.00–10.50 - 0.64 0.44–0.95 - 0.67 0.44–1.02 -
Hypokalaemia 1.10 0.84–1.45 0.565 1.20 0.72–1.99 0.736 0.98 0.79–1.22 0.215 0.99 0.80–1.23 0.089
Hyperkalaemia 1.11 0.88–1.42 - 1.05 0.82–1.35 - 0.83 0.68–1.02 - 0.75 0.58–0.97 -
Hypocalcaemia 0.91 0.78–1.07 0.467 0.89 0.75–1.06 0.426 0.89 0.77–1.03 0.226 0.91 0.79–1.05 0.321
Hypercalcaemia 0.76 0.32–1.84 - 0.03 0.00–1.27884E12 - 0.82 0.51–1.34 - 0.78 0.43–1.42 -
Hypophosphataemia 0.80 0.54–1.20 0.017 0.76 0.48–1.22 0.078 1.01 0.79–1.29 0.199 1.01 0.79–1.29 0.250
Hyperphosphataemia 1.53 1.12–2.11 - 1.44 1.00–2.07 - 0.79 0.61–1.02 - 0.79 0.60–1.04 -
Source: Edendale Hospital diabetes clinic datasheet
OR, odds ratio; aOR, adjusted odds ratio; CI, confidence interval; DM, diabetes mellitus; HbA1c, glycated haemoglobin; PLWH/DM, people living with HIV and diabetes mellitus; HIV, human
immunodeficiency virus.

http://www.sajhivmed.org.za 278 Open Access