Biochemical and renal-metabolic profiles distinguishing primary aldosteronism from essential hypertension: a systematic review and meta-analysis with implications for endocrine hypertension screening

Figures

• 

  Figure 1. PRISMA 2020 flow diagram of study selection. The updated search identified 11,473 records. After 2,198 duplicate records were removed, 9,275 records were screened. Of 162 reports sought for retrieval, 41 were not retrieved and 121 were assessed for eligibility; 47 were excluded after full-text assessment, leaving 74 studies for inclusion. Detailed database yields and categorized full-text exclusion reasons are shown in the diagram.

• 

  Figure 2. Forest plots of serum potassium. (A) PA versus hypertensive controls (59 studies): pooled SMD = −1.15 (95% CI: −1.27 to −1.02), I² = 91.8%, p < 0.001. (B) Unilateral/lateralizing PA versus bilateral PA (17 studies): pooled SMD = −0.85 (95% CI: −1.13 to −0.57), I² = 87.8%, p < 0.001.

• 

  Figure 3. Meta-regression of serum potassium effect size on publication year. The main-text data-quality sensitivity analysis excluded Galetta et al., 2009 because of implausibly small reported SDs (k = 58). The regression coefficient was 0.038 per year (95% CI: 0.021 to 0.055, p < 0.001; R² = 18.2%), indicating attenuation of the PA–hypertensive-control potassium difference in more recent publications. The complete 59-study analysis (slope = 0.042/year; R² = 16.5%) is reported in the Supplementary Materials.

• 

  Figure 4. Subgroup analysis forest plots for serum potassium. (A) By publication period: studies published in or before 2015 (k = 23; SMD = −1.50) versus studies published after 2015 (k = 36; SMD = −0.96); between-subgroup Q = 17.04, p < 0.001. (B) By geographic region: East Asia (k = 36; SMD = −1.07) versus non-East Asia (k = 23; SMD = −1.28); between-subgroup Q = 2.10, p = 0.148.

• 

  Figure 5. Forest plots of serum sodium. (A) PA versus hypertensive controls (31 studies): pooled SMD = 0.51 (95% CI: 0.38 to 0.63), I² = 84.0%, p < 0.001. (B) Unilateral/lateralizing PA versus bilateral PA (7 studies): pooled SMD = 0.43 (95% CI: 0.22 to 0.64), I² = 66.6%, p < 0.001.

• 

  Figure 6. Forest plots of serum creatinine, eGFR, and BUN in PA versus hypertensive controls. (A) Serum creatinine (42 studies): pooled SMD = 0.01 (95% CI: −0.05 to 0.07), I² = 61.2%, p = 0.803. (B) eGFR (36 studies): pooled SMD = 0.02 (95% CI: −0.05 to 0.09), I² = 69.4%, p = 0.625. (C) BUN (8 studies): pooled SMD = −0.07 (95% CI: −0.21 to 0.06), I² = 61.0%, p = 0.287.

• 

  Figure 7. Sensitivity analysis forest plot of eGFR mean difference: PA versus hypertensive controls. Thirty-six studies were included. The pooled MD was 0.34 mL/min/1.73 m² (95% CI: −1.07 to 1.75), I² = 69.4%, p = 0.637. For comparison, the estimate of 3.93 mL/min/1.73 m² reported by Monticone et al. (2020) lies outside the 95% CI of the present analysis.

• 

  Figure 8. Subgroup analysis forest plot of eGFR stratified by exclusion of preexisting renal disease. Studies excluding renal disease (k = 15) yielded SMD = −0.04 (95% CI: −0.14 to 0.06), whereas studies not explicitly excluding renal disease (k = 21) yielded SMD = 0.06 (95% CI: −0.03 to 0.16). The between-subgroup difference was not significant (Q = 2.01, p = 0.156).

• 

  Figure 9. Forest plots of urinary albumin outcomes in PA versus hypertensive controls. (A) UACR/ACR (8 studies): pooled SMD = 0.49 (95% CI: 0.33 to 0.66), I² = 72.8%, p < 0.001. (B) Urinary albumin excretion (9 studies): pooled SMD = 0.73 (95% CI: 0.22 to 1.25), I² = 94.9%, p = 0.005. (C) Sensitivity analysis retaining one preferred albumin-specific indicator per study (19 studies): pooled SMD = 0.60 (95% CI: 0.38 to 0.82).

• 

  Figure 10. Forest plots of serum uric acid. (A) PA versus hypertensive controls (23 studies): pooled SMD = −0.23 (95% CI: −0.35 to −0.12), I² = 78.1%, p < 0.001. (B) Unilateral/lateralizing PA versus bilateral PA (3 studies): pooled SMD = −0.19 (95% CI: −0.40 to 0.03), I² = 13.7%, p = 0.094.

• 

  Figure 11. Cumulative meta-analysis of serum potassium: PA versus hypertensive controls. As studies were added chronologically, the pooled effect estimate stabilized toward the complete-analysis value of SMD = −1.15 (95% CI: −1.27 to −1.02).

• 

  Figure 12. Summary funnel plots for serum sodium, serum creatinine, eGFR, and serum uric acid. Egger's regression was non-significant for serum sodium (p = 0.961), serum creatinine (p = 0.233), eGFR (p = 0.223), and serum uric acid (p = 0.394). Formal publication-bias assessment was not performed for BUN because only eight studies were available.