. 2023 Oct 5;3:131. doi: 10.1038/s43856-023-00359-w

Table 2.

Summary of evidence for treatment effect heterogeneity for SGLT2-inhibitor and GLP1-receptor agonist therapies for cardiovascular outcomes (including heart failure).

	GLP1-RA			SGLT2i			Plain language summary
Cardiovascular disease (CVD)	GRADE EVIDENCE B			GRADE EVIDENCE B
Biomarker	N (observational)	N (RCT)	N (Meta-analysis and pooled RCT)	N (observational)	N (RCT)	N (Meta-analysis and pooled RCT)
Race/ethnicity	0	0	3^76,83,90	0	2^160,161	4^76,78,83,110	No heterogeneity by race/ethnicity for SGLT2is; Potential increased cardiovascular benefit in Asians associated with GLP1-RA use, but results are inconsistent.
History of CVD	3^105–107	8^{125,162–167}	7^{83,90–95,98,99}	4^{104–106,109}	3^86,168,169	5^{2,80,83,95,112}	No consistent impact on SGLT2i or GLP1-RA outcomes.
Age	2^107,108	3^121,165,170	3^83,92,94	2^104,109	1¹⁷¹	4^78–80,83	No consistent heterogeneity by age for SGLT2is; No evidence of age effect for GLP1-RAs.
Sex	3^102,107,108	1¹⁶⁵	5^{83,90,92,94,96,99}	4^{102,104,109,110}	3^171–173	4^{79,83,174,175}	No consistent heterogeneity by sex for SGLT2is or GLP1-RAs.
Renal function	1¹⁰⁸	2^127,176	5^{83,84,90,92,94}	2^104,109	6^{16,117–119,171,177}	5^2,82–84,95	No consistent heterogeneity by renal function on cardiovascular outcomes for SGLT2is; No evidence of heterogeneity by renal function on cardiovascular outcomes for GLP1-RAs.
BMI	0	1¹²⁶	5^{83,90,92,94,97}	1¹¹⁰	2^87,171	3^79,83,97	No consistent heterogeneity by BMI for SGLT2is; Some inconsistent evidence suggests that higher baseline BMI may improve cardiovascular efficacy of GLP1-RAs.
Genetics	0	0	0	0	0	0
Non-routine biomarkers	0	0	0	0	5^89,178–181	0	The greater benefit of SGLT2i in those with high levels of 3 biomarkers: hs Cardiac Troponin T, soluble suppression of tumorigenesis-2 (sST2), and insulin-like growth factor binding protein 7 (IGFBP7) levels
Heart Failure (HF)	GRADE EVIDENCE B			GRADE EVIDENCE B
Ethnicity	0	1¹⁸²	1⁷⁶	0	4^{160,161,183,184}	4^76–79	Possibly greater relative benefit of SGLT2i in Asian and Black compared to white ethnicity; Potential increased efficacy of GLP1-RAs in Asian ethnicity.
Age	0	2^165,170	1⁹⁴	1¹⁰⁴	1¹⁸³	3^79,81	No heterogeneity by age for SGLT2is or GLP1-RAs.
Sex	0	2¹⁸²	1⁹⁴	2^102,104	3^172,173,183	3^79,81,175	No heterogeneity by sex for SGLT2is or GLP1-RAs.
BMI	0	0	1⁹⁴	0	3^87,183,184	2^79,81	No consistent heterogeneity by BMI for SGLT2is or GLP1-RAs.
History of CVD	2^105,106	3^165,166,182	3^93–95	4^{101,104–106}	4^{86,168,169,185}	4^2,77,81,85	No consistent heterogeneity by CVD history for SGLT2is or GLP1-RAs.
History of HF	0	0	0	1¹⁰⁶	1⁸⁸	4^2,77,81,85	No consistent heterogeneity by HF history for SGLT2is; No analysis on heterogeneity by HF history was performed for GLP1-RAs.
HF severity/score	0	0	0	1¹⁰³	3^163,186,187	1⁷⁷	Greater relative benefit of SGLT2i in those with NYHA class II vs class III/IV in one meta-analysis; No analysis of heterogeneity by HF severity/score performed for GLP1-RAs.
Renal function	0	0	1⁹⁴	2^103,104	6^{16,117–119,177,183}	5^{2,77,81,82,85}	No consistent heterogeneity in renal function. A single meta-analysis showed greater SGLT-2 benefit with lower eGFR and higher ACR; No evidence for heterogeneity by renal function for GLP1-RAs.
Genetics	0	0	0	0	0	0
Non-routine biomarkers	0	0	0	0	6^{89,159,178–180,188}	0	No heterogeneity across a variety of non-routine biomarkers. No analysis of heterogeneity by non-routine biomarkers was performed for GLP1-RAs.