Our findings suggest that genetic proxies for ACE inhibitors and CCB have more beneficial associations with kidney function than genetic proxies for other classes of antihypertensive drugs. Kidney dysfunction relates to a higher risk of cerebrovascular disease [
27]; consistently, genetic proxies for CCBs are associated with a lower risk of cerebrovascular disease [
16]. The mechanisms underlying the beneficial associations remain to be explored. Understanding the mechanisms is also helpful when there are varying associations of proxies in different genes, such as in genetic proxies for CCBs and eGFR. In the RCT of ramipril (an ACE inhibitor), its effect on slowing the decline of GFR was not fully explained by lowering blood pressure [
5]. The beneficial associations for ACE inhibitors and CCBs are also possibly through mechanisms beyond lowering blood pressure, such as via the regulation of the renin-angiotensin-aldosterone system (RAAS). RAAS may affect kidney function by several mechanisms, such as via pro-inflammation, endothelial dysfunction, and increasing glomerular capillary pressure [
28]. As such, ACE inhibitors may have a reno-protective role by inhibition of RAAS. However, an RCT of dual blockade of RAAS with ACE inhibitors and ARB did not exhibit more benefits for kidney function than monotherapy [
29], raising the possibility that other pathways might exist. Notably, CKD has an apparent sex disparity, and testosterone has been recognized as a causal factor explaining, or partly explaining, unfavorable kidney function in men [
30,
31]. ACE inhibitors and CCB may also play a role by modulating sex hormones. ACE inhibitors lower free testosterone in men and increase sex hormone-binding globulin in women [
32]. CCBs are also known to have an anti-reproductive effect in men [
33] and to lower testosterone in animal experiments [
34], but the effect remains to be examined in humans.
Despite using MR to minimize confounding and consistency with RCTs [
5‐
7], this study has several limitations. First, MR relies on three assumptions, i.e., the genetic instruments relate to the exposure, are not related to the potential confounders, and the association of the genetic instruments with the outcome is exclusively through the exposure [
35]. To satisfy these assumptions, we used published SNPs related to the expression of genes regulating the relevant antihypertensive target proteins. Moreover, the beneficial associations of genetically proxied ACE inhibition with eGFR and genetic proxies for CCBs with UACR and albuminuria were consistent using different genetic instruments derived from different studies [
11,
12]. Second, weak instruments might bias toward the null; however, the genetic variants we used had
F-statistics > 10. MR estimates are less precise than conventional observational studies, although less prone to confounding, because the genetic variants can only explain a small proportion of the variance in exposure. As such, as previously [
16], we did not conduct an analysis for antihypertensive drugs with only one genetic proxy for the categorical outcome [
16], albuminuria. Although we used by far the largest study for kidney function, i.e., a meta-analysis of the UK Biobank and CKDGen Consortium, the null associations should be interpreted with caution, and we cannot exclude the role of some antihypertensive drugs. Third, most participants in the UK Biobank do not have kidney disease [
36], so the MR estimates from the UK Biobank might be more applicable to the general population than to patients with kidney disease. However, the directions of associations should be consistent across populations. For example, the benefits of ACE inhibitors in kidney function shown in this MR study were also evident in an RCT targeting patients with proteinuric nephropathy [
5]. Fourth, when using the UK Biobank, the estimates might be biased because the genetic predictors for antihypertensives and genetic associations with kidney function were from the same study [
37]. However, the associations using CKDGen were similar, whose participants may not overlap with those in the UK Biobank. Fourth, the associations in Europeans may not apply to other populations, such as Asians. However, causal effects should be consistent across settings, unless the underlying mechanisms and targets vary by setting. Replication in other ancestries will be worthwhile. Fifth, genetic effects might be diluted by compensatory processes or feedback mechanisms [
38]. Such compensation would be expected to mitigate genetic effects, biasing toward the null [
39], but does not explain the associations for specifically genetically proxied ACE inhibitors, CCBs, and BBs. Sixth, the associations are relatively small, which may not be clinically significant, but might be relevant to population health [
40]. Moreover, MR studies assess lifelong effects so the magnitude of effect sizes might not be comparable to short-term effects of taking antihypertensive drugs, so this study is more relevant to assessing the directions of associations than to providing the magnitude of associations. Given the limited evidence from RCTs with generally small sample sizes [
5‐
7], the beneficial associations of genetic proxies for ACE inhibitors and CCBs in comparison with other classes of antihypertensives provide support for the current clinical guidelines on the treatment of hypertension when it occurs with chronic kidney disease. Finally, although men are more vulnerable to kidney dysfunction, we did not conduct sex-specific analysis because we have not identified sex-specific genetic predictors of the effects of antihypertensives.
From the perspective of clinical practice, our findings, together with previous RCTs [
5‐
7], add support to guidelines recommending the use of ACE inhibitors or ACE inhibitor plus CCB in people with hypertension and kidney dysfunction. Our findings also suggest these reno-protective associations are also generalizable to the general population. Genetic proxies for BBs were associated with lower eGFR; however, given the cardiovascular benefit of BBs [
12], this concern should not outweigh its benefits, especially for patients with cardiovascular disease and without other renal comorbidities.