肾功能损害与年龄相关性黄斑变性:系统综述与荟萃分析
慢性肾脏病(CKD)与AMD风险升高相关(OR=1.27,95%CI 1.07-1.50;I²=95.1%),非渗出性AMD关联显著,透析≥90天者关联更强(OR=1.74)。孟德尔随机化支持低eGFR与AMD的因果关联(汇总OR=1.65)。
要点
- CKD与AMD风险升高相关(OR=1.27),但异质性极高(I²=95.1%)
- 非渗出性AMD关联显著(OR=1.46),渗出性AMD不精确
- eGFR<60 mL/min/1.73 m²及透析≥90天者AMD风险更高
核心结论
- 肾功能损害尤其是eGFR降低、血肌酐升高和透析暴露与AMD风险轻度升高相关,非渗出性AMD关联最一致。
- 连续肾功能指标的生物学梯度及孟德尔随机化证据支持眼-肾关联,但因果推断尚不成熟。
研究设计与样本
- 系统检索MEDLINE、Embase、Scopus和Web of Science至2025年8月11日。
- 纳入评估CKD/肾功能指标与AMD关联的人体观察性研究和孟德尔随机化分析。
- 采用随机效应模型(REML及Hartung-Knapp校正)合并OR及95%CI,GRADE框架评级证据确定性。
主要发现
- CKD与AMD风险升高相关(OR=1.27,95%CI 1.07-1.50;I²=95.1%)。
- 按表型:非渗出性AMD关联显著(OR=1.46,1.06-1.99),渗出性AMD不精确(OR=1.71,0.72-4.08),早期AMD无统计学意义(OR=1.32,0.90-1.93)。
- CKD严重程度:eGFR<60 mL/min/1.73 m²者OR=1.40(1.15-1.72),透析≥90天者OR=1.74(1.52-2.00)。
- 连续肾功能指标:eGFR每降低1个SD,OR=1.30(1.11-1.52);孟德尔随机化汇总OR=1.65(1.54-1.76)。
关键数值与推荐
- 整体任何AMD、非渗出性AMD、透析暴露及孟德尔随机化证据的GRADE确定性为中等。
- 多数亚组因异质性和不精确性,证据确定性为低至极低。
临床怎么用
- 对肾功能损害患者尤其是晚期CKD患者,可考虑加强AMD风险监测。
- 支持肾内科与眼科协作,共同管理血管危险因素。
- 鉴于异质性极高,合并估计应视为多样证据的中心趋势而非个体化精确风险。
局限性
- 异质性极高(I²=95.1%),合并估计的精确性受限。
- 观察性证据无法完全确认因果关联,因果推断尚不成熟。
- 多数亚组证据确定性为低至极低。
- 需统一CKD和AMD定义的前瞻性研究确认关联及方向。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
TOPIC: Chronic kidney disease (CKD) and age-related macular degeneration (AMD) share vascular and inflammatory pathways. Clarifying their association could inform risk stratification across nephrology and ophthalmology. CLINICAL RELEVANCE: In adults with renal impairment, especially those with more advanced kidney dysfunction, closer AMD risk surveillance may be warranted, supporting integrated care between ophthalmology and nephrology even though causality cannot be fully confirmed from observational evidence alone. METHODS: We systematically searched MEDLINE, Embase, Scopus, and Web of Science from inception to 11 August 2025. Eligible human observational studies and Mendelian randomization (MR) analyses assessed CKD/renal metrics (eGFR, albuminuria/proteinuria, dialysis) in relation to AMD. Random-effects meta-analyses used REML with Hartung-Knapp adjustments, and pooled evidence using odds ratios (ORs) and 95% confidence intervals (95% CI). Heterogeneity was quantified using I², subgroup analyses explored design, phenotype, and CKD severity, and small-study effects were assessed using funnel plots and Egger's test when appropriate. Certainty of evidence was rated using the GRADE prognostic-factor framework. RESULTS: Nineteen studies contributed to the primary meta-analysis. CKD was associated with higher odds of AMD (OR = 1.27, 95% CI: 1.07-1.50; I² = 95.1%). Effects were similar by design (P for subgroup difference = .84): cohorts OR = 1.29 (0.98-1.69) and cross-sectional studies OR = 1.25 (0.96-1.62). By phenotype, associations were significant for nonexudative AMD (OR = 1.46, 1.06-1.99), while exudative AMD showed an imprecise increase (OR = 1.71, 0.72-4.08), and early AMD was not statistically significant (OR = 1.32, 0.90-1.93). CKD severity analyses suggested significantly higher odds with eGFR <60 mL/min/1.73 m² (OR = 1.40, 1.15-1.72), and a stronger association in dialysis ≥90 days (OR = 1.74, 1.52-2.00). Continuous renal markers aligned with a biological gradient (per SD decrease in eGFR OR = 1.30, 1.11-1.52). Two Mendelian randomization studies supported a causal association between lower eGFR and AMD (pooled OR = 1.65, 1.54-1.76). Overall certainty ranged from moderate (any AMD overall, nonexudative AMD, dialysis exposure, MR) to low/very low for most subgroups due to heterogeneity and imprecision. CONCLUSIONS: Renal impairment, particularly reduced eGFR (<60 mL/min/1.73 m²), elevated serum creatinine, and dialysis exposure, was associated with modestly higher odds of AMD, and the association was most consistent for nonexudative disease. A biological gradient across continuous renal markers and concordant Mendelian randomization evidence support a plausible eye-kidney link, but the very high heterogeneity (I² = 95.1%) means the pooled estimate should be read as a central tendency of a diverse evidence base rather than a precise, individually applicable risk. Causal inference therefore remains premature. In practice, these findings support integrated ocular surveillance in patients with CKD, shared vascular risk reduction across nephrology and ophthalmology, and standardized prospective studies with harmonized CKD and AMD definitions to confirm the association and clarify its direction.
引用
- Alamoudi A, Alnabihi A, Almufarriji N, et al. Renal Impairment and Age-Related Macular Degeneration: A Comprehensive Systematic Review and Meta-Analysis. Am J Ophthalmol. 2026. PMID: 42501958. doi:10.1016/j.ajo.2026.07.046
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。