Meditation for the primary and secondary prevention of cardiovascular disease

Background Interventions incorporating meditation to address stress, anxiety, and depression, and improve self‐management, are becoming popular for many health conditions. Stress is a risk factor for cardiovascular disease (CVD) and clusters with other modifiable behavioural risk factors, such as smoking. Meditation may therefore be a useful CVD prevention strategy. Objectives To determine the effectiveness of meditation, primarily mindfulness‐based interventions (MBIs) and transcendental meditation (TM), for the primary and secondary prevention of CVD. Search methods We searched CENTRAL, MEDLINE, Embase, three other databases, and two trials registers on 14 November 2021, together with reference checking, citation searching, and contact with study authors to identify additional studies. Selection criteria We included randomised controlled trials (RCTs) of 12 weeks or more in adults at high risk of CVD and those with established CVD. We explored four comparisons: MBIs versus active comparators (alternative interventions); MBIs versus non‐active comparators (no intervention, wait list, usual care); TM versus active comparators; TM versus non‐active comparators. Data collection and analysis We used standard Cochrane methods. Our primary outcomes were CVD clinical events (e.g. cardiovascular mortality), blood pressure, measures of psychological distress and well‐being, and adverse events. Secondary outcomes included other CVD risk factors (e.g. blood lipid levels), quality of life, and coping abilities. We used GRADE to assess the certainty of evidence. Main results We included 81 RCTs (6971 participants), with most studies at unclear risk of bias. MBIs versus active comparators (29 RCTs, 2883 participants) Systolic (SBP) and diastolic (DBP) blood pressure were reported in six trials (388 participants) where heterogeneity was considerable (SBP: MD ‐6.08 mmHg, 95% CI ‐12.79 to 0.63, I2 = 88%; DBP: MD ‐5.18 mmHg, 95% CI ‐10.65 to 0.29, I2 = 91%; both outcomes based on low‐certainty evidence). There was little or no effect of MBIs on anxiety (SMD ‐0.06 units, 95% CI ‐0.25 to 0.13; I2 = 0%; 9 trials, 438 participants; moderate‐certainty evidence), or depression (SMD 0.08 units, 95% CI ‐0.08 to 0.24; I2 = 0%; 11 trials, 595 participants; moderate‐certainty evidence). Perceived stress was reduced with MBIs (SMD ‐0.24 units, 95% CI ‐0.45 to ‐0.03; I2 = 0%; P = 0.03; 6 trials, 357 participants; moderate‐certainty evidence). There was little to no effect on well‐being (SMD ‐0.18 units, 95% CI ‐0.67 to 0.32; 1 trial, 63 participants; low‐certainty evidence). There was little to no effect on smoking cessation (RR 1.45, 95% CI 0.78 to 2.68; I2 = 79%; 6 trials, 1087 participants; low‐certainty evidence). None of the trials reported CVD clinical events or adverse events. MBIs versus non‐active comparators (38 RCTs, 2905 participants) Clinical events were reported in one trial (110 participants), providing very low‐certainty evidence (RR 0.94, 95% CI 0.37 to 2.42). SBP and DBP were reduced in nine trials (379 participants) but heterogeneity was substantial (SBP: MD ‐6.62 mmHg, 95% CI ‐13.15 to ‐0.1, I2 = 87%; DBP: MD ‐3.35 mmHg, 95% CI ‐5.86 to ‐0.85, I2 = 61%; both outcomes based on low‐certainty evidence). There was low‐certainty evidence of reductions in anxiety (SMD ‐0.78 units, 95% CI ‐1.09 to ‐0.41; I2 = 61%; 9 trials, 533 participants; low‐certainty evidence), depression (SMD ‐0.66 units, 95% CI ‐0.91 to ‐0.41; I2 = 67%; 15 trials, 912 participants; low‐certainty evidence) and perceived stress (SMD ‐0.59 units, 95% CI ‐0.89 to ‐0.29; I2 = 70%; 11 trials, 708 participants; low‐certainty evidence) but heterogeneity was substantial. Well‐being increased (SMD 0.5 units, 95% CI 0.09 to 0.91; I2 = 47%; 2 trials, 198 participants; moderate‐certainty evidence). There was little to no effect on smoking cessation (RR 1.36, 95% CI 0.86 to 2.13; I2 = 0%; 2 trials, 453 participants; low‐certainty evidence). One small study (18 participants) reported two adverse events in the MBI group, which were not regarded as serious by the study investigators (RR 5.0, 95% CI 0.27 to 91.52; low‐certainty evidence). No subgroup effects were seen for SBP, DBP, anxiety, depression, or perceived stress by primary and secondary prevention. TM versus active comparators (8 RCTs, 830 participants) Clinical events were reported in one trial (201 participants) based on low‐certainty evidence (RR 0.91, 95% CI 0.56 to 1.49). SBP was reduced (MD ‐2.33 mmHg, 95% CI ‐3.99 to ‐0.68; I2 = 2%; 8 trials, 774 participants; moderate‐certainty evidence), with an uncertain effect on DBP (MD ‐1.15 mmHg, 95% CI ‐2.85 to 0.55; I2 = 53%; low‐certainty evidence). There was little or no effect on anxiety (SMD 0.06 units, 95% CI ‐0.22 to 0.33; I2 = 0%; 3 trials, 200 participants; low‐certainty evidence), depression (SMD ‐0.12 units, 95% CI ‐0.31 to 0.07; I2 = 0%; 5 trials, 421 participants; moderate‐certainty evidence), or perceived stress (SMD 0.04 units, 95% CI ‐0.49 to 0.57; I2 = 70%; 3 trials, 194 participants; very low‐certainty evidence). None of the trials reported adverse events or smoking rates. No subgroup effects were seen for SBP or DBP by primary and secondary prevention. TM versus non‐active comparators (2 RCTs, 186 participants) Two trials (139 participants) reported blood pressure, where reductions were seen in SBP (MD ‐6.34 mmHg, 95% CI ‐9.86 to ‐2.81; I2 = 0%; low‐certainty evidence) and DBP (MD ‐5.13 mmHg, 95% CI ‐9.07 to ‐1.19; I2 = 18%; very low‐certainty evidence). One trial (112 participants) reported anxiety and depression and found reductions in both (anxiety SMD ‐0.71 units, 95% CI ‐1.09 to ‐0.32; depression SMD ‐0.48 units, 95% CI ‐0.86 to ‐0.11; low‐certainty evidence). None of the trials reported CVD clinical events, adverse events, or smoking rates. Authors' conclusions Despite the large number of studies included in the review, heterogeneity was substantial for many of the outcomes, which reduced the certainty of our findings. We attempted to address this by presenting four main comparisons of MBIs or TM versus active or inactive comparators, and by subgroup analyses according to primary or secondary prevention, where there were sufficient studies. The majority of studies were small and there was unclear risk of bias for most domains. Overall, we found very little information on the effects of meditation on CVD clinical endpoints, and limited information on blood pressure and psychological outcomes, for people at risk of or with established CVD. This is a very active area of research as shown by the large number of ongoing studies, with some having been completed at the time of writing this review. The status of all ongoing studies will be formally assessed and incorporated in further updates.