The literature search used the Boolean logic algorithm to obtain subject words and free words, searching databases including CNKI, Wanfang, Weipu, PubMed, Embase, Cochrane Library, and Web of Science for articles published up to 23 July 2023. The main subject words were “hypertension, blood pressure, high, high blood pressure, middle age, aerobic training, static training, combined training, and randomized controlled trial.” The literature screening applied PICOS strategies, which have been widely used in evidence-based medicine or practice. In the current study, the literature inclusion criteria were as follows: (1) subjects were middle-aged and older adults, with an average age of 45–64 years, without cardiovascular disease or other diseases; (2) randomized controlled experiments; (3) the exercise group only carried out the planned aerobic exercise or static stretching exercise, while the control group received no exercise intervention and no other intervention measures; (4) the intervention lasted more than 4 weeks; and (5) the aim of the study was mainly to compare the mean and standard deviation (SD) of resting blood pressure at baseline and at the end of the study between the control and exercise groups. The studies identified from the literature screening search were mainly those on the effect of exercise on blood pressure in middle-aged and older adults. Once those studies were identified, two investigators independently decided whether they should be included in the analysis. If the two authors disagreed, a third author intervened.

Data extraction was completed with a prespecified data extraction form that extracted relevant data and details related to the study's subject characteristics, interventions, and primary outcomes for systematic analysis. However, the usual PEDro score of two depends on how the study blinded the group and the treatment. In general, it is difficult to blind subjects in studies of exercise interventions. Therefore, these two points were excluded from the study analysis, and the PEDro assessment was based on eight terms, each scoring one point. The authors independently extracted and examined the data and assessed the quality of each study using the PEDro score.

Network meta-analysis was performed using Stata 16 software and RevMan 5.0 software. The outcome indicators selected were continuous variables, and the mean difference (MD) and 95% CI were used as effect size indicators. When P < 0.05, the inconsistent model was significant. These data cannot be directly analyzed by the consistency model; if P > 0.05, there was no significant difference between the direct and indirect comparisons, and the consistency model was used for analysis. A cumulative ranking probability map (SUCRA) was used to rank and compare the types of interventions. When SUCRA = 1, intervention measures are highly effective, while when SUCRA = 0, intervention measures are highly ineffective (15). According to the SUCRA values of the SBP and DBP indices, a stratified cluster analysis was used to explore the optimal treatment intervention.

A total of 2,652 relevant articles were screened by a preliminary search of each database, and 19 articles, including 2,385 subjects, were finally included in the preliminary screening of titles and abstracts, as well as the rescreening of the full text. The main flow chart of the literature screening is shown in Figure 1.

The 19 included studies mainly focused on hypertension in middle-aged and older adults. All subjects were grouped according to the principle of randomized control and the basic characteristics of the included studies (see Table 1).

All 19 included studies achieved “random assignment,” ITT intention-to-treat analysis, “statistical analysis between groups,” and “point measurement and variation measurement.” It is difficult to assess the overall quality of the study by blinding the subjects and therapists; therefore, these two points were excluded. The PEDro assessment is based on eight factors, and each is worth 1 point. Most PEDro scores were between five and six points, with an average score of ~6 points, and only one study reached a score of eight points. Overall, the results of the quality assessment of the literature were good (see Table 2).

The inconsistency model test of the 19 RCTs included showed that SBP (P = 0.228) and DBP (P = 0.667) were not significant. The results of the local inconsistency test by the node splitting method again resulted in P > 0.05, and the results of indirect comparison and direct comparison of the two indices were shown to be consistent, indicating that the consistency model should be used for analysis.

A total of 19 studies were included, including 11 studies comparing aerobic exercise with a control group, two studies comparing static exercise with a control group, and six studies comparing aerobic exercise with static exercise with the control group. For the network relationship between SBP and DBP of different exercise interventions, the circle area of aerobic exercise was the largest. The edge between aerobic exercise and conventional exercise was wider, indicating that studies comparing aerobic exercise with conventional exercise appeared most frequently among the included studies (Figure 2). Therefore, the effect of aerobic exercise on the hypertension of middle-aged and older adults is still supported by many studies, thus indicating a strong theoretical basis for its support.

The results of the meta-analysis showed the MD value of aerobic exercise intervention [MD = −9.254, P < 0.05, 95% CI (−14.810, −3.698)] and the MD value of static exercise intervention [MD = −10.465, P < 0.05, 95% CI (−18.135, −18.135); −2.794] were better than those of the control group in improving SBP in hypertension. An indirect comparison between the two exercise modes showed no significant difference in SBP between the two groups (P > 0.05). The results of the SUCRA ranking showed that the effect of static exercise on SBP in hypertension was better than that of aerobic exercise (Figure 3).

The results of the meta-analysis showed that the MD value in the aerobic exercise intervention [MD = −1.4096; P > 0.05, 95% CI (−8.2395, 5.4201)] and the MD value in the static exercise intervention [MD = −4.5206, P > 0.05, 95% CI (−14.0436, 5.0023)] had no significant effect on DBP improvement in hypertension (all P > 0.05). An indirect comparison between the two showed no significant difference in exercise mode between the two groups (P > 0.05). The results of the SUCRA ranking showed that static exercise had a better effect on DBP in hypertensive individuals (see Figure 3).

Compared with the corrected funnel plot, the inverted funnel plot was symmetric, and the scattered points were distributed within the range of the inverted funnel plot (see Figure 4), indicating that there was little possibility of a small-sample effect or publication bias.

Many recent studies have found that exercise modes such as isokinetic resistance training, limb stretching exercises, swimming, and running can reduce blood pressure (12, 35). Some studies have shown that aerobic exercise training can lower blood pressure by 5–7 mmHg (36). In this study, aerobic exercise was also found to have a significant effect on lowering SBP in middle-aged and older adults with hypertension, which can reduce the incidence of hypertension to some extent. Engaging in physical activity can also further prevent cardiopulmonary impairment and lower blood pressure (37). The incidence of hypertension increases with age, and the older the individuals are, the more their exercise ability will be affected, and the exercise mode or intensity will be limited (38). The incidence of many chronic diseases, such as cardiovascular disease and hypertension, further increases with age (39, 40). Many forms of aerobic exercise can lower blood pressure in middle-aged and older adults, such as jogging, square dancing, and aerobics, all of which have good effects. Whelton et al. found that aerobic exercise had an overall net effect of reducing SBP and DBP by 3.84 and 2.58 mmHg, respectively (7). Hagberg et al.'s review showed that in hypertensive adults aged 60 or older, SBP and DBP values were reduced by an average of 7.6/8.8 mmHg (41). In this study, aerobic exercise had a significant effect on SBP but not on DBP. This suggests that in this study, aerobic exercise effectively lowered blood pressure in middle-aged and older adults with hypertension, which has certain clinical significance and value for the health and medical conditions of both middle-aged and older adults.

For middle-aged and older individuals, the most common isometric exercises are static stretching exercises, which are primarily defined as sustained muscle contraction (i.e., increased tension) without changes in the length of the involved muscle groups. To date, relatively little attention has been given to the effects of isometric exercise training on resting blood pressure. The results from previous meta-analysis suggest that although the overall sample size is still small, isometric exercise training may produce greater reductions in resting blood pressure (42, 43). The study results also showed that the improvement effect of isometric exercise on the SBP and DBP of middle-aged and older hypertensive individuals was better than that of aerobic exercise on the SBP and DBP of hypertensive individuals, which is similar to the results of the previous meta-analysis. However, middle-aged and older individuals should not excessively perform high-load isometric exercises, as improper training may put pressure on the heart, potentially leading to other diseases. Since isometric exercise is not traditionally recommended for hypertensive subjects, isometric training is generally only applicable to some specific exercise programs. However, small-scale, short-term studies conducted in subjects with normal blood pressure and hypertensive subjects suggest that performing 3–4 short isometric exercises per week can lower systolic and diastolic blood pressure (44). Isometric exercise also has effects on lowering blood pressure in middle-aged and older adults, but its advantages compared to dynamic exercise training (aerobic exercise) may only be limited to blood pressure, and further investigation is still needed regarding other possible aspects. Additionally, middle-aged and older adults need to perform these exercises within their capacity to avoid injuries to the joints and other body parts.

The results of the included literature showed that different exercise intervention modes significantly lower blood pressure in middle-aged and older adults. Aerobic exercise modes such as slow running, hiking, cycling, and brisk walking, as well as static exercises such as stretching, isometric exercises, and combined exercises with low-intensity dynamic and static loading, all have an impact on lowering hypertension in middle-aged and older adults. Currently, international treatment guidelines for primary and secondary prevention of hypertension in China generally recommend lifestyle changes (quitting smoking, weight loss, exercise training, healthy eating, and reducing sodium intake) as the first-line treatment (45). Non-pharmacological treatments can promote the reduction and improvement of blood pressure in middle-aged and older adults, contributing to their physical and mental health development. In daily life, older adults need to engage in reasonable and healthy exercises to cope with the diseases caused by hypertension and provide a suitable and comfortable living environment.