Effect of weight loss on blood pressure changes in overweight patients: A systematic review and meta‐analysis

Abstract

To determine quantitative differences between weight loss and changes in clinic blood pressure (BP) and ambulatory BP in patients with obesity or overweight, the authors performed a meta‐analysis. PubMed, Embase, and Scopus databases were searched up to June 2022. Studies that compared clinic or ambulatory BP with weight loss were included. A random effect model was applied to pool the differences between clinic BP and ambulatory BP. Thirty‐five studies, for a total of 3219 patients were included in this meta‐analysis. The clinic systolic blood pressure (SBP) and diastolic blood pressure (DBP) were significantly reduced by 5.79 mmHg (95% CI, 3.54–8.05) and 3.36 mmHg (95% CI, 1.93–4.75) after a mean body mass index (BMI) reduction of 2.27 kg/m², and the SBP and DBP were significantly reduced by 6.65 mmHg (95% CI, 5.16–8.14) and 3.63 mmHg (95% CI, 2.03–5.24) after a mean BMI reduction of 4.12 kg/m². The BP reductions were much larger in patients with a BMI decrease ≥3 kg/m² than in patients with less BMI decrease, both for clinic SBP [8.54 mmHg (95% CI, 4.62–12.47)] versus [3.83 mmHg (95% CI, 1.22–6.45)] and clinic DBP [3.45 mmHg (95% CI, 1.59–5.30)] versus [3.15 mmHg (95% CI, 1.21–5.10)]. The significant reduction of the clinic and ambulatory BP followed the weight loss, and this phenomenon could be more notable after medical intervention and a larger weight loss.

1. INTRODUCTION

Obesity is regarded as one of the most prevalent diseases in the world. In 2017, the Global Burden of Disease Obesity Collaborators revealed that more than 603.7 million adult individuals were obese and high BMI accounted for 4.0 million deaths worldwide. ¹ Not just in Western countries, where obesity rates are known to be high, but in some Asian countries like China, research estimated that in China the prevalence of overweight (BMI 24.0–28.0 kg/m²) and obesity (BMI ≥28.0 kg/m²) might reach 65·3% in adults, ² which will have rates similar to those in the USA over the next decade with the recent rising trend in overweight. ³ Obesity is a risk factor for hypertension and several cardiovascular diseases, ⁴ and the Global Burden of Disease Obesity Collaborators also estimated that more than two‐thirds of deaths related to high BMI were due to cardiovascular disease. ¹

There is a general agreement that weight reduction is recommended for overweight and obese hypertensive patients, ⁵ a meta‐analysis suggested that after 6–12 months, weight loss reduced systolic blood pressure (SBP) compared to placebo by mean difference (MD) −2.6 mmHg (95% confidence interval (CI) −3.8 to −1.4 mmHg; four trials, 2058 participants) and diastolic blood pressure (DBP) by MD −2.0 mmHg (95% CI −2.7 to −1.2 mmHg; four trials, 2058 participants), ⁶ maintenance of healthy body weight (BMI of approximately 20–25 kg/m² in people <60 years of age; higher in older patients), and waist circumference (<94 cm for men and <80 cm for women) is recommended for nonhypertensive individuals to prevent hypertension, and for patients with hypertension to reduce BP. ⁷

Previous systematic reviews and meta‐analyses have explored the relationship between weight reduction and blood pressure (BP), which have usually examined the clinic BP, but they have neglected the out‐of‐office measurements. Lots of Hypertension Guidelines suggest that ambulatory BP monitoring and/or home BP monitoring are recommended for the diagnosis of hypertension. ² , ⁵ , ⁷ Although ambulatory BP measurement is not as convenient as office BP to reproduce, it can provide the average BP readings over a defined period and is a better predictor of hypertension than office BP. ⁸ In terms of clinical application, office, and ambulatory BP are complementary, we aimed to quantify the effect of weight reduction on BP comprehensively and precisely and to find out what else constructive suggestions can apply to practical applications.

2. METHODS

2.1. Search strategy

A systematic literature search was conducted to identify relevant studies in the PubMed, Embase, and Scopus databases up to June 2022. Only English‐language studies were included. The specific keywords and search strategies were presented in the Supplementary Appendix. In addition, we also checked the reference lists of included studies to identify relevant studies.

2.2. Selection of studies

Two reviewers (SJ, Y and ZY, Z) assessed the eligibility of studies by screening the title, abstract, and even full text of them independently, and the disagreements were settled through discussion. Studies were considered for inclusion if they met the following criteria: (1) included the comparison of clinic or ambulatory BP between weight loss; (2) the mean values and standard deviations (SD) of clinic or ambulatory BP and BMI or weight were reported respectively; (3) all patients have a BMI ≥ 24 kg/m². And the studies were excluded if they met any of the following criteria: (1) incomplete reporting data; (2) unpublished data or conference data; (3) participants who were pregnant. The certainty of evidence was used in the GRADE (grading of recommendations, assessment, development, and evaluation) approach. All statistical analyses were performed using Revman 5.

2.3. Data extraction and collection

After identifying relevant articles, two reviewers (SJ, Y and ZY, Z) extracted the data independently and the disagreements were resolved through discussion. The following data were extracted: study characteristics (authors, year of publication, journal, country/region, study design), baseline information of participants (sample size, mean age, sex, hypertensive status, weight, and BMI), BP measurement (methods and devices of BP measurement contain at clinic or mean 24‐h ambulatory), mean values and SD of clinic and ambulatory BP measurement between weight loss before and after.

2.4. Quality assessment

Risk of bias assessment was carried out independently by two authors (SJ, Y and ZY, Z) utilizing the revised Cochrane risk of bias tool 2 for randomized trials (RoB2 tool). ⁹ Any disagreements were solved by consensus. We used the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system to evaluate the quality of evidence. ¹⁰ , ¹¹

2.5. Statistical analysis

Continuous variables were presented as mean ± SD, and the categorical variables were presented as proportions. We analyzed the differences of clinic BP and ambulatory BP before and after weight loss. A random‐effect model was used and the results were reported as MDs of BP values. Heterogeneity was estimated by a Q test (p < .1) and I2 statistic, with I2 values of 25, 50, and 75% representing mild, moderate, and severe heterogeneity, respectively. Besides, we also performed a subgroup analysis by the percent of BMI reduction, initial BMI values, intervention modalities, sex, and the percent of weight reduction. All statistical analyses were performed using Revman 5 (The Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen, Denmark) and Stata 12 (StataCorp, College Station, Texas).

3. RESULT

3.1. Study selection and characteristics

Nine hundred and forty‐two records were identified through our searching strategy from PubMed, Embase, and Scopus databases (Figure 1). After the removal of 221 duplicates, 217 studies were excluded by title screening and 433 studies were further excluded by abstract screening. Six records were added from reference lists of included studies. A total of 77 studies were eligible for full‐text screening and 42 articles were excluded at this stage for the following reasons: neither clinic nor ambulatory average BP values were included (n = 10); conference data (n = 9); neither BMI nor weight values (n = 4); missing data (n = 18); review (n = 1); and duplicate data (n = 1). In total, 35 studies were included for the meta‐analysis. Among the included studies, six studies contained both clinic and ambulatory BP values, ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ whereas 13 studies contained only clinic BP values, ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ 16 studies contained only ambulatory BP values. ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ A total of 3219 patients were examined, and study populations varied from general group to populations with hypertension and the average weight‐loss time was 7.7 months. The detailed characteristics of included studies were summarized in Table 1.

FIGURE 1.

Flow chart of the study selection.

TABLE 1.

Characteristics of the included studies.

Study	Journal	Sample size	Mean age	Male (%)	Hypertension (%)	Weight (pre)	BMI (pre)	Clinic SBP (pre)	Clinic DBP (pre)	Ambulatory SBP (pre)	Ambulatory DBP (pre)	Weight loss method	Weight loss time (month)
Adachi, H. et al. 2015	Lipids in Health and Disease	101	60.1 ± 11.8	46.5	n.a.	n.a.	27.9 ± 3.5	128.7 ± 14.8	76.0 ± 9.4	n.a.	n.a.	Ezetimibe	6
Amador, N. et al. 2005	Journal of Human Hypertension	25	42.4 ± 1.0	100	n.a.	n.a.	31.8 ± 1.4	131.0 ± 2.9	84.0 ± 2.0	n.a.	n.a.	Losartan	4
Azushima, K. et al. 2015	Atherosclerosis	54	59.2 ± 14.5	51.9	100	82.5 ± 16.4	31.3 ± 5.0	142 ± 19	85 ± 12	140 ± 13	84 ± 10	Bofu‐tsusho‐san	6
Bao, D. Q. et al. 1998	Hypertension	14	53.1 ± 1.9	57.1	100	89.2 ± 4.2	30.9 ± 1.1	132.1 ± 2.9	74.9 ± 1.8	133.1 ± 3.5	76.3 ± 1.9	Fish diet	4
Berney, M. et al. 2021	Front Endocrinol (Lausanne)	25	42 ± 10	28	n.a.	121 ± 19	43 ± 5.3	120 ± 3	79 ± 2	120 ± 8	76 ± 6	RYGB	3
Boerboom, A. et al. 2019	Surgery for Obesity and Related Diseases	72	47.0 ± 7.0	43	29	n.a.	47.0 ± 6.0	146.0 ± 16.0	93.0 ± 8.0	n.a.	n.a.	Standard RYGB	24
Boozer, C. N. et al. 2002	Int J Obes Relat Metab Disord	83	44.5 ± 12.4	22	n.a.	88.1 ± 14.8	31.7 ± 4.0	n.a.	n.a.	119 ± 11	77 ± 8	Herbal	6
Brader, L. et al. 2014	Eur J Clin Nutr	18	54.4 ± 7.5	28	44	90.9 ± 15.5	30.9 ± 3.4	n.a.	n.a.	126.4 ± 9.5	76.8 ± 8.4	Nordic diet	3
Burke, V. et al. 2005	J Hypertens	123	57.1 ± 7.2	45.5	100	86.7 ± 1.2	30.4 ± 2.9	n.a.	n.a.	128 ± 1	77 ± 1	Diet and exercise	4
Careaga, M. et al. 2016	Surg Obes Relat Dis	42	n.a.	n.a.	33.3	n.a.	46.0 ± 5.6	n.a.	n.a.	128 ± 12.5	73 ± 7.8	RYGB	12
Chen, J. L. et al. 2015	J Pediatr Nurs	70	n.a.	n.a.	n.a.	n.a.	24.0 ± 3.5	104.5 ± 8.8	61.9 ± 8.7	n.a.	n.a.	Healthy weight management	2
Conwell, L. S. et al. 2010	Br J Sports Med	15	11.8 ± 0.6	40	n.a.	81.1 ± 4.0	34.5 ± 1.3	120.1 ± 2.1	66.6 ± 2.0	n.a.	n.a.	Physical activity	2.5
Cornelissen, V. A. et al. 2009	J Hypertens	38	n.a.	n.a.	n.a.	75.2 ± 2.8	26.1 ± 0.7	n.a.	n.a.	126.8 ± 1.7	78.7 ± 1.3	High intensity exercise	2.5
Cox, K. L. et al. 1996	J Hypertens	14	41.4 ± 3.5	n.a.	n.a.	97.7 ± 9.1	30.9 ± 2.5	137.7 ± 1.6	86.7 ± 1.3	130.3 ± 3.9	76.6 ± 3.0	Low caloric intake, light exercise	1
Czupryniak, L. et al. 2005	Am J Hypertens	8	35 ± 9.1	37.5	n.a.	129 ± 21.9	44.1 ± 4.7	n.a.	n.a.	154.7 ± 12.3	105.6 ± 8.1	Gastric bypass surgery	2
Daigle, C. R. et al. 2016	Surgery for Obesity and Related Diseases	121	53.5 ± 10.8	13	n.a.	n.a.	47.5 ± 7.5	138.6 ± 17.7	n.a.	n.a.	n.a.	Revisional bariatric surgery	n.a.
Engeli, S. et al. 2005	Hypertension	38	57 ± 3.6	0	n.a.	n.a.	33.1 ± 4.6	n.a.	n.a.	138 ± 12	82 ± 6	Diet and exercise	3.25
Ersoz, H. O. et al. 2004	Int J Obes Relat Metab Disord	28	n.a.	n.a.	0	98.4 ± 15.0	39.3 ± 6.8	n.a.	n.a.	118.1 ± 9.3	73.1 ± 6.2	Metoprolol	3
Faria, A. N. et al. 2004	Arq Bras Cardiol	43	n.a.	n.a.	100	n.a.	40.1 ± 5.5	150.2 ± 18.2	91.0 ± 12.3	n.a.	n.a.	Sibutramine	6
Fortmann, S. P. et al. 1988	Am J Cardiol	42	44 ± 8	n.a.	0	94 ± 8	28 ± 3.7	122 ± 13	78 ± 8	128 ± 10	85 ± 7	Exercise	12
Goulis, D. G. et al. 2004	Int J Obes Relat Metab Disord	45	43.6 ± 12.8	n.a.	n.a.	101.6 ± 22.4	37.6 ± 8.0	134.6 ± 18.2	80.7 ± 11.7	n.a.	n.a.	Home‐centered care	6
Gran, B. et al. 1991	Scand J Prim Health Care	122	n.a.	36.1	n.a.	n.a.	27.0 ± 4.5	150.6 ± 13.4	95.2 ± 5.8	n.a.	n.a.	Physical exercise and diet	24
Gutierrez‐Blanco, D. et al. 2019	Surgery for Obesity and Related Diseases	1330	52.7 ± 5.3	31.1	n.a.	n.a.	43.1 ± 6.9	130.7 ± 14.2	73.0 ± 13.3	n.a.	n.a.	Bariatric surgery	12
Hvidt, K. N. et al. 2014	J Hypertens	61	12.5 ± 1.2	n.a.	n.a.	66.9 ± 10.6	27.2 ± 3.0	110.8 ± 9.2	62.1 ± 5.9	121.2 ± 7.2	70.7 ± 4.8	Lifestyle intervention	12
Ishøy, P. L. et al. 2017	Diabetes Obes Metab	20	37.4 ± 10.7	55	n.a.	118.3 ± 16.0	39.5 ± 3.5	n.a.	n.a.	122.4 ± 14.6	83.7 ± 12.4	Schizophrenia	3
Madero, M. et al. 2015	J Am Soc Hypertens	34	44 ± 9	64.7	n.a.	90 ± 16	30.7 ± 3.6	124 ± 13	84 ± 8	132 ± 12	80 ± 7	Low fructose diet	1
Miller, E. R. et al. 2002	Hypertension	22	53 ± 11	43	100	92.0 ± 14.6	32.8 ± 5.4	134.9 ± 8.9	83.8 ± 5.5	135.3 ± 11.1	83.6 ± 8.6	Lifestyle intervention	2.25
Mori, T. A. et al. 2004	J Hypertens	14	53.1 ± 1.9	57.1	100	89.2 ± 4.2	30.9 ± 1.1	n.a.	n.a.	133.1 ± 3.5	76.3 ± 1.9	Fish diet and lifestyle intervention	4
Nordstrand, N. et al. 2012	Surgery	49	44.4 ± 9.5	67.3	86	134 ± 20	45.5 ± 5.5	n.a.	n.a.	150 ± 18	83 ± 10	RYGB	12
Pekkarinen, T. et al. 1998	Int J Obes Relat Metab Disord	29	40.3 ± 8.3	0	n.a.	96.1 ± 10.9	36.0 ± 2.6	129.6 ± 15.1	83.7 ± 9.1	122.4 ± 12.5	72.9 ± 7.9	VLCD + diet	4.25
Randall, O. S. et al. 2005	J Clin Hypertens (Greenwich)	215	46.7 ± 10.7	23	66	n.a.	42.5 ± 7.5	n.a.	n.a.	129.2 ± 12.8	78.8 ± 9.6	Diet	3
Scherrer, U. et al. 1991	Circulation	19	48 ± 14	52.6	n.a.	86.2 ± 11.2	30.3 ± 3.0	142 ± 14	97 ± 10	n.a.	n.a.	Hypocaloric diet	10
Schiavon, C. A. et al. 2018	Circulation	50	43.1 ± 9.2	18	100	72.7 ± 12.4	26.8 ± 3.7	123.6 ± 13.4	77.6 ± 9.4	122.8 ± 12.9	78.2 ± 11.9	Gastric bypass surgery	12
Schillaci, G. et al. 2003	Am J Hypertens	106	47 ± 11	68	100	82 ± 11	28.2 ± 1.8	145 ± 13	94 ± 6	128 ± 10	82 ± 7	n.a.	48
Shrewsbury, V. A. et al. 2011	BMC Pediatr	129	n.a.	48	n.a.	83.4 ± 14.6	30.9 ± 3.9	119 ± 13	60 ± 9	n.a.	n.a.	Exercise	2

Almost all the included studies had different degrees of bias due to the lack of clarity in methods (Table S2). Sixteen studies have not illustrated whether they enrolled patients consecutively or randomly. Besides, the timing and blinding information of clinic and ambulatory BP measurements were also poorly reported. In terms of concerns regarding applicability, several studies were unclear or high risk of bias in index text and reference standard domains since the absence of detailed descriptions of measurements or nonstandard measurements.

3.2. Comparison between clinic SBP and DBP

Nineteen studies, including a total of 2428 patients, compared clinic SBP before and after weight loss and eighteen studies included 2307 patients compared clinic DBP. The clinic SBP and DBP were significantly reduced by 5.79 mmHg (95% CI, 3.54–8.05) (Figure 2) and 3.36 mmHg (95% CI, 1.93–4.75) (Figure 3) after a mean BMI reduction of 2.27 kg/m². However, there were significant statistical heterogeneities among included studies (I2 = 91%, p<.01 for SBP; I ² = 88%, p<.01 for DBP).

FIGURE 2.

Forest plot of studies that comparing clinic SBP difference between pre weight loss and post weight loss. SBP, systolic blood pressure.

FIGURE 3.

Forest plot of studies that comparing clinic DBP difference between pre weight loss and post weight loss. DBP, diastolic blood pressure.

To further determine whether clinic BP has an influence on the differences between clinic SBP and DBP, subgroup analysis by the mean levels of clinic BP was conducted (Figure S1–S12, see Supplementary Appendix). We found that clinic SBP and DBP in subgroup, which males more than 50% tended to be descend greater than males less than those 50% with only 13 studies included.

Finally, we examined the publication bias by Begg's funnel plots, which indicating no evidence of bias. Further, Begg's test (p>.05) and Egger's test (p>.05) also proved it.

3.3. Comparison between ambulatory SBP and DBP

Twenty‐two studies, including a total of 1072 patients, compared ambulatory SBP and DBP with weight loss. The mean ambulatory SBP and DBP values were significantly reduced by 6.65 mmHg (95% CI, 5.16–8.14) (Figure 4) and 3.63 mmHg (95% CI, 2.03–5.24) (Figure 5). Studies including daytime and night‐time BP indicate that the mean daytime and night‐time ambulatory SBP values were reduced by 4.43 and 4.40 mmHg, meanwhile, the mean daytime and night‐time ambulatory DBP values were reduced by 3.70 and 1.91 mmHg. The statistical heterogeneities were also significant (I ² = 90%, p<.01 for SBP; I ² = 98%, p<.01 for DBP).

FIGURE 4.

Forest plot of studies that comparing ambulatory SBP difference between pre weight loss and post weight loss. SBP, systolic blood pressure.

FIGURE 5.

Forest plot of studies that comparing ambulatory DBP difference between pre weight loss and post weight loss. DBP, diastolic blood pressure.

To further explore the influence of ambulatory BP, subgroup analysis of the mean levels of ambulatory BP was conducted (Figures S13–S24). Similarly, there is a significant tendency in the differences between males more than 50% and males less than 50%. Eight studies, including a total of 409 patients, compared ambulatory SBP before and after weight loss and 12 studies that included 620 patients compared ambulatory DBP showed that medical interventions were more efficient than lifestyle interventions.

All of Begg's funnel plots, Begg's test (p>.05) and Egger's test (p>.05) showed no evidence of publication bias.

3.4. Risk of bias assessment and quality of evidence

The results of the risk of bias assessments are shown in Table S1. We found a low risk of bias in several domains, such as randomization, deviations from the intended interventions, or measurement of the outcome. “Methods for randomization” were generally reported in detail with one exception. “Blinding of participants” was impossible due to the interventions, but the assessment of the outcomes is not likely to be influenced by such knowledge. In the “Missing outcome data” category of the RoB2 tool, some concerns were identified. However, it is not likely that these minor concerns could influence the results of the meta‐analyses. In the “Selection of the reported results” category, some concerns were also found because the published protocol did not include sufficiently detailed data reporting and statistical analysis methods.

The overall quality of the evidence of the analyzed data, based on the GRADE approach, is low to moderate, which is summarized in Table S2. We had to downgrade the quality of evidence for all outcomes because of the substantial and considerable heterogeneity of the results. For two outcomes (before weight loss vs. after weight loss), indirectness was identified because the mean BMI and BP values of the baseline population of several studies differed from those of all others. Serious imprecision was indicated by the wide CIs of various results.

4. DISCUSSION

The main findings of the present meta‐analysis are that: (1) the mean clinic and ambulatory SBP differences between weight loss were 5.79 mmHg (95% CI, 3.54–8.05) and 6.65 mmHg (95% CI, 5.16–8.14), respectively, which were much larger in subgroups with males more than 50% males than in subgroups with less than 50% males; (2) the MDs between clinic and ambulatory SBP in patients with medical interventions were 7.72 mmHg (95% CI, 4.53–10.91) and 8.34 mmHg (95% CI, 3.26–13.43), which were larger than patients with lifestyle interventions, and the BP reductions were correlated with initial weight; (3) in studies with both clinic BP and ambulatory BP, clinic BP decreased more than ambulatory BP, whether systolic or diastolic. To our knowledge, the present study is the first meta‐analysis that compares the changes in BMI, clinic BP, and ambulatory BP before and after weight loss.

Based on office BP, the global prevalence of hypertension was estimated to be 1.13 billion in 2015. ⁵ Hypertension is a major risk factor for stroke, atherosclerosis, and other cardiovascular diseases, and obesity is a major risk factor for hypertension. ⁴⁷ On the other hand, patients with hypertension are also frequently overweight or obese presenting a vicious circle, in which weight gain progressively increases as a consequence of hypertension and it also further aggravates the severity of the BP. Obesity is a known independent risk factor for hypertension. ⁴⁸ Excessive weight gain is associated with hypertension. However, the magnitude and the direction of the association tend to differ on the level of economic development, sex, and race. ⁴⁷ , ⁴⁹ Women tend to have a higher proportion of body fat stored in subcutaneous rather than visceral adipose tissue because of the differences in body composition between men and women. At the same BMI, women will tend to have a considerably higher percentage of body fat than men, ⁵⁰ that may explain why the sex subtype has different outcomes. A meta‐analysis showed that the mean SBP and DBP reductions were associated with an average weight loss, ⁵¹ but weight ignores the effects of height, waistline, and body fat percentage, BMI has a more accurate predictive effect.

Ambulatory BP monitoring provides the average of BP readings over a defined period and offers several advantages, such as stronger prognostic evidence and measurements in real‐life settings, ⁵ numerous studies have used ambulatory BP to improve the accuracy of BP measurements. Our results are especially important because we cite ambulatory BP data from related studies and compare them with the clinic BP data. This approach makes the results more credible. In our study, clinic BP decreased more than ambulatory BP, whether for systolic or diastolic readings. The greater fall in office BP in patients with severely elevated pretreatment BP is caused to a large extent by the reduction of the white coat effect. Because 24‐h ABP is void of white coat and the placebo effect, the changes in ABP are smaller and the pretreatment BP level lower, ⁵² , ⁵³ , ⁵⁴ thereby explaining at least in part why changes in office BP and ambulatory BP are not related to each other in a 1:1 fashion.

There are many things in common between the treatment of obesity and hypertension, such as lifestyle and diet changes. ⁵ A randomized controlled trial in high‐risk individuals on the Mediterranean diet over 5 years showed a 29% cardiovascular risk reduction and a 39% reduction in stroke compared with a low‐fat control diet. ⁵⁵ The Mediterranean diet also significantly reduced ambulatory BP, blood glucose, and lipid levels. ⁵⁶ The clinic DBP reductions in patients with medical interventions [2.97 mmHg (95% CI, 0.13–5.81)] were lower than patients with lifestyle interventions [3.57 mmHg (95% CI, 1.66–5.48)] in this study, showed that lifestyle intervention may have potential benefits over medical intervention.

In addition, our study found that compared with the antihypertensive drug‐treated subgroup, the clinic BP in the nonantihypertensive drug‐treated subgroup decreased significantly, but the ambulatory BP had the opposite result. We think there are several reasons that can explain this result: First, the antihypertensive drug‐treated subgroup had a higher percentage of resistant hypertension, which BP was more difficult to control and less associated with weight, and the nonantihypertensive drug‐treated subgroup contained with low‐to‐mild hypertension, which can control just by lifestyle intervention. ⁵ Second, compared with resistant hypertensive patients, the nonantihypertensive drug‐treated subgroup is more likely to be nervous, thus resulting in higher BP than usual, ambulatory BP monitoring is void of white coat and the placebo effect, which can make the result more reliable. ⁵⁴ Last, the mean age between the antihypertensive drug‐treated subgroup and the nonantihypertensive drug‐treated subgroup in clinic BP and ambulatory BP is also different, in clinic BP, the mean age in the antihypertensive drug‐treated subgroup (42.75) is similar to that of the nonantihypertensive drug‐treated subgroup (44.58), but in ambulatory BP, the former (51.21) is much higher than the latter (44.11), these results suggest that age may also be an important influencing factor.

4.1. Clinical implications and recommendations for future research

As discussed above, the weight and BMI reduction might be different not only in specific BP values but also in clinical implications. However, in clinical practice, some hypertension patients merely accepted antihypertension medication without lifestyle changes due to insufficient patient education or inertia. Thus, an important issue that should be considered in clinical practice is the interventions of hypertension for medication and controlling obesity. Given the differences between an initial BMI of more than 30 kg/m² and less than 30 kg/m² in the comparison of BP before and after weight loss, the antihypertensive effect in mildly obese patients may be less affected by obesity. Compared with Asia, clinic BP, and ambulatory BP decreased more in the European population, which explained that baseline weight is important for BP control. In the discussion above, age could not influence the weight loss in BP change which means old people who are obese also need weight control to decrease BP. To determine the proper management to control hypertension, weight loss in obese and overweight patients should be an important process, further studies will be needed to investigate the differences between body fat percentage with BP. Besides, further prospective randomized trials are also needed to explore the relationship between different target values of ambulatory and clinical outcomes in different obesity populations for better management of hypertensive patients.

4.2. Limitations

There are some limitations in our study. First, our pooling estimates were based on some heterogeneities across the included studies. However, we attempted to explore the reasons for heterogeneities by subgroup analysis according to sex and initial BMI levels, and meta‐regression analyses about several variables. And we finally observed some factors that affected the outcome such as sex, initial weight, and the degree of weight reduction, which might explain a part of the heterogeneity. Second, our study included teenagers who have physiological weight and height gain, which may have influenced the results, but the experiment time was not too long, so the bias can be minimized, and we only selected studies whose sample loss weight successfully, it might produce select bias. In addition, our study did not include antihypertensive drugs in the analysis but instead divided weight loss into medical interventions and lifestyle interventions, as most studies had not clearly stated whether participants were taking antihypertensive drugs, so no further subgroup analysis was possible and more relevant prospective studies need to be conducted. Besides, we have analyzed the differences between regions, but the study in the population of Asia is too little, which may produce bias for our result of region subgroup, the more relevant study should be carried out in Asia. Furthermore, we replaced specific BP values for each period with ambulatory BP averages because we found that many studies had not given specific values. Thus, the data in our study might be less representative of the whole day's BP. We reviewed studies that corporate daytime and night‐time ambulatory BP. Based on studies including daytime and night‐time BP, we found that the mean daytime and night‐time ambulatory SBP values were reduced, but the mean may have less representation. Last, we only reviewed the English literature, which might lead to language bias. However, the main strength of our study is that this is the first meta‐analysis comparing the changes in BMI, clinic BP, and ambulatory BP before and after weight loss.

5. CONCLUSIONS

In conclusion, after the weight loss clinic and ambulatory BP decreased significantly, and the degree of change in clinic BP differs from that of ambulatory BP. The MD between clinic and ambulatory SBP was much larger in the medication intervention subgroup than the lifestyle intervention subgroup, but the outcome of ambulatory DBP was opposite. The differing characteristics of sex, ethnicity, age, and initial weight might explain the diversities. Further studies will be needed to investigate the differences and potential affecting factors between obesity and hypertension.

AUTHOR CONTRIBUTIONS

Designed the analysis: Shijie Yang and Yuqing Zhang. Collected the data: Shijie Yang and Zhanyang Zhou. Performed the analysis: Shijie Yang. Wrote the manuscript: Shijie Yang. Designed the tables: Zhanyang Zhou. Contributed the analysis tools: Zhanyang Zhou. Revised the manuscript: Huanhuan Miao and Yuqing Zhang. Conceived the analysis: Yuqing Zhang.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Supplimentary Information

Yang S, Zhou Z, Miao H, Zhang Y. Effect of weight loss on blood pressure changes in overweight patients: A systematic review and meta‐analysis. J Clin Hypertens. 2023;25:404–415. 10.1111/jch.14661