体重变化对血压影响的纵向队列研究

发布时间：2018-05-17 18:43

本文选题：高血压 + 体重变化　；参考：《山东大学》2017年硕士论文

【摘要】：高血压与肥胖均是全球公共卫生问题,关于肥胖与高血压关系的研究最早可以追溯到20世纪20年代,随后Landsberg等人开始用"obesity-related hypertension "对肥胖相关性高血压进行描述。尽管肥胖直接引起高血压的生物学机制尚未完全清楚,但两者间的关系已基本确立;然而另一个更具公共卫生学意义的危险因素"体重变化"与高血压的发生及血压之间的关系更值得引起研究者的关注。目前,大多数研究仅仅观察体重变化(定义为基线与高血压发病或最后一次随访时的体重之差)这一笼统的单一指标,而实际上体重变化有多种涵义,因而该指标无法从多角度阐明体重变化对高血压影响的模式;还有部分研究只关注某性别的人群或者研究人群的年龄段非常窄,可能会使研究结论的推广存在局限性。此外,有部分研究基于横断面设计或自我报告的人体测量学指标,也会存在偏倚。资料与方法:本研究基于"山东多中心纵向健康管理队列",选取了 2004年～2016年参加体检且年满18岁的城市汉族人群,构建高血压随访队列,体检两次及以上,排除身高、体重、收缩压和舒张压四个变量有缺失值者及首次体检时有高血压、心血管疾病者,最终有87296人(其中,男45482人、女41814人)进入队列。首先定义了五种"体重变化"指标,分别为"绝对体重变化(基线与高血压发病或最后一次随访时的体重之差)"、"相对体重变化(绝对体重变化除以基线体重)"、"年均体重变化(绝对体重变化除以随访时间(年))"、"绝对BMI变化(基线与高血压发病或最后一次随访时的BMI之差)"和"BMI变化速率(以体检时间为自变量、BMI为因变量直线回归的斜率)";每种指标有其不同的流行病学和统计学意义。然后基于中国肥胖标准,采用Cox比例风险回归模型探索城市汉族人群不同性别五种体重变化指标对高血压的影响;进而采用广义估计方程(GEE模型)探究体重变化指标与血压变化的定量关系(由于在计算BMI变化速率时已用到多次随访的数据,故构建GEE模型时,只纳入前四个指标)。最后,求得不同体重变化指标与高血压和血压变化关联性的标准化回归系数,以便为人们科学减肥提供指导依据。研究结果如下:1.BMI和五种体重变化指标的高血压发病密度及其比较。①无论男女,不同BMI分组(分别为低体重组,正常体重组,超重组和肥胖组)的高血压发病密度不同,且发病密度随BMI升高而增加;②无论男女,绝对体重变化、相对体重变化、年均体重变化、绝对BMI变化各分组的高血压发病密度均不同;除绝对体重变化中增重10kg组未发现发病密度间的性别差异外(u=0.384,P=0.701),其余各组均为男性高于女性。2.基线指标分析。所有纳入研究的体检指标均存在性别差异(P0.001);除女性的HDL值高于男性外(P0.001),其它基线指标均为男性高于女性(P0.001)。3.BMI和高血压关联性的Cox比例风险回归模型结果显示:RR值随BMI的增加而增大;无论男女,超重和肥胖均是高血压的危险因素(男性:RR值分别为1.254、1.554;女性:RR值分别为1.245、1.700);低体重是男性高血压的保护因素(RR=0.780)。4.五种体重变化指标与高血压关联性的Cox比例风险回归模型结果显示:①绝对体重变化(以绝对体重变化≤5kg作为参照组)。对于男性,减重10kg、5kg减重≤1Okg及5kg增重≤1Okg的男性高血压发病风险降低(RR值分别为0.680、0.585和0.943),当增重10kg时高血压发病风险增加(RR=1.212);而对于女性,减重是高血压的保护因素(减重1Okg、5kg减重≤1Okg的RR值分别为0.716、0.652),增重是其危险因素(5kg增重≤10kg、增重10kg的RR值分别为1.124、4.005);②相对体重变化(以相对体重变化≤5%作为参照组)。减重10%、5%减重≤10%均能降低男女高血压的发病风险(男性:RR值分别为0.681、0.633;女性:RR值分别0.744、0.694),当增重10%时会增加男女高血压的发病风险(男性:RR=1.083;女性:RR=2.100);③年均体重变化(以年均体重变化≤1kg作为参照组)。年均增重或减重1kg时,男女的RR值均大于1;④绝对BMI变化(以BMI变化1kg/m2作为参照组)。无论男女,BMI减小2kg/m2、1kg/m2BMI减小≤2kg/m2均为高血压的保护因素(男性:RR值分别为0.520、0.703;女性:RR值分别为0.637、0.761);当BMI增大2kg/m2时,女性高血压的发病风险升高(RR=1.870);⑤BMI变化速率。无论男女,BMI变化速率是高血压的危险因素(男性:RR=1.474;女性:RR=1.530);⑥依据标准化偏回归系数结果,无论男女,对高血压发病影响最大的指标均为BMI变化速率(标准化偏回归系数分别为 0.393、0.524)。5.四种体重变化指标与血压变化的相关分析及其GEE模型结果显示:(1)体重变化指标与血压变化的相关分析:①无论男女,体重变化指标与SBP变化和DBP变化均存在正相关关系;②无论男女,体重变化指标与SBP变化的相关系数均大于与DBP变化的相关系数。(2)体重变化指标与血压变化关联性的GEE模型:①对于男性,除绝对BMI变化指标外,其它体重变化指标对SBP变化、DBP变化的影响均较小,绝对体重变化、相对体重变化、年均体重变化每升高一个水平,SBP 平均升高 0.0021mmHg、0.018mmHg、0.0011mmHg,DBP 平均升高 0.0017mmHg、0.014mmHg、0.0007mmHg;绝对 BMI 变化每升高一个单位,SBP平均升高0.0026mmHg,DBP平均升高0.0021mmHg;②对于女性,除绝对BMI变化指标外,其它体重变化指标与DBP变化的关联性均大于其与SBP变化的关联性,绝对体重变化、相对体重变化、年均体重变化每升高一个水平,SBP平均升高 0.0026mmHg、0.0020mmHg、0.0007mmHg,DBP 平均升高 0.4650mmHg、0.3014mmHg、0.1810mmHg;绝对BMI变化每升高一个单位,SBP平均升高0.0022mmHg,DBP平均升高0.3236mmHg。(3)依据标准化偏回归系数结果,无论男女,对SBP变化、DBP变化影响最重要的指标均为绝对BMI变化。研究结论:1.体重变化对血压有影响,但不同体重变化指标与高血压和血压变化的关联性均不相同。2.依据Cox比例风险回归模型标准化偏回归系数结果:①男性各体重变化指标的重要性依次为:BMI变化速率、年均体重变化、绝对BMI变化、绝对体重变化和相对体重变化;②女性各体重变化指标的重要性依次为:BMI变化速率、年均体重变化、相对体重变化、绝对体重变化和绝对BMI变化。3.依据GEE模型标准化偏回归系数结果,无论男女,对SBP变化、DBP变化影响最重要的指标均为绝对BMI变化。
[Abstract]:Hypertension and obesity are all global public health problems. Studies on the relationship between obesity and hypertension can be traced back to 1920s. Then Landsberg and others began to describe obesity related hypertension with "obesity-related hypertension". Although the biological mechanism of obesity directly causes hypertension, the biological mechanism is not completely clear, But the relationship between the two has been basically established; however, another more public health risk factor, "weight change", is more worthy of the attention of the researchers. Most studies are currently only observing body weight changes (defined as the baseline and the body of hypertension or the last follow-up. In fact, there are many meanings of this general single indicator, but in fact, there are many meanings of body weight change, so the index can not explain the model of weight change impact on hypertension from multiple angles. Some studies are based on cross-sectional design or self reporting of anthropometric indicators, and there will be bias. Data and methods: Based on the "Shandong multi center longitudinal health management queue", the study selected the urban Han population aged 18 years old from 2004 to 2016. The follow-up cohort of hypertension was constructed, and two times and above were excluded. The four variables of height, weight, systolic and diastolic pressure were missing, and there were 87296 people (of them, 45482 men and 41814 women) with hypertension and cardiovascular disease at the first physical examination. First, five "body weight changes" were defined, respectively, "the body weight change (the baseline and hypertension or the last follow-up body). Relative weight change (absolute weight change divided by baseline weight) "," annual average weight change (absolute weight change divided by follow-up time (year)) "," "absolute BMI change (baseline and hypertension or the last follow-up of the BMI difference)" and "BMI change rate (with the time of physical examination as the independent variable, BMI as the slope of the variable linear regression) "Each index has its different epidemiological and statistical significance. Then based on the Chinese obesity standard, the Cox proportional risk regression model is used to explore the effect of five kinds of body weight changes on the hypertension in the urban Han population, and then the generalized estimation equation (GEE model) is used to explore the quantitative correlation between the index of body weight change and the change of blood pressure. In order to provide guidance for people to lose weight, the results are as follows: 1.BMI and five kinds of weight. The results are as follows: the results are as follows: 1.BMI and five kinds of weight. (1) the density of hypertension in different BMI groups (low body recombination, normal weight group, super recombinant and obesity group) was different, and the density of the disease was increased with the increase of BMI; and the change of absolute weight, relative weight, annual weight, absolute BMI change The incidence of high blood pressure in each group was different, except the sex difference between the weight gain 10kg group and the weight gain group (u=0.384, P=0.701), and the other groups were all higher than the female.2. baseline analysis. All the examination indexes in the study had sex difference (P0.001), and the HDL value of women was higher than that of the male (P0.0). 01) the Cox proportional risk regression model of other baseline indicators that were male higher than women (P0.001).3.BMI and hypertension showed that the RR value increased with the increase of BMI; both male and female, overweight and obesity were risk factors for hypertension (male: RR, 1.254,1.554, RR value, respectively 1.245,1.700); low weight was male. The Cox proportional risk regression model of the five weight change indices of hypertension (RR=0.780).4. and the association of hypertension showed: (1) the absolute weight change (with the absolute weight change < 5kg as the reference group). For men, the risk of hypertension in men with weight loss 10kg, 5kg weight less than 1Okg and 5kg weight gain less than 1Okg (RR values, respectively) For 0.680,0.585 and 0.943), the risk of hypertension increased when weight gain was 10kg (RR=1.212); for women, weight loss was a protective factor for hypertension (weight loss 1Okg, RR value of 5kg weight less than 1Okg respectively 0.716,0.652), and weight gain was a risk factor (5kg increase weight < < 10kg, 10kg RR value); and relative weight change (relative weight) Weight change less than 5% as a reference group. Weight loss 10%, 5% weight loss less than 10% can reduce the risk of hypertension in men and women (male: RR value is 0.681,0.633, female: RR respectively 0.744,0.694), when the weight gain 10% will increase the risk of hypertension in men and women (male: RR=1.083; female: RR=2.100); and (3) annual weight change (average annual weight change is less than) 1kg as a reference group). When the average annual weight gain or weight loss 1kg, the RR value of both men and women is greater than 1; (4) the absolute BMI change (BMI change 1kg/m2 as the reference group). Regardless of the male and female, BMI decreases 2kg/m2,1kg/m2BMI less than 2kg/m2 is the protective factor of hypertension (male: RR value is 0.520,0.703; female: RR values respectively); when enlargement The risk of hypertension in women (RR=1.870); (5) the rate of change in BMI. The rate of change in BMI is a risk factor for hypertension (male: RR=1.474; female: RR=1.530). (6) according to the standardized partial regression coefficient, both male and female, the most influential index of hypertension is the rate of BMI change (standardized partial regression coefficient) The correlation analysis of four kinds of weight change indexes and blood pressure changes in.5. and the results of GEE model showed that: (1) the correlation analysis between the index of weight change and the change of blood pressure: (1) there is a positive correlation between the index of body weight change and the change of SBP and the changes of DBP in both men and women; (2) the relationship between the index of body weight change and the change of SBP in both men and women The correlation coefficient was greater than that of DBP. (2) the GEE model of the correlation between the weight change index and the blood pressure change: (1) for the male, except for the absolute BMI change index, the influence of the other weight change indexes on the change of SBP, the change of DBP, the change of absolute weight, the average annual weight change of one level, the average of SBP rise High 0.0021mmHg, 0.018mmHg, 0.0011mmHg, DBP increase 0.0017mmHg, 0.014mmHg, 0.0007mmHg, and the absolute BMI changes each unit, SBP increase averagely 0.0026mmHg, DBP average increase 0.0021mmHg. Association, absolute weight change, relative weight change, average annual weight change per level, SBP average increase 0.0026mmHg, 0.0020mmHg, 0.0007mmHg, DBP increase 0.4650mmHg, 0.3014mmHg, 0.1810mmHg; absolute BMI change per unit, SBP average increase 0.0022mmHg, DBP average rise (3) according to standardized partial return The result of return coefficient, the most important index of SBP change and DBP change for both men and women is absolute BMI change. Conclusion: 1. body weight changes have influence on blood pressure, but the correlation of different weight changes and hypertension and blood pressure change is different.2. according to the Cox proportional risk regression model standardized partial regression coefficient results: (1) male The importance of the index of body weight change was BMI change rate, annual weight change, absolute BMI change, absolute weight change and relative weight change, and the importance of female body weight change indexes were BMI change rate, annual weight change, relative weight change, absolute weight change and absolute BMI change.3. based on GEE model standard. The results of the normalized partial regression coefficient showed that the most important indicators of SBP change and DBP change were absolute BMI changes, both male and female.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R544.1

【相似文献】