P,N, Lee Statistics and Computing Ltd, Surrey, UK.
Biomark Res. 2013 Sep 13;1(1):26. doi: 10.1186/2050-7771-1-26.
A higher concentration of high density lipoprotein cholesterol (HDL-C) in ex-smokers than smokers has consistently been observed. Better evidence of quitting effects comes from within-subject changes. We extend an earlier meta-analysis to quantify the reduction, and investigate variation by time quit and other factors. We conducted Medline and Cochrane searches for studies measuring HDL-C in subjects while still smoking and later having quit. Using unweighted and inverse-variance weighted regression analysis, we related changes (in mmol/l) to intra-measurement period, and estimated time quit, and to study type, location and start year, age, sex, product smoked, validation of quitting, baseline HDL-C, baseline and change in weight/BMI, and any study constraints on diet or exercise. Forty-five studies were identified (17 Europe, 16 North America, 11 Asia, 1 Australia). Thirteen were observational, giving changes over at least 12 months, with most involving >1000 subjects. Others were smoking cessation trials, 12 randomized and 20 non-randomized. These were often small (18 of <100 subjects) and short (14 of <10 weeks, the longest a year). Thirty studies provided results for only one time interval. From 94 estimates of HDL-C change, the unweighted mean was 0.107 (95% CI 0.085-0.128). The weighted mean 0.060 (0.044 to 0.075) was lower, due to smaller estimates in longer term studies. Weighted means varied by time quit (0.083, 0.112, 0.111, 0.072, 0.058 and 0.040 for <3, 3 to <6, 6 to <13, 13 to <27, 27 to <52 and 52+ weeks, p=0.006). After adjustment for time quit, estimates varied by study constraint on diet/exercise (p=0.003), being higher in studies requiring subjects to maintain their pre-quitting habits, but no other clear differences were seen, with significant (p<0.05) increases following quitting being evident in all subgroups studied, except where data were very limited. For both continuing and never smokers, the data are (except for two large studies atypically showing significant HDL-C declines in both groups, and a smaller decline in quitters) consistent with no change, and contrast markedly with the data for quitters. We conclude that quitting smoking increases HDL-C, and that this increase occurs rapidly after quitting, with no clear pattern of change thereafter.
曾吸烟者的高密度脂蛋白胆固醇（HDL-C）浓度高于吸烟者，这一现象一直存在。来自于个体内变化的戒烟效果的证据更好。我们扩展了之前的荟萃分析，以量化这种降低，并调查了时间戒烟和其他因素的变化。我们对仍在吸烟并随后戒烟的研究对象中测量 HDL-C 的研究进行了 Medline 和 Cochrane 搜索。使用未加权和倒数方差加权回归分析，我们将变化（mmol/l）与测量期间、时间戒烟以及研究类型、地点和开始年份、年龄、性别、吸烟产品、戒烟验证、基线 HDL-C、体重/ BMI 的基线和变化以及对饮食或运动的任何研究限制联系起来。确定了 45 项研究（欧洲 17 项，北美 16 项，亚洲 11 项，澳大利亚 1 项）。其中 13 项为观察性研究，至少在 12 个月内有变化，其中大多数涉及> 1000 名受试者。其他是戒烟试验，其中 12 项为随机试验，20 项为非随机试验。这些试验通常较小（18 项< 100 名受试者），较短（14 项< 10 周，最长一年）。30 项研究仅提供了一个时间间隔的结果。在 94 项 HDL-C 变化的估计中，未加权平均值为 0.107（95%CI 0.085-0.128）。加权平均值 0.060（0.044 至 0.075）较低，这是由于长期研究中的估计值较小。加权平均值因时间戒烟而变化（<3 周、3-<6 周、6-<13 周、13-<27 周、27-<52 周和 52+周时分别为 0.083、0.112、0.111、0.072、0.058 和 0.040，p=0.006）。在调整时间戒烟后，研究限制对饮食/运动的影响存在差异（p=0.003），需要受试者保持戒烟前习惯的研究中的估计值更高，但未发现其他明显差异，除了数据非常有限的情况外，所有研究亚组都有明显（p<0.05）的戒烟后升高。对于继续吸烟者和从不吸烟者，数据（除了两项大型研究异常地显示两组均有显著的 HDL-C 下降，以及戒烟者下降幅度较小外）均表明没有变化，与戒烟者的数据形成鲜明对比。我们得出的结论是，戒烟会增加 HDL-C，并且这种增加在戒烟后迅速发生，此后没有明显的变化模式。
