Abstract
在不同群体中获得性免疫缺陷综合征(以下简称“艾滋病”)的防控需求具有显著的异质性,这为制订和实施防控策略带来了极大挑战。动态评估不同群体的异质性和疾病进展轨迹至关重要。潜分类增长模型是将纵向数据拟合为不同发展轨迹的N个亚组,以识别和分析不同亚组的发展轨迹,从而为疾病的防控策略提供新视角。潜分类增长模型在艾滋病防控领域的应用中显现出显著优势,尤其在深入理解和分析艾滋病流行病学特征、风险行为、心理研究、检测的异质性和动态变化等方面,总结其优势和局限可为艾滋病的精准防控提供科学依据。
Keywords: 潜分类增长模型, 获得性免疫缺陷综合征, 预防控制, 应用进展
Abstract
The prevention and control requirements for HIV/AIDS vary significantly among different populations, posing substantial challenges to the formulation and implementation of intervention strategies. Dynamically assessing the heterogeneity and disease progression trajectories of various groups is crucial. Latent class growth model (LCGM) serves as a statistical approach that fits a longitudinal data into N subgroups of individual development trajectories, identifying and analyzing the progression paths of different subgroups, thereby offering a novel perspective for disease control strategies. LCGM has shown significant advantages in the application of HIV/AIDS prevention and control, especially in gaining a deeper understanding and analysis of epidemiological characteristics, risk behaviors, psychological research, heterogeneity in testing, and dynamic changes. Summarizing the advantages and limitations of applying LCGM can provide a reliable basis for precise prevention and control of HIV/AIDS.
Keywords: latent class growth model, acquired immunodeficiency syndrome, prevention and control, application progress
获得性免疫缺陷综合征(以下简称“艾滋病”)是由人类免疫缺陷病毒(human immunodeficiency virus,HIV)引起的性传播疾病,至今仍是全球公共卫生领域面临的重大挑战[1-2]。尽管联合国艾滋病规划署(United Nations Programme on HIV/AIDS,UNAIDS)报告[3]指出防控措施已取得显著成效,但2022年全球仍有3 900万人感染HIV,68万人因HIV相关并发症而死亡。艾滋病的疾病进展与生物学和社会心理因素的动态变化密切相关,不同群体具有较大的异质性[4-6]。如何精确评估不同群体HIV感染的异质性和疾病进展轨迹,从而更有效地进行早期干预和管理,已成为当前疾病防控的关键问题。传统的疾病评估方法主要侧重于单一时间点的数据分析,这在一定程度上限制了对疾病进程动态变化的深入理解[7-9]。潜分类增长模型(latent class growth model,LCGM)能够揭示不同群体在不同阶段中的变化趋势,并通过纵向数据分析,追踪疾病的发展路径[10-11]。本文旨在总结LCGM在艾滋病防控策略中的应用进展,深入探讨LCGM应用于艾滋病领域的优势和局限,以期为艾滋病的精准防控提供科学依据。
1. LCGM概述
1.1. 模型简介
LCGM是一种高级统计学方法,用于分析纵向数据中不同潜在亚组随时间发展的变化轨迹。该模型的核心优势在于能够处理多种数据分布,揭示总体中存在的不同发展模式,从而成为分析纵向数据的有效方法。LCGM也称轨迹模型(trajectory model,TM)[12-13]和群组轨迹模型(group-based trajectory model,GBTM)[14],强调其在识别具有相似发展轨迹的群体或类别方面的应用。传统的数据分析模型假定总体发展趋势相同[15],然而在现实研究中很难满足这一条件[16]。LCGM避免了这一局限,不仅可以分析总体的发展趋势,而且能够揭示总体中亚组的发展轨迹,为探究纵向数据的发展模式提供了更有价值的研究视角和分析方法[17]。
1.2. 模型原理
LCGM基于一个核心假设,即数据中存在潜在的异质性,总体中可能存在多个亚组,每个亚组都有其独特的发展轨迹或模式[18]。该模型允许各亚组之间展现不同的发展趋势,但同一亚组内的个体遵循相同的发展趋势,而不同亚组间的轨迹表现出显著差异[19-20]。具体来说,在亚组内部,个体的发展轨迹(即截距和斜率)是一致的,而在不同亚组之间,这些轨迹可能存在显著差异[21-22]。LCGM的核心价值在于它充分关注和研究了群体的异质性。
1.3. 模型拟合和评估
LCGM能够拟合不同类型数据的分布[23]。计数数据可以拟合Poisson分布或零膨胀Poisson分布模型;二分类数据可以拟合二分类logit分布模型;而存在一定范围内或近似正态分布的连续数据,可以拟合删截正态分布模型[16, 24]。模型拟合过程中的一个关键步骤是确定亚组的数量及其轨迹,此步骤需要多次拟合以确认最佳的亚组数量和轨迹形状,一般从较少的亚组数量开始,逐步增加,以便识别出数据中的潜在模式[25-26]。每个亚组先从高阶模型开始拟合,若无统计学意义,则移除非显著的参数,从而简化模型,找到最佳的模型复杂度与数据拟合之间的平衡[27]。评估模型的拟合效果可以通过若干指标进行判断,如贝叶斯信息标准(Bayesian information criterion,BIC)、贝叶斯因子对数值和平均验后分组概率(average posterior probability,AvePP)等[19, 26, 28]。
2. LCGM在医学领域中的应用
LCGM提供了一种分析异质性总体不同发展模式的统计学方法,能够识别总体中的潜在亚群,并揭示在一段时间内亚群的发展轨迹,为研究者深入剖析总体及其亚群内部的动态发展提供了科学依据[29-31]。该模型现已应用于疾病管理、护理学、心理学和临床治疗等多个方面,用于分析亚群和总体在多个时间点的变化轨迹,揭示总体中不同亚群的轨迹特征[17, 25, 32-35]。例如,Bhavani等[34]使用LCGM探究多中心队列中的新型冠状病毒肺炎住院患者的生命体征轨迹,并评估由此产生的亚型的临床特征和治疗效果,发现新型冠状病毒肺炎的不同亚型可能是制订治疗方案的重要依据。另外,Wang等[36]通过LCGM探讨乳腺癌患者支持性护理需求的轨迹模式。该研究发现:支持性护理需求展现出3种不同的变化轨迹,即“高需求下降组”“高需求持续组”和“低需求持续组”,并且这些轨迹与患者的年龄、教育水平、子女数量等因素有关。因此,医疗保健提供者可以根据这些发现为不同特征的患者制订个性化的干预措施。
3. LCGM在艾滋病防控领域中的应用
艾滋病的防控已成为全球性问题,防控措施包括艾滋病相关知识的普及、艾滋病的检测和诊断及抗逆转录病毒治疗等多个方面,在实际应用中仍然面临着诸多挑战,缺乏对不同人群的特定关注和针对性措施。2021年联合国大会发布的《关于艾滋病毒和艾滋病问题的政治宣言》[37]强调制订有针对性的艾滋病预防、宣传教育和护理服务方案的重要性,以满足不同重点人群的需求并提高公众对疾病的认知。在这一背景下,LCGM的应用变得尤为关键。该模型能够深入分析现有数据,精确识别不同群体在艾滋病防控中的特定需求和风险,从而为制订更精准、更具针对性的防控策略提供数据支持。
3.1. 艾滋病的流行病学特征
在分析艾滋病的流行病学特征时,LCGM通常用于探索群体中的亚群在特定时间的疾病变化和发展路径。俞秋嫣等[38]运用LCGM分析通过非婚异性性行为传播的HIV感染者和艾滋病患者的流行病学特征及其潜类别,结果表明患者可以分为4个具有独立流行病学特征的亚组人群。该研究提示艾滋病防控策略应关注不同非婚异性性行为人群特征的差异,并可根据不同亚组人群的综合特征来指导艾滋病经性传播防治策略的制订。
3.2. 艾滋病的风险因素分析
在艾滋病风险行为的探究中,LCGM已被证实是一种可靠的数据分析工具,能够分析特定人群的异质性及风险行为的演变轨迹。一项采用LCGM分析男男性行为(men who have sex with men,MSM)人群无保护性行为轨迹特征的研究[39]发现:在一定的时间跨度内,个体的风险行为表现出不同的变化(高风险组和低风险组),且与社会人口学变量(年龄、性别、社会经济地位)和心理健康变量(心理创伤)密切相关。孟静[40]应用LCGM探究MSM人群HIV预期污名的发展趋势及其影响因素,发现HIV预期污名的发展趋势存在群体异质性,可分为高水平污名稳定组、中等水平污名缓慢下降组和低水平污名快速下降组,影响HIV预期污名潜在类别的因素为学历和是否有子女。由此可见,LCGM能有效区分并描述HIV感染风险行为的变化差异,为制订预防措施和干预策略提供针对性指导。
3.3. 艾滋病的相关心理研究
在艾滋病的相关心理研究中,LCGM展现出显著的应用价值。有研究[15]应用LCGM识别MSM人群焦虑、抑郁发展轨迹的类别分组,并进一步探讨各组轨迹的影响因素。结果表明焦虑和抑郁的发展轨迹在不同个体间表现出明显的异质性,且与人口学特征、知识、态度等因素存在相关性。Li等[41]对新诊断为HIV阳性的MSM人群进行了为期1年的前瞻性队列研究,以探索该人群抑郁症状的变化轨迹以及相关因素。通过应用LCGM,该研究将MSM的抑郁症状分为3个潜在类别:非抑郁组、轻度抑郁组和持续中重度抑郁组。上述研究表明,通过应用LCGM,研究者能够深入探讨艾滋病关键人群心理因素发展的多样性和复杂性,并识别与其发展轨迹相关的多种因素。但现有研究在理论方面仍有欠缺,后续研究可结合心理学和社会学理论,深化模型应用,拓宽研究视角。
3.4. 艾滋病的检测及诊断
艾滋病检测在艾滋病防控中至关重要。LCGM的应用能够帮助医务人员深入理解个体在HIV检测行为及其相关变量上的发展轨迹。这些轨迹有助于揭示个体在HIV检测自我效能和检测意愿等方面的变化。一项在山东省进行的研究[42]招募了404名MSM者,实施个体干预和社区干预,并测量HIV检测自我效能水平;通过运用LCGM,研究者分析了HIV检测自我效能的发展轨迹及影响发展轨迹的因素。该研究结果显示在HIV检测自我效能方面,MSM群体呈现出2种不同的发展轨迹即干预反应组和干预无反应组,职业因素可能是影响这些发展轨迹的关键变量。
3.5. 其他
LCGM还可应用于艾滋病的治疗依从性研究中,具有识别患者依从性变化的作用。Liu等[43]通过实时监测使用含有可摄入传感器(ingestible sensor,IS)的抗逆转录病毒药物的患者依从性,分析了在治疗过程中患者依从性的变化轨迹。结果显示IS系统的使用与提高药物依从性和较低的病毒载量相关。这一发现为改善HIV治疗效果提供了新的技术途径,强调了实时监测技术在提升患者治疗依从性中的潜在价值。后续研究应进一步验证该系统在不同患者群体中的适用性和效果。Jovanović等[44]探索了塞尔维亚HIV的流行趋势,结果表明塞尔维亚的HIV感染率呈现出上升趋势,存在5个潜在类别,其中2个类别主要与MSM传播、乙型肝炎病毒和其他性传播感染有关。
4. 结语与展望
LCGM是一种基于潜在变量的统计学方法,用于识别和描述数据中的潜在类别和轨迹[45]。在艾滋病防控领域的应用中显现出显著优势,尤其在深入理解和分析艾滋病流行病学特征、风险行为、心理研究、检测的异质性和动态变化等方面[46-49]。LCGM能够深入挖掘现有数据,更精确地识别不同亚群在艾滋病防控中所面临的特定需求和风险,为艾滋病的防控提供更精准和个性化的指导。
尽管LCGM在艾滋病领域的应用中有诸多优势,但也存在一定的局限性。首先,LCGM的准确性和可靠性依赖于数据质量,如果数据存在缺失、异常等问题,可能影响模型的拟合和解释。其次,模型的选择和解释可能受到研究设计、实施阶段的假设和约束等的限制。例如在多中心艾滋病队列研究中,尽管研究包含了较大的样本量且是多中心研究,但仍可能未能充分考虑影响认知轨迹的其他关键变量[50]。最后,LCGM只能描述轨迹之间的差异,不能解释轨迹之间的因果关系,需要结合其他理论和方法进行深入分析。此外,在艾滋病的研究中,个体的行为和疾病进程往往受到多种因素的共同影响,如何准确建模处理复杂的关系,仍是一个待解决的问题。
在艾滋病防控领域中,LCGM还有许多尚未充分挖掘的应用潜力。首先,在艾滋病患者的生活质量、药物依赖、社会支持等方面,可以进一步探索这些方面的轨迹特征和潜在类别之间的影响因素,为艾滋病的综合防控提供更多元的研究视角。其次,目前应用LCGM的艾滋病患者心理健康的研究主要集中于焦虑、抑郁等负性心理的探讨,但积极心理学的研究也十分重要,未来研究可尝试应用LCGM探索积极心理因素在艾滋病患者中的发展轨迹(如乐观、希望和心理弹性等),为全面了解艾滋病患者心理因素提供依据。最后,未来研究可以结合深度学习和其他机器学习方法,开发能够处理更复杂关系和大规模数据的模型,进一步拓展LCGM在艾滋病防控研究中的应用。
基金资助
湖南省重点研发计划项目(2021SK2031);中南大学中央高校基本科研业务费专项资金(2023ZZTS0851)。This work was supported by the Key Research and Development Plan Project of Hunan Province (2021SK2031) and the Fundamental Research Funds for the Central Universities of Central South University (2023ZZTS0851), China.
利益冲突声明
作者声称无任何利益冲突。
作者贡献
翟咪咪、李亚敏 论文构思、撰写与修改;刘苏顺、李云霞、刘义婷、李莉 论文构思,文献收集;雷先阳 论文构思与审阅。所有作者阅读并同意最终的文本。
Footnotes
http://dx.chinadoi.cn/
原文网址
http://xbyxb.csu.edu.cn/xbwk/fileup/PDF/202404621.pdf
参考文献
- 1. Abdool Karim SS, Barre-Sinoussi F, Varmus HE, et al. Threatening the global AIDS response-obstacles to PEPFAR’s reauthorization[J]. N Engl J Med, 2023, 389(13): 1159-1161. 10.1056/NEJMp2310330. [DOI] [PubMed] [Google Scholar]
- 2. Gostin LO. US global AIDS programme is under threat[J]. BMJ, 2023, 382: 1929. 10.1136/bmj.p1929. [DOI] [PubMed] [Google Scholar]
- 3. UNAIDS . World AIDS Day 2023 Fact Sheet[EB/OL]. (2023-07-13)[2023-09-27]. https://www.unaids.org/en/resources/documents/2023/UNAIDS_FactSheet.
- 4. May MT, Hogg RS, Justice AC, et al. Heterogeneity in outcomes of treated HIV-positive patients in Europe and North America: relation with patient and cohort characteristics[J]. Int J Epidemiol, 2012, 41(6): 1807-1820. 10.1093/ije/dys164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Sharma M, Ong JJ, Celum C, et al. Heterogeneity in individual preferences for HIV testing: a systematic literature review of discrete choice experiments[J]. EClinical Medicine, 2020, 29: 100653. 10.1016/j.eclinm.2020.100653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Ataguba JEO, Birungi C, Cunial S, et al. Income inequality and pandemics: insights from HIV/AIDS and COVID-19-a multicountry observational study[J/OL]. BMJ Glob Health, 2023, 8(9): e013703[2023-10-02]. 10.1136/bmjgh-2023-013703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hong CL, Ochoa AM, Wilson BDM, et al. The associations between HIV stigma and mental health symptoms, life satisfaction, and quality of life among Black sexual minority men with HIV[J]. Qual Life Res, 2023, 32(6): 1693-1702. 10.1007/s11136-023-03342-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Robinson A, Cooney A, Fassbender C, et al. Examining the relationship between HIV-related stigma and the health and wellbeing of children and adolescents living with HIV: a systematic review[J]. AIDS Behav, 2023, 27(9): 3133-3149. 10.1007/s10461-023-04034-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Mireles L, Horvath KJ, Guadamuz TE, et al. The moderating role of social support and HIV stigma on the association between depression and ART adherence among young Thai men who have sex with men[J]. AIDS Behav, 2023, 27(9): 2959-2968. 10.1007/s10461-023-04018-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Murphy DA, Brecht ML, Herbeck DM, et al. Trajectories of HIV risk behavior from age 15 to 25 in the national longitudinal survey of youth sample[J]. J Youth Adolesc, 2009, 38(9): 1226-1239. 10.1007/s10964-008-9323-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. You RJ, Li WJ, Ni LH, et al. Study on the trajectory of depression among middle-aged and elderly disabled people in China: Based on group-based trajectory model[J]. SSM Popul Health, 2023, 24: 101510. 10.1016/j.ssmph.2023.101510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hamid JS, Roslin NM, Paterson AD, et al. Using a latent growth curve model for an integrative assessment of the effects of genetic and environmental factors on multiple phenotypes[J]. BMC Proc, 2009, 3(Suppl 7): S44. 10.1186/1753-6561-3-S7-S44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Cusack SE, Bountress KE, Lind MJ, et al. Trauma exposure, alcohol consumption, and sleep quality: a latent growth curve model[J]. J Am Coll Health, 2022, 70(7): 2126-2134. 10.1080/07448481.2020.1845181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. 毛昂. 基于GBTM的四川省成人艾滋病抗病毒治疗后CD4+ T淋巴细胞变化轨迹模式及其影响因素研究[D]. 成都: 成都医学院, 2023. [Google Scholar]; MAO Ang. Study on trajectory patterns and influencing factors of CD4+ T lymphocyte changes after antiretroviral therapy in adults with AIDS in Sichuan Province based on GBTM[D]. Chengdu: Chengdu Medical College, 2023. [Google Scholar]
- 15. 吴旦. 基于轨迹分析模型的MSM人群焦虑、抑郁及影响因素研究[D]. 重庆: 重庆医科大学, 2020. [Google Scholar]; WU Dan. Study on anxiety, depression and influencing factors of MSM population based on trajectory analysis model[D]. Chongqing: Chongqing Medical University, 2020. [Google Scholar]
- 16. 冯国双, 于石成, 刘世炜. 轨迹分析模型在追踪数据分析中的应用[J]. 中国预防医学杂志, 2014, 15(3): 292-295. 10.16506/j.1009-6639.2014.03.009. [DOI] [Google Scholar]; FENG Guoshuang, YU Shicheng, LIU Shiwei. Application of trajectory analysis model in tracking data analysis[J]. Chinese Preventive Medicine, 2014, 15(3): 292-295. 10.16506/j.1009-6639.2014.03.009. [DOI] [Google Scholar]
- 17. Thomas AJ, Mitchell ES, Pike KC, et al. Stressful life events during the perimenopause: longitudinal observations from the Seattle midlife women’s health study[J]. Womens Midlife Health, 2023, 9(1): 6. 10.1186/s40695-023-00089-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. 黄明明, 叶璇, 陈丽萍, 等. 初中生校园受欺负发展轨迹的潜类别增长模型分析[J]. 中国心理卫生杂志, 2021, 35(4): 315-320. 10.3969/j.issn.1000-6729.2021.04.009. [DOI] [Google Scholar]; HUANG Mingming, YE Xuan, CHEN Liping, et al. Developmental trajectory of campus victimization in middle school students based on Latent class growth model analysis[J]. Chinese Mental Health Journal, 2021, 35(4): 315-320. 10.3969/j.issn.1000-6729.2021.04.009. [DOI] [Google Scholar]
- 19. 邓锐斌. 基于潜类别增长模型的孕期负性情绪对妊娠结局影响研究[D]. 重庆: 重庆医科大学, 2021. [Google Scholar]; DENG Ruibin. Study of the effect of negative emotions on pregnancy outcome based on the latent class growth model[D]. Chongqing: Chongqing Medical University, 2021. [Google Scholar]
- 20. Wang X, Lu JJ, Liu Q, et al. Childhood experiences of threat and deprivation predict distinct depressive symptoms: a parallel latent growth curve model[J]. J Affect Disord, 2022, 319: 244-251. 10.1016/j.jad.2022.09.061. [DOI] [PubMed] [Google Scholar]
- 21. Birkeland MS, Holt T, Ormhaug SM, et al. Perceived social support and posttraumatic stress symptoms in children and youth in therapy: a parallel process latent growth curve model[J]. Behav Res Ther, 2020, 132: 103655. 10.1016/j.brat.2020.103655. [DOI] [PubMed] [Google Scholar]
- 22. Ni YH, Tein JY, Zhang MQ, et al. The need to belong: a parallel process latent growth curve model of late life negative affect and cognitive function[J]. Arch Gerontol Geriatr, 2020, 89: 104049. 10.1016/j.archger.2020.104049. [DOI] [PubMed] [Google Scholar]
- 23. Rodríguez-Cano R, Paulus DJ, Zvolensky MJ, et al. Depressive symptoms in the trajectory of craving during smoking cessation treatment: a latent growth curve model[J]. Am J Drug Alcohol Abuse, 2018, 44(4): 472-479. 10.1080/00952990.2018.1423687. [DOI] [PubMed] [Google Scholar]
- 24. Li HC, Tucker JD, Ma W, et al. Mediation analysis of peer norms, self-efficacy, and condom use among Chinese men who have sex with men: a parallel process latent growth curve model[J]. Arch Sex Behav, 2020, 49(1): 287-297. 10.1007/s10508-019-1459-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. 麻雨婷, 张宝山. 老年人负性情绪的联合发展轨迹与心理资源的关系: 基于潜类别增长的分析[J]. 西南大学学报(自然科学版), 2022, 44(10): 10-20. 10.13718/j.cnki.xdzk.2022.10.002. [DOI] [Google Scholar]; MA Yuting, ZHANG Baoshan. Joint trajectories of negative emotions and psychological resource in older adults: an analysis based on latent class growth model[J]. Journal of Southwest University. Natural Science Edition, 2022, 44(10): 10-20. 10.13718/j.cnki.xdzk.2022.10.002. [DOI] [Google Scholar]
- 26. 路卓. 基于潜类别增长模型女性孕产期抑郁发展轨迹及影响因素研究[D]. 重庆: 重庆医科大学, 2020. [Google Scholar]; LU Zhuo. Study on the development track and influencing factors of maternal depression based on latent category growth model[D]. Chongqing: Chongqing Medical University, 2020. [Google Scholar]
- 27. Liao HP, Pan XF, Yin XQ, et al. Decreased COVID-related adaptive behavior and increased negative affect: a multivariate latent growth curve model[J]. J Health Psychol, 2022, 27(9): 2115-2128. 10.1177/13591053211021651. [DOI] [PubMed] [Google Scholar]
- 28. 马玲, 丁枫, 李嘉. 在线知识社区中用户付费行为的发展轨迹: 基于潜类别增长模型的研究[J]. 管理评论, 2023, 35(3): 136-147. 10.14120/j.cnki.cn11-5057/f.2023.03.010. [DOI] [Google Scholar]; MA Ling, DING Feng, LI Jia. The trajectory of user payment behavior in online knowledge communities: a study based on latent class growth model[J]. Management Review, 2023, 35(3): 136-147. 10.14120/j.cnki.cn11-5057/f.2023.03.010. [DOI] [Google Scholar]
- 29. Rosenberg M, Houghton S, Hunter SC, et al. A latent growth curve model to estimate electronic screen use patterns amongst adolescents aged 10 to 17 years[J]. BMC Public Health, 2018, 18(1): 332. 10.1186/s12889-018-5240-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Barkmann C, Helle N, Bindt C. Is very low infant birth weight a predictor for a five-year course of depression in parents? A latent growth curve model[J]. J Affect Disord, 2018, 229: 415-420. 10.1016/j.jad.2017.12.020. [DOI] [PubMed] [Google Scholar]
- 31. Liu LF, Weng RH, Wu J. Exploring factors influencing residents’ health outcomes in long-term care facilities: 1-year follow-up using latent growth curve model[J]. Qual Life Res, 2014, 23(9): 2613-2627. 10.1007/s11136-014-0710-z. [DOI] [PubMed] [Google Scholar]
- 32. Yuan Y, Chen SM, Lin CJ, et al. Association of triglyceride-glucose index trajectory and frailty in urban older residents: evidence from the 10-year follow-up in a cohort study[J]. Cardiovasc Diabetol, 2023, 22(1): 264. 10.1186/s12933-023-02002-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Song YX, Huang YC, Li YY, et al. Risk factors for poor progression of addictive Internet use across different COVID-19 periods in China[J]. Am J Addict, 2023, 32(6): 593-605. 10.1111/ajad.13464. [DOI] [PubMed] [Google Scholar]
- 34. Bhavani SV, Robichaux C, Verhoef PA, et al. Using trajectories of bedside vital signs to identify COVID-19 subphenotypes[J]. Chest, 2024, 165(3): 529-539. 10.1016/j.chest.2023.09.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. 任珉. 基于潜分类增长模型妊娠期体重轨迹的妊娠期高血压疾病、妊娠期糖尿病和妊娠结局预测[D]. 天津: 天津医科大学, 2019. [Google Scholar]; REN Min. Prediction of hypertensive disorder complicating pregnancy, gestational diabetes mellitus and pregnancy outcome based on latent classification growth model[D]. Tianjin: Tianjin Medical University, 2019. [Google Scholar]
- 36. Wang SX, Li Y, Zhang EM, et al. Trajectory patterns and influencing factors of supportive care needs in Chinese elderly breast cancer patients[J]. Support Care Cancer, 2023, 31(10): 558. 10.1007/s00520-023-08003-y. [DOI] [PubMed] [Google Scholar]
- 37. 75th Session of the United Nations General Assembly . Political Declaration on HIV and AIDS[EB/OL]. (2021-06-08)[2023-09-27]. https://www.unaids.org/sites/default/files/media/documents/2021_political-declaration-on-hiv-and-aids_zh.pdf.
- 38. 俞秋嫣, 徐鹏, 林鹏, 等. 2015—2017年广东省江门市HIV感染者和艾滋病患者经非婚异性性传播流行特征及其潜类别分析[J]. 中华预防医学杂志, 2018, 52(12): 1269-1275. 10.3760/cma.j.issn.0253-9624.2018.12.015. [DOI] [PubMed] [Google Scholar]; YU Qiuyan, XU Peng, LIN Peng, et al. Epidemiological characteristics and latent class analysis of non-marital heterosexual behaviors among human immunodeficiency virus/acquired immunodeficiency syndrome individuals in Jiangmen, Guangdong Province between 2015 and 2017[J]. Chinese Journal of Preventive Medicine, 2018, 52(12): 1269-1275. 10.3760/cma.j.issn.0253-9624.2018.12.015. [DOI] [PubMed] [Google Scholar]
- 39. 林倚伊. MSM人群无保护性行为轨迹特征及PrEP性行为去抑制化研究[D]. 重庆: 重庆医科大学, 2017. [Google Scholar]; LIN Yiyi. Trajectory characteristics analysis of high-risk sexual behaviours and sexual risk compensation in HIV pre-exposure prophylaxis clinical trial among men who have sex with men[D]. Chongqing: Chongqing Medical University, 2017. [Google Scholar]
- 40. 孟静. 基于轨迹模型的MSM人群HIV预期污名发展趋势及影响因素研究[D]. 济南: 山东大学, 2023. [Google Scholar]; MENG Jing. Study on the development trend and influencing factors of anticipated HIV stigma among MSM population based on trajectory model[D]. Jinan: Shandong University, 2023. [Google Scholar]
- 41. Li X, Liu Y, Han J, et al. Trajectories of depressive symptoms in young and middle-aged men who have sex with men with new HIV-diagnosis: a 1-year prospective cohort study in Beijing, China[J]. Front Public Health, 2023, 11: 1244624. 10.3389/fpubh.2023.1244624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. 孟静, 程春晓, 林玉玺, 等. 基于潜类别增长模型的男男性行为人群HIV检测自我效能的趋势分析[J]. 中华预防医学杂志, 2023, 57(1): 29-34. 10.3760/cma.j.cn112150-20220509-00459. [DOI] [Google Scholar]; MENG Jing, CHENG Chunxiao, LIN Yuxi, et al. Tajectories of the self-efficacy of HIV testing among MSM based on latent class growth model[J]. Chinese Journal of Preventive Medicine, 2023, 57(1): 29-34. 10.3760/cma.j.cn112150-20220509-00459. [DOI] [PubMed] [Google Scholar]
- 43. Liu HH, Wang Y, Huang YL, et al. Ingestible sensor system for measuring, monitoring and enhancing adherence to antiretroviral therapy: an open-label, usual care-controlled, randomised trial[J]. EBio Medicine, 2022, 86: 104330. 10.1016/j.ebiom.2022.104330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Jovanović L, Šiljić M, Ćirković V, et al. Exploring evolutionary and transmission dynamics of HIV epidemic in Serbia: bridging socio-demographic with phylogenetic approach[J]. Front Microbiol, 2019, 10: 287. 10.3389/fmicb.2019.00287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Hsu HY, Lin JJH, Skidmore ST, et al. Evaluating fit indices in a multilevel latent growth curve model: a Monte Carlo study[J]. Behav Res Methods, 2019, 51(1): 172-194. 10.3758/s13428-018-1169-6. [DOI] [PubMed] [Google Scholar]
- 46. Nokes KM. Applying the chronic illness Trajectory Model to HIV/AIDS[J]. Sch Inq Nurs Pract, 1991, 5(3): 197-204. [PubMed] [Google Scholar]
- 47. 黄冰雪. 基于轨迹分析模型的MSM人乳头瘤病毒感染风险分类及影响因素分析[D]. 乌鲁木齐: 新疆医科大学, 2018. [Google Scholar]; HUANG Bingxue. Risk classification and influencing factors analysis of MSM human papillomavirus infection based on trajectory analysis model[D]. Urumqi: Xinjiang Medical University, 2018. [Google Scholar]
- 48. 林倚伊, 钟晓妮, 彭斌, 等. 基于轨迹分析模型的男男性行为人群无保护性行为特征分析[J]. 重庆医科大学学报, 2017, 42(2): 209-213. 10.13406/j.cnki.cyxb.000979. [DOI] [Google Scholar]; LIN Yiyi, ZHONG Xiaoni, PENG Bin, et al. Analysis of high-risk sexual behaviors among men who have sex with men based on trajectory model[J]. Journal of Chongqing Medical University, 2017, 42(2): 209-213. 10.13406/j.cnki.cyxb.000979. [DOI] [Google Scholar]
- 49. 黄冰雪, 桑国耀, 妥小青, 等. 轨迹分析模型在男男性行为人群人乳头瘤病毒感染状态变化趋势研究中的应用[J]. 浙江大学学报(医学版), 2018, 47(2): 150-155. 10.3785/j.issn.1008-9292.2018.04.07. [DOI] [Google Scholar]; HUANG Bingxue, SANG Guoyao, Xiaoqing TUO, et al. Trajectory modeling for estimating the trend of human papillomavirus infection status among men who have sex with men[J]. Journal of Zhejiang University. Medical Sciences, 2018, 47(2): 150-155. 10.3785/j.issn.1008-9292.2018.04.07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Popov M, Molsberry SA, Lecci F, et al. Brain structural correlates of trajectories to cognitive impairment in men with and without HIV disease[J]. Brain Imaging Behav, 2020, 14(3): 821-829. 10.1007/s11682-018-0026-7. [DOI] [PMC free article] [PubMed] [Google Scholar]