Weight Fluctuations Tied to Higher Dementia Risk, Meta-Analysis Shows

Understanding a New Clue in Dementia Prevention

Dementia is one of the most pressing global health challenges, affecting millions of older adults and placing heavy burdens on families and healthcare systems. As the world’s population ages, researchers continue searching for modifiable risk factors that might help delay or prevent the onset of dementia.

A new systematic review and meta-analysis has now provided evidence that large fluctuations in body weight or body mass index (BMI) over time — known as weight variability — may be linked to a higher risk of developing dementia.

The study, registered with PROSPERO and conducted according to PRISMA guidelines, brings together data from nearly 4.23 million participants across nine cohort studies. Its conclusion is striking: people with the highest levels of body weight variability had a 36% higher risk of dementia compared with those whose weight remained more stable.

Why Look at Weight Variability?

The authors note that previous research on body weight and dementia risk has often focused on a person’s average weight or whether they were underweight, normal weight, or obese. However, many individuals experience weight fluctuations due to lifestyle changes, health conditions, or aging.

“Body weight variability has been recognized as an indicator of metabolic instability and long-term health outcomes,” the paper explains. The authors aimed to clarify whether such fluctuations could also be an early warning sign for dementia.

How the Study Was Done

The researchers searched PubMed, Embase, and Web of Science databases up to March 2025. They included cohort studies that examined long-term weight variability and later dementia diagnosis in adults. Studies that duplicated data, had unclear variability definitions, or lacked risk estimates were excluded.

In total, ten datasets from nine independent cohorts were analyzed, covering populations from Korea, Israel, the United States, and Australia. Follow-up periods ranged up to 36 years, and most studies adjusted for age, sex, and health behaviors.

Weight variability was measured using several indicators — such as standard deviation (SD), coefficient of variation (CV), average successive variability (ASV), and variability independent of the mean (VIM). Participants were typically grouped into the highest versus lowest categories of weight fluctuation.

Main Findings: Greater Weight Swings, Higher Dementia Risk

Across all studies, high weight variability was significantly associated with dementia incidence. The pooled relative risk (RR) was 1.36 (95% CI 1.27–1.46).

When weight change was measured by body weight alone, the risk was slightly higher (RR 1.45), and when measured by BMI variability, it remained significant (RR 1.34).

The link held true for both Alzheimer’s disease (RR 1.32) and vascular dementia (RR 1.40). Results were consistent across sex, follow-up duration, and study design, though some heterogeneity was observed (I² = 84%).

Interestingly, studies that did not adjust for baseline body weight or BMI showed stronger associations (RR 1.66) than those that did (RR 1.32). This suggests that both average weight and weight stability matter in understanding dementia risk.

Consistent Trends Across Populations

Despite differences in study design and population size, most included cohorts reported similar trends. For instance, the Korean National Health Insurance Service studies contributed a large portion of participants but were supported by findings from Western cohorts.

Prospective studies — those that followed participants forward in time — showed a higher relative risk (1.51) than retrospective designs (1.27). The authors interpret this as additional support for the observed association.

No Evidence of Publication Bias

The analysis found no indication that small or negative studies were missing from the literature. A visual inspection of the funnel plot appeared symmetrical, and the Egger’s test did not detect bias (p = 0.22). This strengthens confidence in the overall findings.

Possible Explanations and Limitations

The authors discuss several possible interpretations. Weight fluctuations may reflect underlying metabolic or inflammatory processes that also affect brain health. Alternatively, reverse causality cannot be ruled out — early cognitive changes might lead to unintentional weight loss or gain years before dementia diagnosis.

Because dementia develops gradually, the study notes that “reverse causation cannot be completely excluded.” Some included studies lacked long-term lag analyses, meaning that early-stage dementia might have influenced body weight variability.

Heterogeneity among studies also remains high. Differences in measurement methods (e.g., SD, CV, ASV, VIM), population characteristics, and diagnostic criteria could contribute to variation in results. Still, sensitivity analyses — including leave-one-out tests — confirmed the stability of the overall conclusion.

Practical Implications

Although the research does not prove that weight fluctuations directly cause dementia, it highlights weight stability as a potential marker of healthy aging.

Maintaining relatively consistent body weight might reflect better metabolic regulation, nutrition, and lifestyle stability — all factors known to influence cognitive health.

The authors suggest that “body weight variability could serve as an early indicator for dementia risk,” and that clinicians should consider monitoring weight patterns, not just static BMI values, when assessing long-term health.

Strengths and Novelty of the Study

This is the first meta-analysis to comprehensively examine the link between body weight variability and dementia risk. It includes the largest pooled sample to date, a rigorous search strategy, and thorough sensitivity analyses.

The study adheres to PRISMA standards, assesses bias using the Newcastle–Ottawa Scale, and conducts subgroup analyses for dementia subtype, follow-up length, and exposure metric.

By synthesizing a vast dataset, the paper provides a clearer, evidence-based picture of an emerging risk factor that has been largely overlooked in dementia research.

Looking Ahead

While further studies are needed to confirm causality and explore mechanisms, the findings add a valuable piece to the complex puzzle of dementia prevention. Long-term stability in body weight might become one of several lifestyle markers to track in midlife and older adulthood.

The authors conclude:

“High intra-individual variability in body weight or BMI is associated with an increased risk of dementia.”

They emphasize that more prospective studies with standardized measures and lag-time analyses are necessary to confirm this link and to determine whether maintaining stable weight can help protect against cognitive decline.

Takeaway for Researchers

• High weight variability → 36% higher dementia risk (RR 1.36).

• Consistent across body weight and BMI, Alzheimer’s and vascular dementia.

• Findings robust across subgroups and study designs.

• Heterogeneity remains high; reverse causation possible.

• Weight stability may represent a modifiable indicator of long-term brain health.

 

The translation of the preceding English text in Chinese:

 

理解预防痴呆的新线索

痴呆症是全球最紧迫的健康挑战之一，影响着数以百万计的老年人，并对家庭和医疗系统造成沉重负担。随着全球人口老龄化的加速，研究人员持续寻找可改变的风险因素，以帮助延缓或预防痴呆的发生。

一项新的系统综述和荟萃分析提供了新的证据：长期内体重或体质指数（BMI）的剧烈波动——即所谓的体重变异性——可能与更高的痴呆风险相关。

该研究已在 PROSPERO 注册，并遵循 PRISMA 指南进行，汇总了来自九项队列研究、约423万名参与者的数据。研究结论引人注目：体重波动最大的个体，其患痴呆的风险比体重较稳定者高出36%。

为什么要研究体重变异性？

作者指出，以往关于体重与痴呆风险的研究通常关注个体的平均体重，或其是否处于体重过轻、正常或肥胖状态。然而，许多人在生活方式变化、健康状况波动或衰老过程中会经历体重起伏。

论文中写道：“体重变异性已被认为是代谢不稳定性和长期健康结局的指标。” 作者旨在探讨这种体重波动是否也可能是痴呆的早期预警信号。

研究是如何进行的

研究人员检索了截至2025年3月的 PubMed、Embase 和 Web of Science 数据库，纳入了研究成年人体重长期波动与随后痴呆诊断关系的队列研究。重复数据、体重变异定义不清或缺乏风险估计的研究被排除。

最终，共纳入来自九个独立队列的十个数据集，覆盖韩国、以色列、美国和澳大利亚的人群。随访时间最长达36年，大多数研究调整了年龄、性别和健康行为等变量。

体重变异通过多种指标衡量——如标准差（SD）、变异系数（CV）、平均连续变异（ASV） 和 与均值无关的变异（VIM）。参与者通常被分为体重波动最大的组与最小的组进行比较。

主要发现：体重波动越大，痴呆风险越高

在所有研究中，高体重变异性与痴呆发病显著相关。合并的相对风险（RR）为1.36（95% CI 1.27–1.46）。

当以体重变化衡量时，风险略高（RR 1.45）；以 BMI 波动衡量时，风险仍显著（RR 1.34）。

这一关联在**阿尔茨海默病（RR 1.32）和血管性痴呆（RR 1.40）**中均得到验证。结果在性别、随访时间及研究设计上保持一致，尽管存在一定异质性（I² = 84%）。

值得注意的是，未调整基线体重或BMI的研究显示更强的关联（RR 1.66），而调整后的研究为（RR 1.32）。这表明，平均体重与体重稳定性在理解痴呆风险时都具有重要意义。

不同人群中的一致趋势

尽管研究设计与样本规模各异，大多数队列研究报告了相似的趋势。例如，韩国国家健康保险服务研究贡献了大量样本，但其结果也得到西方人群研究的支持。

前瞻性研究（即随访参与者的研究）显示的相对风险更高（1.51），而回顾性研究为1.27。作者将此视为对观察到的关联的进一步支持。

未发现发表偏倚的证据

分析结果显示，文献中没有遗漏小样本或阴性研究的迹象。漏斗图外观对称，Egger 检验未发现偏倚（p = 0.22），这增强了对整体结论的信心。

可能的解释与局限性

作者讨论了几种可能的解释。体重波动可能反映了潜在的代谢或炎症过程，这些过程也会影响脑部健康。另一方面，反向因果关系不能被完全排除——早期认知变化可能在确诊痴呆前多年导致无意的体重增加或减少。

由于痴呆发展缓慢，研究指出：“反向因果关系无法完全排除。” 部分研究缺乏长期滞后分析，意味着痴呆早期阶段可能已影响体重波动。

研究间的异质性依然较高。测量方法（如SD、CV、ASV、VIM）、人群特征及诊断标准的差异，可能导致结果差异。然而，敏感性分析（包括逐一剔除研究的测试）证实了总体结论的稳定性。

实践意义

虽然该研究未能证明体重波动直接导致痴呆，但结果强调了体重稳定性作为健康老龄化潜在标志的重要性。

保持相对稳定的体重可能反映出更好的代谢调节、营养状况及生活方式稳定性——这些因素均被认为影响认知健康。

作者指出：“体重变异性可作为痴呆风险的早期指标。” 医务人员在评估长期健康时，应关注体重变化模式，而不仅仅是静态的BMI数值。

研究的优势与创新性

这是首个系统探讨体重变异性与痴呆风险关联的荟萃分析。研究样本量最大，采用严格的检索策略和多重敏感性分析。

研究遵循 PRISMA 标准，使用 Newcastle–Ottawa 量表评估偏倚，并对痴呆类型、随访年限及暴露指标进行了亚组分析。

通过整合庞大的数据集，本文为一个在痴呆研究中长期被忽视的新风险因素提供了更清晰的、基于证据的图景。

展望未来

尽管仍需更多研究来确认因果关系并探讨潜在机制，这项研究的结果为痴呆预防的复杂拼图增添了重要一块。长期保持体重稳定可能成为中年及老年阶段可追踪的生活方式指标之一。

作者总结道：
“个体内体重或BMI的高变异性与痴呆风险增加相关。”

他们强调，需要更多采用标准化测量及滞后分析的前瞻性研究，以验证这一关联，并确定保持体重稳定是否有助于防止认知能力下降。

研究者要点总结

• 高体重变异性 → 痴呆风险增加36%（RR 1.36）

• 在体重与BMI、阿尔茨海默病与血管性痴呆中结果一致

• 结论在不同亚组和研究设计中均具稳健性

• 异质性仍高，存在反向因果可能

• 体重稳定性可能是长期脑健康的可干预指标

---

Reference:

Sitian Fang, Lewei Guan, Huimin Jian, Xi-jian Dai, Lianggeng Gong