University of North Carolina-Chapel Hill Center for Health Promotion and Disease Prevention, Chapel Hill, NC, USA.
Department of Epidemiology, University of California-Berkeley, School of Public Health, Berkeley, CA, USA.
Cancer Causes Control. 2024 Jan;35(1):77-92. doi: 10.1007/s10552-023-01761-4. Epub 2023 Aug 10.
The association between diet quality, captured by the Mediterranean Diet Score (MDS), and mortality was studied among 1184 individuals diagnosed with head and neck cancer (HNC) who reflected on the year preceding diagnosis about their usual diet using National Cancer Institute's Diet History Questionnaire (DHQ).
Intakes of nine dietary components were scored and summed to construct the MDS (sample: median = 4; range (0-9); lower MDS reflected poorer diet quality; 5-year survival probability = 0.62). Cox regression estimated 5-year hazard ratios (HR) and 95% confidence intervals (95CI) for all-cause mortality and for HNC-specific death for contrasts of MDS quintiles. Effect measure modification (EMM) by tumor features [human papillomavirus (HPV) positivity; anatomic site] and sociodemographic behavioral factors [race, body mass index (BMI), smoking, alcohol consumption] was explored.
The 5-year [HR (95CI); P-trend] for all-cause mortality and HNC-specific mortality for highest versus lowest MDS quintile contrasts were [0.51 (0.33, 0.80); 0.014] and [0.43 (0.22, 0.85); 0.004], respectively. A unit increase in MDS adherence resulted in a 15% reduction of the 5-year HR for HNC-specific death for tumors located at the oral cavity [HR (95CI): 0.85 (0.75, 0.96)]. Poor diet quality (MDS ≤ 4) interacted with lower BMI (kg/m < 25) and separately with ever-using alcohol to produce 5-year HRs for all-cause and HNC-specific mortality that were statistically significantly larger than the sum of the individual HRs representing each combination (Poor diet quality + lower BMI; Poor diet quality + ever-using alcohol).
Greater adherence to a Mediterranean diet pattern prior to HNC diagnosis may reduce post-diagnosis mortality.
本研究通过使用国家癌症研究所饮食史问卷（DHQ），对 1184 名确诊为头颈部癌症（HNC）的患者进行了回顾性分析，调查了在诊断前一年，他们的饮食质量（通过地中海饮食评分（MDS）来评估）与死亡率之间的关系。
对九种饮食成分的摄入量进行评分和汇总，以构建 MDS（样本中位数=4；范围（0-9）；较低的 MDS 反映较差的饮食质量；5 年生存率=0.62）。Cox 回归估计了所有原因死亡率和 HNC 特异性死亡率的 5 年危险比（HR）和 95%置信区间（95%CI），并对比了 MDS 五分位数。通过肿瘤特征[人乳头瘤病毒（HPV）阳性；解剖部位]和社会人口学行为因素[种族、体重指数（BMI）、吸烟、饮酒]探讨了效果衡量修饰（EMM）。
对于最高与最低 MDS 五分位组的对比，5 年全因死亡率和 HNC 特异性死亡率的 [HR（95%CI）；P-趋势] 分别为 [0.51（0.33，0.80）；0.014] 和 [0.43（0.22，0.85）；0.004]。MDS 依从性增加一个单位，与口腔部位肿瘤的 HNC 特异性死亡的 5 年 HR 降低 15%[HR（95%CI）：0.85（0.75，0.96）]。较差的饮食质量（MDS≤4）与较低的 BMI（kg/m<25）和单独的饮酒相互作用，导致全因死亡率和 HNC 特异性死亡率的 5 年 HR 明显大于每个组合（较差的饮食质量+较低的 BMI；较差的饮食质量+饮酒）的单独 HR 之和。
在诊断 HNC 之前，更严格地遵循地中海饮食模式可能会降低诊断后的死亡率。
