Scholarlink

Background <p>Seasonal variation in glycemic control is well documented in temperate settings, but evidence from tropical, high-humidity environments is limited.</p> Objective <p>We examined intra-annual fluctuations in fasting plasma glucose (FPG) among adults with type 2 diabetes in Mumbai, India.</p> Methods <p>We conducted a retrospective cohort analysis of 835 patients (2011–2020). Weekly mean FPG values were decomposed into trend, seasonal, and residual components. Seasonal dynamics were modelled using SARIMA; candidate models were compared by AIC/BIC and residual diagnostics. Predictive validity (January–August 2019) was summarised with MAE, RMSE, and MAPE. To relate glycemia to climate, we fit a generalised additive mixed model (GAMM) with thin-plate splines for temperature–rainfall, adjusting for season, sex, age, and comorbidity.</p> Results <p>Time-series decomposition revealed a steady upward drift in FPG until 2016, after which levels stabilised, overlaid with a distinct and recurring annual cycle. The best-fitting SARIMA model demonstrated pronounced short-term and seasonal dependencies, with a significant MA (1) effect (−0.968, <i>p</i> &lt; 0.001) and strong seasonal autoregressive terms at 52 and 104&#xa0;weeks (−0.515 and −0.332, both <i>p</i> &lt; 0.001). Non-seasonal AR terms were not significant (<i>p</i> &gt; 0.50). Model performance was robust in out-of-sample validation (MAE = 0.627; RMSE = 2.843; MAPE = 0.494 per cent). Complementary GAMM analysis reinforced this pattern, showing consistent FPG declines during the monsoon (June–September) and pronounced peaks during the pre- and post-monsoon months of April–May and October–November.</p> Conclusions <p>The study found that FPG follows a seasonal cycle, peaking in hot pre-/post-monsoon months and dipping during the cooler monsoon, with an overall upward drift over time. These patterns highlight the need for season-aware diabetes management, including adjusted targets and proactive interventions before high-risk periods.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Seasonal variations in fasting plasma glucose among individuals with type 2 diabetes in Mumbai, India

  • 期刊论文

  • 发表日期：

    2025年11月17日

查看全文
  • Puja Goswami,
  • Dilip Thandassery,
  • Abdul Fathah,
  • Yogesh Shejul,
  • Anjali Kulkarni

摘要

Background

Seasonal variation in glycemic control is well documented in temperate settings, but evidence from tropical, high-humidity environments is limited.

Objective

We examined intra-annual fluctuations in fasting plasma glucose (FPG) among adults with type 2 diabetes in Mumbai, India.

Methods

We conducted a retrospective cohort analysis of 835 patients (2011–2020). Weekly mean FPG values were decomposed into trend, seasonal, and residual components. Seasonal dynamics were modelled using SARIMA; candidate models were compared by AIC/BIC and residual diagnostics. Predictive validity (January–August 2019) was summarised with MAE, RMSE, and MAPE. To relate glycemia to climate, we fit a generalised additive mixed model (GAMM) with thin-plate splines for temperature–rainfall, adjusting for season, sex, age, and comorbidity.

Results

Time-series decomposition revealed a steady upward drift in FPG until 2016, after which levels stabilised, overlaid with a distinct and recurring annual cycle. The best-fitting SARIMA model demonstrated pronounced short-term and seasonal dependencies, with a significant MA (1) effect (−0.968, p < 0.001) and strong seasonal autoregressive terms at 52 and 104 weeks (−0.515 and −0.332, both p < 0.001). Non-seasonal AR terms were not significant (p > 0.50). Model performance was robust in out-of-sample validation (MAE = 0.627; RMSE = 2.843; MAPE = 0.494 per cent). Complementary GAMM analysis reinforced this pattern, showing consistent FPG declines during the monsoon (June–September) and pronounced peaks during the pre- and post-monsoon months of April–May and October–November.

Conclusions

The study found that FPG follows a seasonal cycle, peaking in hot pre-/post-monsoon months and dipping during the cooler monsoon, with an overall upward drift over time. These patterns highlight the need for season-aware diabetes management, including adjusted targets and proactive interventions before high-risk periods.