Military personnel, as a group executing special missions, have sleep health that directly affects psychological adaptability and combat effectiveness. Sleep reactivity is a key indicator reflecting an individual’s vulnerability to sleep disturbances under stress or stimulation. However, existing studies have mostly adopted a variable-centered perspective, neglecting the possible heterogeneous categories of sleep reactivity within the military population, and the moderating mechanisms of different categories in the emotion-sleep relationship remain unclear. This study aims to explore the latent categories of sleep reactivity in military personnel and analyze the moderating effects of different categories on the relationship between emotional states and sleep quality.

A cross-sectional study design was adopted. A total of 1352 military personnel from a certain unit were recruited as participants using convenience sampling from March to April 2025. The survey instruments included Ford Insomnia Response to Stress Test-Chinese Version (FIRST-C), Positive Affect Subscale from the Positive and Negative Affect Schedule (PANAS), Anxiety subscale from the Depression Anxiety Stress Scales (DASS), and the Insomnia Severity Index(ISI). Spearman correlation analysis was used to examine the correlations among the study variables; Latent profile analysis (LPA) was employed to explore the latent categories of sleep reactivity in military personnel; Analysis of variance (ANOVA) was utilized to compare differences in sleep problems, positive affect, and anxiety among military personnel with different categories of sleep reactivity. Interaction terms were constructed to analyze the moderating effects of the high, moderate, and low reactivity groups in the relationship between positive affect and sleep, as well as between anxiety and sleep,

Sleep reactivity was significantly positively correlated with total ISI score (r=0.546, 95%CI: 0.506 to 0.583, P<0.01) and anxiety (r=0.390, 95%CI: 0.343 to 0.436, P<0.01), and also negatively correlated with positive affect (r=-0.309, 95%CI: -0.358 to =0.259, P<0.01). Total ISI score was negatively correlated with positive affect (r=-0.348, 95%CI: -0.396 to -0.299, P<0.01), and positively with anxiety (r=0.454, 95%CI: 0.410 to 0.497, P<0.01). Positive affect was negatively correlated with anxiety (r=-0.364, 95%CI: -0.410 to -0.315, P<0.01). Sleep reactivity among military personnel could be classified into 3 categories: high, moderate, and low reactivity, accounting for 6.0%, 41.7%, and 52.3% of the total participants, respectively. The high reactivity group exhibited significantly higher ISI scores (mean difference: high vs low =7.217, 95%CI: 6.425 to 8.009, P<0.001; high vs medium =5.047, 95%CI: 4.244 to 5.849, P<0.001) and anxiety scores (mean difference: high vs low =6.515, 95%CI: 6.209 to 6.819, P<0.001; high vs medium =5.770, 95%CI: 5.461 to 6.079, P<0.001) than both the medium and low reactivity groups. Compared to the low reactivity group, the medium reactivity group had significantly higher ISI scores (mean difference: medium vs low =2.170, 95%CI: 1.789 to 2.551, P<0.001) and anxiety scores (mean difference: =0.745, 95%CI: 0.597 to 0.891, P<0.001). Regarding positive affect scores, the high reactivity group scored significantly lower than the low and medium reactivity groups (high vs low =-16.905, 95%CI: -18.114 to -15.695, P<0.001; high vs medium =-2.692, 95%CI: -3.917 to -1.466, P<0.001), and the medium reactivity group scored significantly lower than the low reactivity group (mean difference =-14.212, 95%CI: -14.795 to -13.630, P<0.001). The high reactivity group showed a greater impact of anxiety on sleep than the low reactivity group (β=-0.213, 95%CI: -0.366 to -0.059, P<0.01). In the effect of positive affect on sleep, the interaction term between positive affect and the moderate reactivity group had no significant regression effect on ISI (β=-0.073, 95%CI: -0.249 to 0.103), and the interaction term between positive affect and the high reactivity group also had no significant regression effect on ISI (β=-0.145, 95%CI: -0.383 to 0.093).

Sleep reactivity among military personnel can be categorized into 3 types: low reactivity, moderate reactivity and high reactivity. Compared with the military personnel with low sleep reactivity, those with high reactivity have sleep quality more vulnerable to anxiety. This finding suggests that in military mental health practice, priority should be given to anxiety management for those with high sleep reactivity in order to reduce their risk of insomnia.