Background <p>High-intensity interval training (HIIT) was originally designed to improve athletic performance, but a growing body of research over the past decade has highlighted its positive impact on various other health outcomes. However, concerns exist regarding HIIT’s suitability for those unaccustomed to regular exercise, as its high intensity may impact tolerance and adherence. Therefore, this study aimed to assess the associations between various objective adherence metrics and the resultant changes in peak oxygen uptake (V̇O_2PEAK) and lactate threshold (LT) in older adults completing six-month home-based HIIT.</p> Methods <p>Healthy older adults (<i>n</i> = 233, 60–84 years, 54% female) were randomized to six-month, thrice-weekly home-based HIIT or a passive control group. Adherence in the HIIT group was objectively monitored using a Polar watch and heart rate sensor and quantified as frequency (sessions/week), intensity (minutes ≥ 80% HR_PEAK/session), duration (session duration, min), and total adherence (cumulative MET-min; overall exercise volume summed across sessions from session duration and mean %HR_PEAK). For each metric, adherence was expressed as a percentage of completion relative to the planned amount. V̇O_2PEAK and LT were assessed using a modified Balke treadmill protocol to volitional exhaustion. To account for multicollinearity, partial least squares regression (PLSR) models assessed associations between adherence metrics and changes in V̇O_2PEAK or LT.</p> Results <p>The PLSR analysis, accounting for baseline V̇O_2PEAK, age, sex, and country where data-collection took place, revealed a positive association between post-test V̇O_2PEAK and total adherence (Selectivity fractions (SF) = 0.70 [0.19; 0.93]), frequency adherence (SF = 0.54 [0.08; 0.86]), and intensity adherence (SF = 0.44 [0.07; 0.80]), but not with duration adherence (SF = 0.11 [− 0.08; 0.54]). For LT, PLSR analysis revealed no associations with any adherence metrics.</p> Conclusions <p>Superior HIIT adherence (i.e., total, frequency, and intensity) was associated with larger V̇O_2PEAK gains in older adults. Notably, total adherence demonstrated the strongest predictive contribution among the adherence metrics, suggesting that overall exercise volume may be especially relevant for improving V̇O_2PEAK in this context. In contrast, adherence metrics did not predict changes in LT, despite the exercise group experiencing significant LT improvements, indicating that LT adaptations were not strongly explained by variation in HIIT adherence in this sample.</p> Trial registration <p>ClinicalTrials.gov NCT07443189 (Registration date: 02.03.2026, retrospectively registered).</p>