Supplementary Table 5. Threshold-gradient lag analyses excluding early osteoarthritis events. Threshold Lag period Weekend warrior, HR (95% CI); P Regularly active, HR (95% CI); P ≥ 50% 1-year lag 0.922 (0.874, 0.973); 0.003 0.908 (0.850, 0.970); 0.004 2-year lag 0.924 (0.872, 0.979); 0.007 0.907 (0.845, 0.974); 0.007 3-year lag 0.930 (0.873, 0.991); 0.024 0.903 (0.836, 0.976); 0.010 ≥ 55% 1-year lag 0.924 (0.872, 0.979); 0.008 0.912 (0.861, 0.966); 0.002 2-year lag 0.913 (0.857, 0.972); 0.005 0.924 (0.868, 0.983); 0.012 3-year lag 0.918 (0.857, 0.983); 0.015 0.924 (0.863, 0.989); 0.022 ≥ 60% 1-year lag 0.910 (0.854, 0.969); 0.003 0.922 (0.873, 0.974); 0.004 2-year lag 0.898 (0.839, 0.962); 0.002 0.931 (0.878, 0.987); 0.017 3-year lag 0.908 (0.843, 0.978); 0.011 0.929 (0.871, 0.991); 0.025 ≥ 65% 1-year lag 0.910 (0.846, 0.978); 0.010 0.920 (0.873, 0.970); 0.002 2-year lag 0.901 (0.834, 0.975); 0.009 0.924 (0.874, 0.978); 0.007 3-year lag 0.907 (0.833, 0.987); 0.024 0.926 (0.870, 0.985); 0.015 ≥ 70% 1-year lag 0.935 (0.862, 1.014); 0.106 0.914 (0.868, 0.962); < 0.001 2-year lag 0.920 (0.842, 1.004); 0.063 0.918 (0.869, 0.971); 0.003 3-year lag 0.928 (0.843, 1.022); 0.129 0.919 (0.865, 0.977); 0.007 ≥ 75% 1-year lag 0.949 (0.863, 1.044); 0.286 0.913 (0.868, 0.961); < 0.001 2-year lag 0.923 (0.832, 1.024); 0.130 0.918 (0.869, 0.970); 0.002 3-year lag 0.943 (0.843, 1.055); 0.305 0.918 (0.865, 0.975); 0.005 ≥ 80% 1-year lag 0.901 (0.801, 1.014); 0.083 0.919 (0.874, 0.967); 0.001 2-year lag 0.876 (0.770, 0.997); 0.044 0.922 (0.874, 0.974); 0.003 3-year lag 0.879 (0.764, 1.012); 0.073 0.925 (0.871, 0.981); 0.009 ≥ 85% 1-year lag 0.878 (0.751, 1.026); 0.101 0.920 (0.875, 0.967); < 0.001 2-year lag 0.875 (0.739, 1.035); 0.119 0.921 (0.872, 0.972); 0.003 3-year lag 0.912 (0.761, 1.093); 0.319 0.921 (0.869, 0.977); 0.006 ≥ 90% 1-year lag 0.782 (0.637, 0.959); 0.018 0.922 (0.877, 0.969); 0.001 2-year lag 0.794 (0.637, 0.988); 0.039 0.922 (0.874, 0.973); 0.003 3-year lag 0.844 (0.669, 1.064); 0.151 0.923 (0.871, 0.979); 0.008 ≥ 95% 1-year lag 0.906 (0.679, 1.210); 0.504 0.918 (0.873, 0.965); < 0.001 2-year lag 0.852 (0.618, 1.174); 0.327 0.919 (0.871, 0.970); 0.002 3-year lag 0.907 (0.646, 1.274); 0.575 0.921 (0.869, 0.977); 0.006 HRs were estimated using Cox proportional hazards models after excluding participants with follow-up of no more than the stated 1-, 2-, 3-year lag interval, thereby excluding OA events occurring within that interval. The exposure was the physical activity pattern defined separately at each top-two-day threshold: inactive (<150 min/week of MVPA), WW (>=150 min/week and at least the stated percentage of total weekly MVPA on the two most active days), or regularly active (>=150 min/week and below the stated percentage). Inactive participants were the reference group. The outcome was incident overall OA, defined as the first recorded ICD-10 M15-M19 diagnosis during follow-up among participants without OA at baseline. Models adjusted for age at accelerometer assessment, sex, ethnicity, education, Townsend deprivation index, smoking status, alcohol intake frequency, and BMI. Abbreviations: BMI, body mass index; CI, confidence interval; HR, hazard ratio; ICD-10, International Classification of Diseases, 10th Revision; MVPA, moderate-to-vigorous physical activity; OA, osteoarthritis; WW, weekend warrior.