The prevalence of late-onset patients in this study (10.63%) was comparable to the prevalence rates observed in many previous reports (2–17%) [8,9,10, 12,13,14,15,16,17,18,19,20,21, 23, 29, 30]. In addition, a very high prevalence (39%) was reported by Alonse et al. from Spain [31]. The lower female-to-male ratio of 5.9:1 in the late-onset patients, compared to the ratio of 13.8:1 in the early-onset patients in this study, was comparable to many previous reports (2.6–9.7:1 and 5.7–18.1:1, respectively) [9,10,11,12,13, 15, 16, 19, 21, 23, 24, 29,30,31]. Interestingly, reports from China by Choi et al. [14] and Egypt by Mehat et al. [20] found a higher female to male ratio in the late-onset group when compared with early-onset patients. The lower female-to-male ratio in the late-onset patients might be due partly to a lack of estrogenic hormones during the menopausal period, as estrogen can trigger the development of SLE in susceptible individuals [32].

Despite many studies reporting on the difference in clinical manifestations between the late- and early-onset SLE patients, information on the initial clinical presentation between the two groups was limited. This study found that mucocutaneous, articular and hematologic involvements were the common presenting symptoms, which was similar to a report from China by Boddaert et al., who found that polyarthritis, serositis and malar rashes were among the common manifestations in their late-onset patients [19]. However, it was difficult to conclude that mucocutaneous and articular manifestations were more common initial presentations in the late-onset patients than in the early-onset patients, as they were common clinical manifestations in both age groups.

The duration from first clinical symptoms to SLE diagnosis generally took longer in the late-onset patients than in the early-onset patients [10, 12, 17, 21, 33]. The diagnostic delay in late-onset SLE patients might be related partly to SLE being uncommon at ages ≥ 50 years. The clinical presentation (e.g., hematologic or cardiopulmonary involvement) might not be seen typically in early-onset patients, thus causing the physician to search for other more common diseases, particularly malignancies, that occur in elderly individuals. In this study, no difference between the two groups was found in the duration from first clinical symptoms to diagnosis, which was similar to a report by Alonso et al. [31].

The differences in clinical and laboratory manifestations, disease activity, treatment and outcomes between the late- and early-onset SLE patients have been determined by several groups. Unfortunately, a majority of these studies were not controlled but instead compared the clinical manifestations directly between late- and early-onset patients and mainly compared the cumulative clinical manifestations, which showed differences in the results. Of those studies in which information on clinical manifestations at diagnosis was available, the late-onset patients had less mucocutaneous [24, 29, 30, 33], vasculitis [30], renal [12, 29, 30], and central nervous system involvement [30], less hypocomplementemia [9, 24, 29, 30], less positive anti-dsDNA [24, 29], and lower mean number of ACR classification criteria [9] and SLEDAI-2K scores [29] than the early-onset patients.

Regarding cumulative manifestations at the last visit, the late-onset SLE patients had lower rates of mucocutaneous [10, 12, 17, 19, 21, 23, 33, 34], renal [10, 12, 16, 17, 19, 23, 33, 34], arthritic [17, 19], neurological [12, 17, 34] and hematologic involvement, and particularly less leukopenia and thrombocytopenia [10, 17, 34]. They also had less hypocomplementemia [9, 12, 16, 17, 19, 31], less positive anti-dsDNA [17, 31], less positive anti-Smith antibody (anti-Sm) [12, 17, 31], and a lower mean number of ACR classification criteria [9, 12, 19, 21] and SLEDAI-2K scores [9, 17]. In contrast, the late-onset patients had more Sicca or Sjogren’s symptoms [10, 16, 31] and cardiopulmonary involvement [10, 21] than the early-onset patients.

Among the few controlled studies (Supplementary Table 2), Bertoli et al. [13] used data from the LUMINA (Lupus in minorities: nature versus nurture) cohort and matched the early-onset patients by sex and disease duration (randomly selected). They found that the late-onset patients had more comorbidit