Patterns of neuromyelitis optica spectrum disorder attacks in different age groups and sexes depending on the status of immunosuppressive therapy: A retrospective cohort study

Abstract

Background and purpose

The association between onset age and sex with relapse risk in neuromyelitis optica spectrum disorder (NMOSD) remains inconclusive. We aimed to describe the clinical features of patients with NMOSD in different age groups and sexes and to analyse relapse characteristics pre‐ and post‐immunosuppressive therapy (IST).

Methods

Patients with NMOSD were retrospectively reviewed from our clinical centre's database. Demographic and clinical data, attack presentation, and disease course pre‐ and post‐IST were investigated. We also analysed the effect of onset age on the annualized relapse rate and relapse risk according to sex and IST status. Interactions on the additive scale between onset age and sex were analysed. A restricted cubic spline was used to analyse potential nonlinear correlations. Longitudinal changes in the Expanded Disability Status Scale score across NMOSD attacks were analysed using linear mixed‐effect models.

Results

In total, 533 patients experienced 1394 attacks pre‐IST and 753 relapses post‐IST. Older age at onset was correlated with more myelitis attacks but fewer optic neuritis attacks, with no sex‐related differences in attack presentation. Pre‐IST, relapse risk increased with age at onset in women, while a U‐shaped correlation between onset age and relapse risk was found in men. Post‐IST, an inverted U‐shaped association between the predicted relapse risk and onset age was observed in women. Conversely, a negative correlation between the predicted relapse risk and onset age was found in men. Overall, a higher ratio of myelitis attacks was found post‐IST.

Conclusions

Patients of different onset ages and sexes had different relapse patterns before and after IST.

INTRODUCTION

Neuromyelitis optica spectrum disorder (NMOSD) is an autoimmune inflammatory disease of the central nervous system [1]. NMOSD is mediated by a pathogenic immunoglobulin G (IgG) antibody to aquaporin 4 (AQP4) in more than 80% of cases [2]. Although the pathology of AQP4‐seronegative NMOSD is still unclear, the clinical features of AQP4‐seropositive NMOSD and AQP4‐seronegative NMOSD are quite different [3, 4, 5]. NMOSD is characterized by a high rate of relapse and disability. The disability in patients with NMOSD is relapse‐dependent [6]; thus, it is of vital importance to prevent relapses.

Many studies have described the association between age at onset, sex, and relapse in patients with NMOSD; however, inconsistent results have been revealed. Most previous studies have focused on patients who have received maintenance therapy, while the natural history of the disease in patients has been less studied. In the present study, we aimed to describe the demographic and clinical features of the natural history in patients with NMOSD in different age groups and sexes and to analyse the changes in relapse patterns after immunosuppression therapy (IST).

METHODS

Study design and patients

In this retrospective cohort study, we reviewed patients diagnosed with NMOSD at the Department of Neurology, West China Hospital, between January 2014 and September 2022. The inclusion criteria were as follows: (i) NMOSD diagnosis according to the 2015 diagnostic criteria [7] with AQP4‐IgG seropositivity and (ii) complete clinical data with detailed records of each relapse. Patient exclusion criteria were as follows: (i) age of onset < 18 years; (ii) refusal to undergo follow‐up; and (iii) uncertain treatment records. In the analysis of risk of relapse, patients with a monophasic course and a disease duration in the natural history of < 12 months were excluded. As pregnancy is a strong risk factor for NMOSD relapses [8], patients who experienced pregnancy‐related NMOSD attacks were all excluded to eliminate the effect of pregnancy. In the analysis of the disease course after IST, patients who did not receive IST were excluded.

Data collection

All patients diagnosed with NMOSD were registered in our database. Face‐to‐face follow‐up was regularly performed every 3–6 months; the features and presentation of each relapse were recorded, and the Expanded Disability Status Scale (EDSS) score was accessed by a trained neurologist. Data on previous therapeutic regimens, including the date of initiation of maintenance IST, were collected. IST included the most commonly used immunosuppressants at our centre, such as rituximab 2000 mg/year, azathioprine 2–3 mg/kg/day, and mycophenolate mofetil 1000–1500 mg/day, with or without oral glucocorticoids. All patients underwent AQP4 serology using a commercial cell‐based assay (EUROIMMUN AG, Luebeck, Germany) [9, 10].

Outcome measurement

The disease duration before and after IST was calculated according to the date of onset, date of IST initiation, and date of the last follow‐up. The annualized relapse rate (ARR; number of relapses per patient‐year) was calculated based on the recorded information, and the first attack of onset was not considered in the calculation of the ARR. Recurrent relapses were defined as all NMOSD attacks that occurred after the initial onset. A relapse was defined as a new worsening of neurological function lasting more than 24 h in the absence of other identifiable causes and occurring more than 30 days after a previous attack or a new lesion confirmed via magnetic resonance imaging. A pregnancy‐related attack was defined as an NMOSD attack during the pregnancy or within 1 year after delivery or abortion.

Statistical analysis

To describe the demographic and clinical features of different age groups, patients were divided into four groups according to age at onset: 18–29 years, 30–39 years, 40–49 years, and ≥50 years. Categorical variables are presented as frequencies (percentages). Continuous variables are presented as mean ± standard deviation (SD) or, if they are not normally distributed, as median (interquartile range). The Kolmogorov–Smirnov test was applied to verify the normal distribution of continuous variables. The Kruskal–Wallis test or analysis of variance was used to analyse the differences in means and medians among groups; we did not perform multiple comparisons as the study was exploratory.

We also used a generalized linear model to explore the associations among onset age, sex, and ARR. The Anderson–Gill (AG) proportional hazard model was used to analyse recurrent events in terms of ‘time to subsequent relapse’ [11]; restricted cubic splines with four knots at the 5th, 35th, 65th and 95th centiles were used to flexibly model the association between onset age and recurrent relapses in women and men, respectively. We repeated the above analysis to explore the impact of age at onset on recurrent relapse in each sex after IST. Sex, presentation at onset, initial severe attack (defined as visual acuity equal to or worse than 20/200 [12, 13], or EDSS score > 6.0 [14]), ARR before IST, and active relapse before IST (defined as at least two relapses within 1 year [15]) were adjusted as covariates in the multivariable analysis.

Finally, we divided patients into two groups: an early‐onset NMOSD (EO‐NMOSD) group (onset age < 50 years) and a late‐onset NMOSD (LO‐NMOSD) group (onset age ≥ 50 years). We examined interaction on the additive scale based on an epidemiological definition [16]. This interaction was reported according to the four steps recommended by Knol et al. [17] and the relative excess risk due to the interaction (RERI) was reported; a RERI of 0 indicates perfect additivity (no interaction) [18, 19]. For different IST statuses, we analysed the longitudinal change in EDSS score over relapses in patients with EO‐NMOSD and LO‐NMOSD using linear mixed‐effect models, where the identity of each patient and the type of attack were adjusted as random intercepts. Sex, an initial severe attack, ARR before IST, and EDSS score at IST initiation were controlled as fixed effects (if necessary).

Some events were analysed together to improve the power and robustness of the statistical analysis: unilateral optic neuritis (ON) and bilateral ON were analysed together; attacks of the cerebral and brain stem were analysed together because the number of cerebral attacks was insufficient.

All statistical analyses were performed using R, V.3.6.2. Statistical significance was set at a p value < 0.05.

Standard protocol approvals, registrations, and patient consents

This study was approved by the Ethics Committee of Sichuan University (approval number: 2018[Shen]29) and was performed in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all the patients.

RESULTS

Study population

Among 646 patients with NMOSD, 68 were excluded for being AQP4‐IgG seronegative and 45 were excluded for being aged <18 years; ultimately, 533 patients with 1394 attacks and 861 relapses before IST (with the first attack excluded) were included in the analysis of demographic and clinical characteristics in the disease's natural history. Since the results of survival analysis can be affected by over‐censoring, we excluded 67 patients with a monophasic course and disease duration in the natural history of < 12 months. Nine women with 27 relapses were then excluded for having pregnancy‐related attacks. Thus, 457 patients with 834 relapses were included in the analysis of recurrent relapse before IST and the interaction on the additive scale. To analyse the risk of recurrent relapse after IST, we excluded five patients who did not receive any IST; thus, 452 patients with 753 relapses after IST were included in the analysis of the risk of recurrent relapse after IST. Figure 1 shows the inclusion and exclusion criteria.

FIGURE 1.

Flow chart for the inclusion and exclusion of patients. AQP4, aquaporin 4; IgG, immunoglobulin G; IST, immunosuppression therapy.

Clinical features in the natural history

Transverse myelitis (TM) was the most common presentation of the first attack, followed by ON, and 115 patients (22%) presented with multifocal onset (Table 1). The presentation of the first attack differed among patients in different age groups: ON onset was more common in younger patients, while the opposite result was observed in the onset of TM. The ratios of area postrema syndrome (APS), brain stem/cerebral syndrome (BS), and multifocal onset did not differ among age groups. No differences were found in the presentation of the first attack between women and men (Table S3).

TABLE 1.

Demographic and clinical data on the natural history of patients according to age group.

Characteristic	Total, N = 524	18–29 years, N = 120	30–39 years, N = 113	40–49 years, N = 158	≥50 years, N = 133	p value
Onset age, years, mean (SD)	41 (13)	24 (4)	35 (3)	45 (3)	57 (7)	<0.001
Sex, n (%)						0.979
Female	467 (89)	107 (89)	102 (90)	140 (89)	118 (89)
Male	57 (11)	13 (11)	11 (9.7)	18 (11)	15 (11)
Presentation of the first onset, n (%)
ON	204 (39)	61 (51)	38 (34)	59 (37)	46 (35)	0.023
TM	314 (60)	59 (49)	67 (59)	97 (61)	91 (68)	0.02
APS	92 (18)	26 (22)	19 (17)	32 (20)	15 (11)	0.107
BS	57 (11)	10 (8.3)	14 (12)	20 (13)	13 (9.8)	0.636
Multifocal	115 (22)	31 (26)	20 (18)	36 (23)	28 (21)	0.499
Clinical course
Initial severe attack, n (%)	158 (30)	40 (33)	28 (25)	44 (28)	46 (35)	0.287
EDSS score at first attack, mean (SD)	3.16 (2.05)	2.96 (1.75)	2.82 (1.87)	3.13 (2.03)	3.67 (2.39)	0.02
ARR before IST, median (IQR)	0.43 (0.00, 1.00)	0.35 (0.07, 1.00)	0.51 (0.15, 1.00)	0.51 (0.00, 1.00)	0.43 (0.00, 1.00)	0.443
Active relapse, n (%)	264 (50)	60 (50)	69 (61)	76 (48)	59 (44)	0.059

Abbreviations: APS, area postrema syndrome; ARR, annualized relapse rate; BS, brain stem/cerebral syndrome; IST, immunosuppression therapy; ON, optic neuritis; TM, transverse myelitis.

In the natural history, a total of 1394 NMOSD attacks (first attack included) were recorded. The analysis of attacks before IST showed a higher ratio of ON and TM attacks in younger and older patients, respectively (Table S1), and the ratio of APS attacks was lower in patients aged > 50 years. The proportion of type of attacks in the natural history analysis did not differ between men and women (Table S2).

The severity of onset was significantly different among age groups (Table 1), and EDSS score at first attack was higher in patients with an older onset age, although the ratio of patients who had a severe first attack did not differ among age groups. No difference in EDSS score for the first attack was observed between the sexes (Table S3). A similar result was found in the analysis of the EDSS scores for the first four relapses (excluding the first attack). The EDSS score for relapses was significantly higher in patients aged > 50 years and lower in patients aged < 30 years (Figure S1a), while it did not differ between men and women (Figure S1b).

In terms of natural history, 50% of patients experienced active NMOSD relapses; the ratio of patients who experienced active NMOSD relapses was significantly higher in patients aged 30–39 years but lower in patients aged > 50 years, while the ARR in the natural history was not different among age groups (Table 1, Figure S2b). The rate of active relapses did not differ between the sexes; however, a higher ARR before IST was observed in men (Figure S2a).

Correlation between onset age and relapse patterns in natural history varies between the sexes

To explore the relapse patterns in different age groups and the two sexes, multivariable Poisson regression was performed. Although the ARR in the natural history did not differ among age groups, after adjusting for presentation and severity of the first attack, a positive correlation between age at onset, sex, and ARR before IST was found (Table 2).

TABLE 2.

Association between onset age and ARR for each subtype of relapses and overall annualized relapse rate before immunosuppression therapy.

	Overall ARR		ARR for ON		ARR for TM		ARR for APS		ARR for BS		ARR for Multifocal relapse
	RR (95% CI)	p value	RR (95% CI)	p value	RR (95% CI)	p value	RR (95% CI)	p value	RR (95% CI)	p value	RR (95% CI)	p value
Age group
18–29	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–
30–39	1.61 (1.33–1.94)	<0.001	1.17 (0.86–1.58)	0.33	1.44 (1.13–1.83)	<0.001	2.68 (1.25–5.77)	0.01	0.66 (0.35–1.26)	0.21	0.88 (0.54–1.43)	0.60
40–49	1.56 (1.29–1.88)	<0.001	1.14 (0.84–1.55)	0.40	1.5 (1.19–1.9)	<0.001	2.57 (1.18–5.58)	0.02	0.88 (0.48–1.63)	0.69	1.32 (0.86–2.04)	0.21
≥50	1.74 (1.43–2.13)	<0.001	1.28 (0.92–1.78)	0.14	1.91 (1.5–2.44)	<0.001	0.89 (0.28–2.84)	0.85	1.50 (0.80–2.82)	0.20	1.54 (0.98–2.43)	0.06
Sex
Female	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–	1 (ref.)	–
Male	1.34 (1.06–1.69)	0.01	1.54 (1.06–2.22)	0.02	1.16 (0.85–1.58)	0.34	1.23 (0.44–3.48)	0.69	2.15 (1.05–4.42)	0.04	1.83 (1.08–3.11)	0.03

Note: Adjusted for initial severe attack and presentation of onset (ON, TM, APS, BS, and multifocal).

Abbreviations: APS, area postrema syndrome; ARR, annualized relapse rate; BS, brain stem/cerebral syndrome; CI, confidence interval; IST, immunosuppression therapy; ON, optic neuritis; ref., reference; RR, rate ratio; TM, transverse myelitis.

Compared with patients with an onset age of 18–29 years, patients with an older onset age showed a higher overall ARR, ARR for TM relapse, and ARR for APS relapse, but the ARR for APS relapse in patients aged > 50 years was not different from that in patients aged 18–29 years. No correlation between the ARR and ON, BS, or multifocal relapse or between the ARR and onset age was found. Compared with women, men had a significantly higher overall ARR and ARR for ON, BS, and multifocal relapse, whereas no association between ARR for TM, APS relapse, and sex was observed.

We evaluated the impact of age at onset on recurrent relapse using AG regression in men and women (Table S4). In different sexes, the correlation between age at onset and risk of relapse was completely different. In female patients, older onset age was a risk factor for recurrent relapses; however, an opposite correlation was observed in male patients. Additionally, we noticed that in men, although the risk of relapse in patients in older age groups was lower compared with the youngest age group, patients aged > 50 years showed a trend towards elevated risk of relapse compared with patients aged 40–49 years (hazard ratio [HR] 0.80 vs. 0.52), indicating a potential nonlinear association between onset age and recurrent relapses.

We used restricted cubic splines with four knots at the 5th, 35th, 65th and 95th percentiles to explore potential correlations. A J‐shaped association between onset age and predicted risk of relapse was found in women (Figure 2a); in patients with onset age of < 30 years, the predicted risk of relapse remained stable and ultimately increased with a higher onset age (p, non‐linear < 0.001), whereas we found a U‐shaped correlation in men (Figure 2b), which showed a substantial reduction in risk and reached the lowest risk around 50 years, and then started to increase rapidly thereafter (p, non‐linear < 0.001).

FIGURE 2.

Association between the predicted risk of relapse and onset age in different sexes in the natural history. (a) Female and (b) male patients. Hazard ratios (HRs) are indicated by solid lines and 95% confidence intervals (CIs) by shaded areas. Horizontal black dotted line indicates HR = 1, and four knots were placed at the 5th, 35th, 65th, and 95th centiles.

Interaction between onset age and sex in the natural history

To analyse the interaction between onset age and sex, we recoded onset age as a binary variable: patients with an onset age of < 50 years were coded as EO‐NMOSD, while those aged > 50 years were coded as LO‐NMOSD. By adding an interaction term to the AG model, we assessed the interaction on the additive scale between late onset and sex. The results are summarized in Table S5.

Among EO‐NMOSD patients, men showed a higher risk of relapse compared with women (HR 1.14, 95% confidence interval [CI] 1.3–1.16), while in LO‐NMOSD patients, the risk of relapse was not different between men and women (HR 0.99, 95% CI 0.86–1.02). LO‐NMOSD was a risk factor for relapse for female patients (HR 1.24, 95% CI 1.23–1.25), while it did not influence the risk of relapse for male patients (HR 1.07, 95% CI 1.04–1.11). The joint effect estimates for risk of relapse were lower than those for the individual effect, and a negative interaction on the additive scale was found between LO‐NMOSD and male sex (RERI = −0.16, 95% CI −0.2 to −0.12).

Clinical features after IST

To explore the clinical features of the different therapy statuses, we analysed the proportion of each presentation of NMOSD attacks before and after IST (Figure S3). In general, compared to the natural history, the propensity of the involved areas differed after IST. The proportion of TM attacks increased after IST, and the proportion of all other types of attacks decreased; however, in male patients and patients with LO‐NMOSD, the presentation of attacks did not differ after IST.

The EDSS scores for the first four relapses after IST are shown in Figure S4. After IST, the severity of attacks differed among age groups, and the EDSS score for relapses was significantly higher in patients aged > 50 years. No difference in the EDSS score for relapses was observed between the sexes. After IST, the ARR significantly decreased among age groups and sexes, and no difference in ARR was found among those groups (Figure S5).

Correlation between onset age and relapse patterns after IST in different sexes

Although no difference in ARR after IST was found among age groups and sexes, our previous analysis showed different effects of age of onset on the risk of relapse in different sexes; therefore, we analysed the association between age of onset and the risk of relapse after IST. Five patients who did not receive IST were excluded, and 461 patients with 769 relapses after IST were included. In the multivariate analysis, ARR and active relapses in the natural history were additionally adjusted for as covariates.

After IST, an older onset age in men was associated with a lower risk of recurrent relapses compared with the youngest group; the risk in patients aged > 50 years decreased by 80% (p < 0.001), while in women, the risk of recurrent relapses increased in the first two groups and then decreased in the oldest group (Table 3). The association was flexibly modelled using the restricted cubic splines, and the curves are shown in Figure 3. In female patients, an inverted U‐shaped correlation between the predicted risk of recurrent relapses and onset age was observed: the risk of relapse gradually increased and peaked at around 35 years, then substantially decreased afterwards, and finally remained flat. By contrast, a strong negative association between the predicted risk of relapse and onset age was found in male patients.

TABLE 3.

Association between onset age and overall risk of relapse, analysed using the Anderson–Gill model.

Sex	Age group	Risk of relapse
		HR	95% CI	p value
Female	18–29	1 (reference)	–	–
	30–39	1.8	1.34–2.46	<0.001
	40–49	1.5	1.14–1.92	<0.001
	≥50	0.7	0.52–1.07	0.054
Male	18–29	1 (reference)	–	–
	30–39	0.4	0.17–0.76	0.01
	40–49	0.2	0.07–0.39	<0.001
	≥50	0.2	0.1–0.56	<0.001

Note: Adjusted by initial severe attack, presentation of first onset (ON, TM, APS, and BS), ARR before IST, and active relapses before IST.

Abbreviations: APS, area postrema syndrome; ARR, annualized relapse rate; BS, brain stem/cerebral syndrome; CI, confidence interval; HR, hazard ratio; IST, immunosuppression therapy; ON, optic neuritis; TM, transverse myelitis.

FIGURE 3.

Association between predicted risk of relapse after IST and onset age. Hazard ratios (HRs) are indicated by solid lines and 95% confidence intervals (CIs) by shaded areas. Horizontal black dotted line indicates HR = 1, and the onset age of 18 years was used as the reference. (a) A U‐shaped association between predicted risk of relapse after immunosuppression therapy (IST) and onset age was found in female patients. (b) In male patients, the predicted risk of relapse after IST decreased with a higher onset age.

Longitudinal change in EDSS score over relapses in different age groups before and after IST

In all patients, with an increase in the number of relapses, increasing severity was observed in the natural history (Figure S6). However, patients with LO‐NMOSD experienced disability progression over relapses significantly more rapidly (B = 0.42, SE = 0.07, p < 0.001 in the natural history; B = 0.18, SE = 0.05, p < 0.001 after IST) than those with EO‐NMOSD (B = 0.19, SE = 0.05, p < 0.001 in the natural history; B = 0.14, SE = 0.05, p < 0.001 after IST) both pre‐ and post‐IST.

DISCUSSION

We found that an older onset age was associated with a higher prevalence of TM but a lower prevalence of ON, which was observed not only in the first attack but also in the subsequent relapses. Similar results were reported in other studies [20, 21, 22]; these results support the idea that there are differences in anatomical susceptibility in patients with different onset ages. The presentation of onset/all attacks was not different between women and men, consistent with the results in a German cohort [23]. Two retrospective cohort studies observed a higher TM prevalence but lower ON prevalence in male patients [24, 25]; in these cohorts, male patients were significantly older at onset, while in our cohort, the onset age was not different between men and women, which explains the different results among several cohorts, indicating that the preference of affected areas was associated with age rather than sex. Interestingly, we observed a higher prevalence of TM attacks after IST, which may be attributed to increasing age during follow‐up. Thus far, IST has not been reported to affect the presentation of NMOSD attacks.

Whether onset age is a predictor of relapse in NMOSD remains debatable. Previous studies have produced conflicting results [4, 20, 26, 27, 28]. Conflicting results were also obtained when studying the association between sex and relapse rate [26, 29, 30]. In addition, a recent study suggested that age of onset was associated with response to IST [31]; thus, it was necessary to evaluate the correlation between age of onset and sex in different IST statuses. Although several studies have reported ARR before IST initiation, the predictors of relapse in natural history are still less studied. Our study provides data to further understand this issue.

Several studies suggested that older age was associated with a lower risk of relapse [4, 26], while some studies defined early‐ and late‐onset NMOSD according to whether the onset age was over or under 50 years and found no differences in ARR between the two groups [28, 32, 33]. Another study defined very late‐onset NMOSD (VLO‐NMOSD) as onset age > 70 years and found that the attack interval was significantly shorter in patients with VLO‐NMOSD than in those with EO‐NMOSD and LO‐NMOSD. We observed different relapse patterns in patients with different onset ages, sexes, and treatment statuses. In the natural history, a completely positive correlation between onset age and risk of relapse was found in women, which was contrary to several studies [4, 26, 34]. In these studies, the differences in sex and treatment status were not considered, which might have led to bias.

The analysis of the interaction between age of onset and sex also supported the difference in relapse risk between sexes, and we found that the impact of age of onset on risk of relapse was modified by sex (RERI < 0), indicating an opposite effect of age of onset on relapse risk in different sexes. Our work suggests a potential high‐relapse‐risk population: in untreated patients, females with LO‐NMOSD showed the highest risk of relapse compared with other patients, suggesting that more attention should be paid to these patients.

Our study revealed different relapse patterns under various treatment conditions. After IST, an inverted U‐shaped curve and a monotonically decreasing curve of the correlation between age at onset and risk of relapse were observed in women and men separately (Figure 3). Compared with the natural history curves (Figure 2), the relative risk of patients aged > 50 years was obviously reduced, indicating a better response to IST in these patients (especially patients with LO‐NMOSD). Our findings was supported by those of a Korean cohort [31], in which a younger age of onset was related to a poor response to azathioprine and mycophenolate mofetil. Consistent with previous studies, we observed a higher EDSS score for NMOSD attacks in patients with LO‐NMOSD, regardless of their IST status. Given the potentially devastating consequences of NMOSD attacks, the higher risk of mortality, and the better IST response, patients with LO‐NMOSD can benefit more from early and highly effective IST.

The prevalence of NMOSD was much higher in women than in men, with a female–male ratio range of 1.2 to 8.8 [35, 36, 37]. Female sex was considered the strongest risk factor for NMOSD, particularly in AQP4‐IgG‐positive NMOSD [1]. Interestingly, after IST, we found a relapse peak at childbearing age (30–40 years) in women, and no such peak was found in men, strongly suggesting sex differences in relapse patterns.

A major strength of this study is that the data were obtained from a very large cohort. Additionally, we exploratively evaluated predictors for relapse in the natural history and analysed the potential interaction between onset age and sex, as well as the nonlinear correlation between onset age and the risk of relapses.

The primary limitation of this study was its retrospective design. Furthermore, although we analysed the disease course in natural history, the median observation period was still relatively short (23 months). In addition, our study explored the impact of age on the risk of relapse using age at onset rather than age at each relapse, which limits the interpretation of the results. Furthermore, the type of maintenance therapy is related to the risk of relapse; however, many patients experienced an alternation of therapy, which made it difficult to analyse the association between the type of IST and relapses. Future prospective studies should focus on this issue. Finally, due to the rarity of NMOSD in men, the number of male patients included in this study was relatively low, and the results may not reflect the true clinical features of male patients.

In conclusion, the presentation of NMOSD attacks was related to age at onset, and relapse patterns were associated with age at onset, sex, and treatment status. Future large prospective cohort studies are needed to investigate the features of relapse with respect to age and sex.

AUTHOR CONTRIBUTIONS

Hongyu Zhou: Conceptualization; investigation; data curation; resources; funding acquisition; writing – review and editing. Wenqin Luo: Conceptualization; investigation; writing – original draft; methodology; writing – review and editing; formal analysis. Ziyan Shi: Funding acquisition; data curation; resources. Lingyao Kong: Data curation; resources. Xiaofei Wang: Funding acquisition; data curation; resources.

FUNDING INFORMATION

This work was funded by: the Department of Science and Technology of Sichuan Province (Grant Number: 2022YFS0315 to ZHY); 1·3·5 Project for Disciplines of Excellence–Clinical Research Incubation Project, West China Hospital, Sichuan University (Grant Number: 21HXFH041 to ZHY); the National Natural Science Foundation of China (Grant Number: 82201494 to SZY); and the Natural Science Foundation of Sichuan province (Grant Number: 022NSFSC1432 to WXF).

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflicts of interest.

Supporting information

FIGURE S1 EDSS of the first four relapses (first attack excluded) in different age groups and sexes. EDSS, expanded disability status scale; ns, no significance; *represented p value < 0.05.

FIGURE S2. ARR in the natural history among age groups and sex, presented as median (IQR). The ARR before IST is higher in male patients compared with patients aged 18–29 years. The ARR is not significantly different in other age groups. ARR, annualized relapse rate; IST, immunosuppression therapy; NS, no significance; *represented p value < 0.05.

FIGURE S3. Presentation of attacks in different treatment statuses. (A) All participants; (B) female patients; (C) male patients; (D) patients with EO‐NMOSD; and (E) patients with LO‐NMOSD. APS, area postrema syndrome; BS, brain stem/cerebral syndrome; EO‐NMOSD, early onset NMOSD; LO‐NMOSD, late onset NMOSD; ON, optic neuritis; TM, transverse myelitis.

FIGURE S4. EDSS of the first four relapses after IST. EDSS, Expanded Disability Status Scale; IST, immunosuppression therapy; ns, no significance; **represented p value < 0.01; ***represented p value < 0.001.

FIGURE S5. ARR after IST among age groups and sexes, presented as median (IQR). ARR, annualized relapse rate; IST, immunosuppression therapy; NS, no significance.

FIGURE S6. Plots of longitudinal change in EDSS over relapses. The thinner lines show individual‐based values, and thicker lines show fitted linear mixed effect models in the nature history (A) and after IST (B). EDSS, Expanded Disability Status Scale; EO‐NMOSD, early onset NMOSD; IST, immunosuppression therapy; LO‐NMOSD, late onset NMOSD.

TABLE S1. Presentation of 1394 attacks before IST (the first onsets included) of 533 patients. APS, area postrema syndrome; BS, brain stem/cerebral syndrome; IST, immunosuppression therapy; ON, optic neuritis; TM, transverse myelitis.

TABLE S2. Demographic and clinical data of natural history of patients by sex. APS, area postrema syndrome; BS, brain stem/cerebral syndrome; ON, optic neuritis; TM, transverse myelitis.

TABLE S3. Demographic and clinical data of natural history of patients by sex. APS, area postrema syndrome; ARR, annualized relapse rate; BS, brain stem/cerebral syndrome; ON, optic neuritis; TM, transverse myelitis.

TABLE S4. Association between onset age and overall risk of relapse in the sex subgroups, analysed by the A–G model. CI, confidence interval; HR, hazard ratio.

TABLE S5. Interaction on the additive scale between onset age and sex in natural history. RERI < 0 with a 95% CI that does not contain 0 indicates a negative interaction on the additive scale. CI, confidence interval; EO‐NMOSD, early onset NMOSD; HR, hazard ratio; LO‐NMOSD, late onset NMOSD; RERI, relative excess risk due to interaction.

Luo W, Shi Z, Kong L, Wang X, Zhou H. Patterns of neuromyelitis optica spectrum disorder attacks in different age groups and sexes depending on the status of immunosuppressive therapy: A retrospective cohort study. Eur J Neurol. 2024;31:e16178. doi: 10.1111/ene.16178

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.