Introduction

Dyslipidemia is a major modifiable risk factor for cardiovascular disease, and its prevalence continues to increase in Japan, together with rapid population aging [1]. Despite the widespread use of lipid-lowering agents such as statins, ezetimibe, and proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, real-world data from large administrative databases have consistently shown that a substantial proportion of high-risk patients fail to achieve guideline-recommended low-density lipoprotein cholesterol (LDL-C) targets [2]. Recent real-world observational data from cardiology outpatient clinics in Turkey similarly showed that many patients remain above target despite intensified statin-based therapy, and that adding ezetimibe improved but did not normalize goal attainment [3]. This residual risk highlights the limitations of pharmacotherapy alone and underscores the need for sustainable, lifestyle-based strategies that can be implemented as part of routine care.

Complementary and non-pharmacological approaches, such as structured counseling on diet, physical activity, weight management, and smoking cessation, have attracted growing interest as adjuncts to conventional dyslipidemia management. These lifestyle-based strategies are increasingly embedded in chronic care programs and reimbursement schemes, including Japan’s new lifestyle disease management fee.

The Japanese government has recognized the importance of addressing the social determinants of health in reducing and managing the risk of non-communicable diseases, such as cardiovascular disease, and has made changes to the medical reimbursement system to reflect this [4]. In Japan, most residents are covered by public insurance schemes, and patients generally pay a fixed copayment at the point of care. The remainder is reimbursed directly from insurers to medical facilities based on the national fee schedule. In 2024, the long-standing “specific disease management fee” for conditions such as hypertension, diabetes, and dyslipidemia was replaced by a new “lifestyle disease management fee,” which took effect on 1 June 2024 [5]. These “management fees” are paid to providers as add-on fees for ongoing, structured chronic disease management, rather than as direct payments to patients. The design of the management fee categories may therefore influence provider behavior, adoption, and adherence to chronic care programs. Eligibility for the lifestyle-related disease management fee, which is applicable to outpatient visits only and billed on a per-visit basis, requires physicians to work with patients to create and periodically review a structured Treatment Plan Sheet [5].

This covers both biomedical targets (e.g., LDL-C levels) and broader lifestyle and social determinants (e.g., sleep, physical activity, work conditions, body weight). This reform therefore comprised both a technical change in coding and a broader shift in the institutional framework for chronic disease care. The new management fee emphasizes comprehensive lifestyle management with integrated goals and includes lifestyle counseling, rather than the previous approach in which comorbidities were effectively treated in parallel with standard pharmacotherapy and brief, disease-focused counseling during routine outpatient visits. Pharmacotherapy remains guideline-based and disease-appropriate under the new system. The intended added value lies in the more holistic, goal-oriented, and lifestyle-focused management framework, rather than in changes to drug regimens [6]. Overall, the fee revision was designed to encourage a more comprehensive, collaborative approach, with patients and physicians working together to create and review treatment plans. In theory, this approach should enhance adherence and outcomes. However, as far as can be ascertained, no study to date has examined the effect of these changes on patient outcomes related to dyslipidemia. Therefore, questions remain, such as how often LDL-C goals are achieved in routine care, and whether there have been reductions in major adverse cardiac events.

To address this gap, we conducted a nationwide retrospective cohort study using the Medical Data Vision (MDV) Diagnosis Procedure Combination (DPC) database, which includes approximately 30 million patients across acute care hospitals in Japan. We explored patterns of dyslipidemia care and outcomes before and after the 2024 fee revision to identify gaps in current care and potential target populations for future lifestyle-related interventions. Given the short post-revision observation window and observational design of the present study, our objective was not to estimate the causal effect of the fee revision but to provide an early descriptive snapshot of care patterns and outcomes surrounding implementation of the new lifestyle disease management fee.

Methods

Results

Study population and baseline characteristics

From approximately 30 million patients in the MDV DPC database, we identified 590,000 adults with dyslipidemia who had at least one lipid measurement and at least 6 months of follow-up between 2021 and 2024 (Table 1). Most visits occurred in large acute care hospitals participating in the DPC system, including university hospitals and tertiary centers, with the remainder seen at non-university acute care hospitals. Among the 590,000 patients with dyslipidemia and at least 6 months of follow-up, approximately 472,000 (roughly 80%) had visits only under the specific disease management fee (M1), approximately 38,000 (roughly 6%) had visits only under the lifestyle disease management fee (M2), and approximately 79,000 (roughly 14%) had visits during both periods; this indicated that the M1 and M2 visit counts partially reflected the same patients over time.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics of the outpatient cohort with dyslipidemia, by management fee category (unit of observation: outpatient visits).

https://doi.org/10.1371/journal.pone.0358270.t001

The outpatient cohort comprised 23,600,010 visit-level records, of which 21,000,013 visits were reimbursed under the specific disease management fee (M1) and 2,599,997 under the lifestyle disease management fee (M2). Accordingly, 551,000 unique patients (472,000 with M1 visits only plus 79,000 with visits in both periods) contributed the 21,000,013 M1 visits, and 117,000 unique patients (38,000 with M2 visits only plus 79,000 with visits in both periods) contributed the 2,599,997 M2 visits; the M1 and M2 visit counts therefore describe outpatient encounters, not distinct individuals. The mean patient age was similar across categories (67.4 years in M1 and 66.9 years in M2), with roughly one-third of visits involving patients aged 65–74 years and approximately 9%–10% involving those aged ≥ 75 years. Men accounted for just over half of visits in both groups (52.4% in M1 and 53.8% in M2).

Although fewer patients contributed to M2 and its post-fee-revision observation window was shorter, the average number of outpatient visits per patient was higher in M2, resulting in a comparable total number of visits across the two fee categories over the study period. Each outpatient visit was treated as a separate observation, and individual patients could contribute multiple visits to either fee category depending on the timing of their follow-up relative to the 2024 fee revision. Comorbidities were common: hypertension was recorded in 68.1% of M1 visits and 65.4% of M2 visits and diabetes mellitus in 39.1% and 37.7%, respectively. Prior cardiovascular disease was also frequent, with 12.4% versus 12.3% of visits associated with previous myocardial infarction, 9.1% versus 8.8% with previous ischemic stroke, and 8.3% versus 8.1% with previous heart failure in M1 and M2.

The distribution of cardiovascular risk categories was nearly identical between fee categories, with primary prevention accounting for 69.1% of M1 visits, 68.5% of M2 visits, and secondary prevention for 31.1% and 31.5%, respectively. Baseline lipid profiles were also highly comparable: mean TC was 217.8 mg/dL in M1 and 216.6 mg/dL in M2, mean LDL-C was 138.1 versus 137.7 mg/dL, mean HDL-C was 52.6 versus 52.4 mg/dL, and mean TG was 155.7 versus 154.7 mg/dL, with similar median and interquartile range values. Current smoking was reported in approximately one-fifth of visits (20.5% in M1 and 20.1% in M2). Across these variables, SMDs were generally below 0.1, indicating a good balance in baseline demographic, clinical, and lipid characteristics between M1 and M2, despite the large sample size.

Pharmacological lipid-lowering therapy and on-treatment lipid levels

During both study periods, most outpatient visits involved active pharmacological management of dyslipidemia (Table 2). Statins were prescribed in 81.9% of M1 visits and 82.3% of M2 visits; ezetimibe in 27.6% and 27.7%, respectively; fibrates in 11.9% of visits in both groups; and PCSK9 inhibitors in 2.5% of visits in each category. On-treatment lipid profiles were modestly improved as compared with baseline, but substantial residual dyslipidemia persisted.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Lipid profile and treatment patterns by management fee category.

https://doi.org/10.1371/journal.pone.0358270.t002

Mean TC declined to 195.5 mg/dL in M1 and 194.9 mg/dL in M2, and mean LDL-C dropped to 114.8 mg/dL and 114.3 mg/dL, respectively. Mean HDL-C increased slightly to 54.8 mg/dL in M1 and 54.4 mg/dL in M2, whereas mean TG decreased to 141.8 mg/dL and 140.9 mg/dL, respectively. Differences in on-treatment lipid parameters between M1 and M2 were small and not clinically meaningful; SMDs were < 0.10 for all lipid variables, and no statistically significant differences were detected using t-tests or analysis of variance after adjustment for the large sample size. These findings suggest that the initial transition from specific disease management fee (M1) to lifestyle disease management fee (M2) did not substantially alter visit-level treatment intensity or on-treatment lipid levels within the observation window.

Age-stratified LDL-C goal attainment and impact of fee transition

Across 23,600,010 outpatient visits, LDL-C goal attainment declined progressively with advancing age under both fee categories. The lowest attainment rates were observed among visits involving patients aged ≥ 75 years in both M1 and M2 (Fig 1). In age-stratified analyses (< 65, 65–74, and ≥ 75 years), LDL-C goal attainment was consistently highest in those aged < 65 years, lower in those aged 65–74 years, and lowest in those aged ≥ 75 years, reflecting the greater burden of comorbidities and more refractory dyslipidemia in older patients.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Low-density lipoprotein cholesterol (LDL-C) goal attainment by age and management fee category.

Age-stratified LDL-C goal attainment among outpatient visits under the specific disease management fee (M1) and lifestyle disease management fee (M2).Each bar represents the proportion of outpatient visits at which patients had achieved the LDL-C target, stratified by age group and fee category. A total of 21,000,013 outpatient visits (from 551,000 unique patients) were classified as M1 and 2,599,997 visits (from 117,000 unique patients) as M2; the numbers shown within the bars are outpatient visits, not individual patients. The higher number of M1 visits primarily reflects the longer pre-revision observation period and the inclusion of patients already under follow-up prior to the 2024 revision, rather than a systematically higher visit frequency per patient. Error bars are 95% confidence intervals calculated using the number of unique patients contributing visits in each stratum as the effective sample size, so as to account for clustering of repeated visits for the same patient.

https://doi.org/10.1371/journal.pone.0358270.g001

In each age stratum, the difference in LDL-C goal attainment between M1 and M2 was small, with overlapping 95% CIs and odds ratios close to 1.0 (all 95% CIs included 1.0). These age-stratified comparisons were tested using chi-square tests and supplemented with SMDs, which were uniformly < 0.10, indicating no material imbalance in visit-level LDL-C control between fee categories. Taken together, the similarity in lipid-lowering treatment patterns, on-treatment lipid levels, and age-stratified LDL-C goal attainment between M1 and M2 suggests that, within the short follow-up period after the 2024 fee revision, the transition from specific disease management fee to lifestyle disease management fee did not lead to detectable improvements in LDL-C control at the visit level.

Major adverse cardiovascular events

Cardiovascular outcomes were assessed at the patient level (Table 3). The 551,000 patients who contributed M1 visits were followed for a mean of 2.8 years, yielding 1,542,800 patient-years of observation. Because the lifestyle disease management fee was introduced on 1 June 2024, the 117,000 patients who contributed M2 visits could be followed for a mean of only 0.5 years, yielding 58,500 patient-years, approximately 26-fold fewer patient-years than in M1. Incidence rates of major adverse cardiovascular events were substantial and numerically similar between fee periods. For the composite of myocardial infarction, ischemic stroke, and cardiovascular death, there were 42,242 events in M1 (7.7% of M1 patients) and 1591 events in M2 (1.4% of M2 patients), corresponding to incidence rates of 27.38 (95% CI 27.12–27.64) and 27.20 (95% CI 25.88–28.57) per 1000 patient-years, respectively. Given the markedly unequal and very short post-revision observation window, this similarity should be regarded as descriptive only and must not be read as evidence that the fee revision had no effect on cardiovascular outcomes.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Patient-level major adverse cardiovascular events in the study population, by management fee period.

https://doi.org/10.1371/journal.pone.0358270.t003

All-cause mortality also occurred at similar rates in both periods (16.33 vs. 16.21 per 1000 patient-years; 25,194 deaths in M1 and 948 in M2), with overlapping 95% CIs. Myocardial infarction and ischemic stroke each contributed materially to the event burden, with incidence rates of 10.03 per 1000 patient-years for myocardial infarction in both periods (15,474 vs. 587 events) and 7.82 versus 7.97 per 1000 patient-years for ischemic stroke in M1 and M2, respectively (12,065 vs. 466 events). Hospitalization for heart failure was frequent, with incidence rates of 10.88 per 1000 patient-years in M1 and 11.26 per 1000 patient-years in M2 (16,786 vs. 659 events). These consistently high rates of cardiovascular events and mortality, despite high statin use and broadly similar management under M1 and M2, underscore a substantial residual risk that appears not to have been mitigated during the initial implementation period of the lifestyle disease management fee, suggesting the need for more comprehensive risk-reduction strategies extending beyond pharmacotherapy alone.

Discussion

This nationwide real-world analysis of 590,000 adults with dyslipidemia who contributed 23.6 million outpatient visits in Japanese acute care hospitals showed that LDL-C goal attainment remains suboptimal at the visit level, despite widespread use of statins and combination lipid-lowering therapy, and that this pattern persisted after the 2024 fee schedule revision. Suboptimal lipid control was evident before this change, and the initial post-implementation period showed no clear improvements in LDL-C goal attainment or in major acute cardiovascular events in this observational setting. The present findings should be interpreted with caution, given the limited post-implementation follow-up and potential for residual confounding. However, these findings suggest that measures beyond fee revision alone will be needed to close the gap between guideline recommendations and routine care. This is broadly consistent with prior Japanese and international reports documenting persistent underachievement of LDL-C targets among high-risk patients [10,11].

The 2024 fee structure was introduced to support more holistic disease management, including explicit attention to lifestyle and social factors, and to promote patient involvement in treatment planning [5]. In principle, such changes could enhance adherence and improve outcomes. However, across the 23.6 million outpatient visits contributed by these patients, JAS LDL-C targets were met at only approximately 58% of visits, with consistently lower attainment in secondary than in primary prevention. These results underscore a substantial residual cardiovascular risk that is unlikely to be mitigated by pharmacotherapy alone and highlight the need for scalable lifestyle-oriented strategies that can be embedded in routine care [4,8,11,12]. Our findings are also consistent with contemporary real-world evidence from other health systems, in which a large share of patients attending specialist outpatient clinics remained above target despite intensification of statin-based regimens [3].

The baseline characteristics in this study illustrate both the scale of dyslipidemia care and the complexity of comorbidity in an aging society. Patients under both fee schemes had a mean age of approximately 67 years; two-thirds of visits involved coexisting hypertension, nearly 40% involved diabetes mellitus, and approximately one-third involved secondary prevention. Despite broadly similar risk profiles between M1 and M2, on-treatment lipid levels remained far from optimal, with mean LDL-C levels of approximately 115 mg/dL and more than 40% of visits failing to meet JAS LDL-C goals (Table 2). Statins were prescribed in more than 80% of visits, and ezetimibe or fibrates were frequently added, indicating that many clinicians had already escalated patients’ pharmacologic regimens. Nevertheless, a large fraction of patients remained above the target, a pattern consistent with claims-based studies from Japan and other countries [13–15].

Cardiovascular outcomes were similarly sobering. Patient-level incidence rates of major acute cardiac events and all-cause mortality were high and numerically comparable between M1 and M2, with approximately 27 events per 1000 patient-years for major acute cardiac events and approximately 11 hospitalizations for heart failure per 1000 patient-years in both periods. These event rates, observed at the patient level over a mean follow-up of 2.8 years and approximately 1.54 million patient-years of observation in M1, underscore that contemporary drug-centered dyslipidemia management has not eliminated the cardiovascular risk in this high-risk population. Importantly, no measurable improvement in LDL-C control was detected after the transition to M2 in mid-2024; for cardiovascular events, only 58,500 patient-years were accumulated between June and December 2024, which was insufficient to detect a change in event rates; thus, the absence of an observed difference is uninformative rather than reassuring.

Several factors may explain the apparent lack of early improvement in our study. First, there was only a brief post-implementation period between the introduction of fee changes in June 2024 and the end of follow-up in December 2024; such a practice change typically requires more time to become embedded. For example, implementation studies of evidence-based practices have reported substantial increases in uptake between 7–9 months and 2.5 years after initial introduction, a time frame that extends beyond our observation window [16]. Second, multiple barriers at the patient, professional, and organizational levels may impede effective implementation of shared, lifestyle-focused care. A systematic review of shared decision-making in older adults with multimorbidity found that poor health status and cognitive or physical impairments reduced patient participation; time constraints and limited communication skills among health care professionals also hindered shared decision-making [17]. Other studies suggest that older adults with a greater burden of health problems may be less likely to favor or engage in shared decision making [18]. At the organizational level, constraints on staff time, workload, and competing priorities in busy outpatient settings likely make structured lifestyle counseling difficult to deliver consistently [17]. Lastly, the financial incentive associated with the lifestyle disease management fee may have been insufficient to offset the additional workload and logistic complexity required to implement structured counseling. Because these mechanisms cannot be disentangled using claims data alone, our findings should be regarded as hypothesis generating. Nevertheless, longer-term follow-up remains important, given the accumulating evidence that structured lifestyle-based interventions can reduce cardiovascular risk in high-risk populations [19–21].

Our results will help to identify groups that can be prioritized for additional support and/or lifestyle interventions. In our study, LDL-C goal attainment declined with advancing age, and patients in secondary prevention were less likely to reach targets than those in primary prevention. Therefore, older adults and patients with prior cardiovascular events appear particularly vulnerable to residual risk despite pharmacotherapy and may benefit most from tailored lifestyle interventions [18]. Paradoxically, these same groups are among those less likely to actively participate in health care decisions, underscoring the need for proactive, patient-centered approaches versus reliance on patient-initiated engagement alone.

In the above context, complementary and non-pharmacological approaches have attracted increasing interest as adjuncts to conventional dyslipidemia management [22]. For example, a recent systematic review of seven randomized controlled trials reported that hot spring or sauna interventions, particularly when combined with exercise, produced modest but clinically relevant reductions in TC and LDL-C among younger or middle-aged adults (mean age < 60–65 years) whereas the effects were small or non-significant in older or multimorbid cohorts [23]. Although such interventions cannot be evaluated directly using claims-based data, they illustrate the potential for structured, context-specific lifestyle programs to complement drug therapy, particularly in settings where thermal and spa resources are readily available.

This study has several strengths. We leveraged one of the largest hospital-based claims and laboratory datasets in Japan, capturing 23.6 million outpatient visits from 590,000 patients and providing a large volume of observations with which to describe LDL-C control across age, risk, and fee strata. The analysis spans both pre- and early post-2024 reimbursement regimes, offering an initial view of how the new lifestyle disease management fee aligns with emerging lifestyle-oriented strategies. However, for detecting changes in clinical practice and outcomes, the post-implementation period is relatively short, and a lag between policy changes and frontline counseling practices is likely. Therefore, we cannot rule out the possibility that longer follow-up will reveal differences in LDL-C control or cardiovascular event rates between M1 and M2. Future analyses should incorporate more recent data not available at the time of this study.

Several limitations should also be considered. The MDV database primarily includes DPC-participating acute care hospitals and may underrepresent care in smaller clinics or non-DPC settings, potentially limiting generalizability of the findings. The database does not capture detailed information on lifestyle behaviors or non-pharmacological interventions, including diet, physical activity, structured counseling, or complementary approaches such as hot spring or sauna therapy. Therefore, we could not evaluate how the fee revision influenced lifestyle-related care processes and could not disentangle the effects of pharmacologic intensification from behavioral change. Residual confounding is inevitable in observational claims analyses; we lacked data on diet, physical activity, sleep, and other lifestyle factors that influence lipid profiles and cardiovascular risk. Some misclassification of fee category during the early implementation phase is also possible. Overall, the present analysis should be viewed as an early snapshot of real-world practice patterns and outcomes surrounding the fee revision, rather than a definitive policy evaluation of the lifestyle disease management fee.

Several further limitations follow directly from the study design and data source. First, the M1 and M2 periods are neither independent nor of comparable length: M1 spans January 2021 to May 2024 (41 months) whereas M2 spans only June to December 2024 (7 months), and 79,000 patients (13% of the cohort) contributed observations to both periods. The M1 versus M2 contrasts are therefore repeated cross-sectional descriptions of the same evolving population rather than comparisons of independent groups, and they remain vulnerable to secular trends, calendar-time effects, seasonal variation in lipid testing, and within-patient correlation of repeated visits. We deliberately refrained from formal hypothesis testing of policy effects for this reason, reported conservative CIs based on the number of unique patients, and interpreted all M1 versus M2 differences as descriptive. Second, the cardiovascular outcome comparison is intrinsically limited by the accumulation of only 58,500 patient-years after implementation; a difference in event rates could not have been detected within this window, even if one had existed, and our findings should not be read as evidence of no effect. Third, the MDV database is built on DPC-participating acute care hospitals and therefore excludes most small clinics, primary care practices, and other non-DPC facilities, where a large share of routine dyslipidemia care and much of the uptake of the lifestyle disease management fee is likely to occur. The cohort is consequently enriched for older, multimorbid, hospital-managed patients, and the findings may not generalize to the overall Japanese population with dyslipidemia. Fourth, the central components of the new management strategy including structured lifestyle counseling, shared goal setting recorded on the Treatment Plan Sheet, medication adherence, dietary modification, physical activity, sleep, smoking behavior, and body weight are not directly measurable using administrative claims. Adherence could only be approximated from prescription continuity, and the delivery or quality of counseling could not be observed at all. Fifth, laboratory results are reported only for tests performed and recorded at participating institutions, so lipid values may not be missing at random, and coding of the fee category during the early implementation phase may be subject to misclassification.

From a clinical and health policy perspective, the 2024 shift to a lifestyle disease management fee represents a timely institutional opportunity to integrate structured lifestyle programs, including exercise, nutrition, sleep counseling, and appropriate complementary thermal interventions into dyslipidemia care. Our findings suggest that any impact of the fee change on clinical outcomes may take time to emerge and may be difficult to detect using routine claims data alone. Future research with longer follow-up, more granular clinical and lifestyle information, and quasi-experimental or pragmatic trial designs, where feasible, will be needed to more rigorously evaluate the long-term effects of the new fee structure. Pragmatic cluster-randomized or stepped-edge trials embedded in M2-based care pathways could test whether adding specific lifestyle or complementary interventions can yield incremental improvements in LDL-C control and major acute cardiac events beyond drug therapy alone. By explicitly linking reimbursement structures, real-world risk profiles, and evidence-based non-pharmacological options, such studies could inform future guideline and policy development for the management of dyslipidemia in aging societies.

Acknowledgments

The insurance claims data were provided by Medical Data Vision Co., Ltd. (Tokyo, Japan). The authors obtained the data under a research agreement, and the company did not participate in the study design, analysis, or interpretation of the results. The authors thank Edanz (https://jp.edanz.com/ac) for critically reviewing and editing a draft of this manuscript.