Partial obesity paradox regarding mortality in patients hospitalized with diastolic or systolic heart failure

ABSTRACT

Introduction

A phenomenon known as the obesity paradox has been reported in patients with heart failure (HF). The goal of this study is to characterize this observation in systolic (SHF) and diastolic (DHF) HF.

Methods and results

We used the National Inpatient Sample (NIS) database for 2016–2020. We evaluated mortality based on body weight. A total of 7,364,023 with SHF and 10,064,223 with DHF were found in the NIS database. All-cause inpatient mortality was lowest in overweight followed by obesity and morbid obesity, whereas mortality was highest in cachexia for SHF and DHF (mortality: overweight 2.56%, obese 3.12%, morbidly obese 3.70%, normal weight 5.60%, and cachexia 15.22%; p < 0.001) and DHF patients (mortality: overweight 2.08%, obese 2.43%, morbidly obese 2.93%, normal weight 4.58%, and cachexia 14.25%; p < 0.001). This relationship remains similar after multivariate analysis (SHF patients: overweight OR: 0.49 (0.41–0.58), obesity OR: 0.64 (0.62–0.66), morbid obesity OR: 0.85 (0.83–0.88), and cachexia OR: 2.78 (2.67–2.90); p < 0.001; DHF patients: overweight OR: 0.47 (0.40–0.56), obesity OR: 0.61 (0.59–0.63), morbid obesity OR: 0.83 (0.81–0.85), and cachexia OR: 3.09 (2.96–3.23); p < 0.001).

Conclusions

All-cause inpatient mortality in SHF and DHF is lowest in overweight populations followed by obese and morbidly obese populations, whereas cachexia has the highest mortality.

Plain Language Summary

We studied the death rate of patients with heart failure based on weight categories. We studied 7,364,023 patients with weak heart muscle and heart failure and 10,064,223 with normal or near normal heart muscle. The death rate in patients in both categories was lowest in overweight, obese, and morbid obesity, whereas it was highest in underweight compared to normal weight. These are patients admitted to a hospital with a diagnosis of heart failure. The reason for our finding is unclear, but higher energy reserves in obese patients may offer some protection against death. However, it is crucial to note that our data does not apply to healthy individuals or heart patients who are not ill enough to be hospitalized, and, therefore, we should not encourage weight gain in individuals of normal weight.

1. Introduction

The incidence and prevalence of obesity has risen dramatically in the United States and has reached pandemic levels worldwide [1,2]. The disease has been extensively studied and excess visceral adiposity (visceral fat is composed of several adipose depots including mesenteric, epididymal white adipose tissue pericardial and perirenal fat) has been shown to impose a variety of systemic metabolic derangements, such as increasing pro-inflammatory signaling cascades, renin-angiotensin-aldosterone system activation, insulin resistance, and blood vessel sympathetic activation [3,4], leading to many chronic diseases. In particular, it serves as a significant risk factor in the pathogenesis of cardiovascular disease (CVD) [3–6] including hypertension, coronary heart disease, atrial fibrillation, and heart failure (HF). Other paradoxical effects of risk factors have also been observed. For example, metabolic syndrome might offer protection against the occurrence of dementia among patients with atrial fibrillation [7].

Still, our collective understanding of weight and obesity continues to grow. Although it continues to serve as a major comorbidity in the development of CVD, studies have shown that in patients with established CVD, overweight and obese populations display a better prognosis in comparison to those with normal weight [5,6,8–10]. This “obesity paradox” in CVD is especially apparent in HF, although questions regarding this phenomenon persist and warrant further examination. Some have suggested that in patients with HF, mortality decreases with increasing weight [9–13], whereas others have proposed that there is a specific point where the detrimental effects of obesity are minimized and the beneficial effects of the obesity paradox are maximized [5,6,8,14]. Still, results are unclear, and the point of weight optimization remains uncertain. Furthermore, due to obesity’s complex interplay with various metabolic systems, comorbidities of obesity make studying its effects on HF challenging and bring uncertainty to our current understanding of the relationship. BMI is an incredibly quick, easy, and useful tool for healthcare workers to get a snapshot of a patient’s potential body type and allows researchers to have an abundance of data to analyze. Cachexia is a devastating illness that clinically presents progressive weight loss with alterations in body composition and disturbed homeostasis of several body systems. In particular, cachexia is characterized by a loss of muscle mass, which may or may not be accompanied by a loss of fat mass. The goal of this study is to evaluate and characterize the obesity paradox in both SHF and DHF patients with the largest available inpatient database of adult patients while eliminating potential confounders by adjusting baseline characteristics and comorbidities of patients.

2. Method

2.1. Collected information

This study contains data from patients aged over 18 years old that were admitted to NIS hospitals from 2016 to 2020 with ICD-10 diagnoses codes for SHF (I50.2, I50.20, I50.21, I50.22, I50.23) or DHF (I50.3, I50.30, I50.31, I50.32, and I50.33). Coding of diastolic vs systolic heart failure was based on the discretion of coding physicians taking together ejection fraction and clinical scenario. Defining diastolic heart failure includes patients who present with heart failure and have at least had a normal or low normal ejection fraction, defined as an ejection fraction of over 50% based on the discretion of coding clinicians. An ejection fraction of 40–49% is the gray zone that could have also been coded as diastolic heart failure. ICD-10 codes regard patients who were admitted with acute heart failure. Heart failure was coded as primary and secondary diagnoses of heart failure during index admission. Cardiogenic shock had separate codes not included in this study. These patients were categorized according to ICD-10 diagnosis codes used for the following body weights: cachexia (R64), overweight (E66.3), obesity (E66.9, E66.8, E66.0), and morbid obesity (E66.01, E66.2). These ICD-10 codes correspond to different body mass indexes (BMI) and are defined as follows: <18.5 is considered underweight (cachectic), 18.5 to <25.0 is considered normal weight, 25.0 to <30.0 is considered overweight, 30.0 to <40.0 is considered obese, and >40.0 is considered morbidly obese. Comorbidity data was collected on these patients also using ICD-10 diagnosis codes (Table 1) for the following comorbidities of HF: diabetes, hypertension, chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), ST elevation myocardial infarction (STEMI), non-ST elevation myocardial infarction (NSTEMI), and old myocardial infarction (MI). Patient demographics and details of their hospital visits were additionally collected including age, length of stay, total charges, sex, race, insurance, hospital bed size, ownership of hospital, hospital location and region, and median household income (Tables 2 and 3). NIS database is publicly available without any patient identifiers exempt from IRB.

Table 1.

ICD-10 diagnosis codes used in searching the NIS database.

SHF	I50.2, I50.20-I50.23
DHF	I50.3, I50.30-I50.33
Cachexia	R64
Overweight	E66.3
Obesity	E66.9, E66.8, E66.0
Morbid Obesity	E66.01, E66.2
Smoking	F17.20, Z72.0, Z87.891
Diabetes	E08-E13
Hypertension	I10, I11.0, I11.9, I120, I129, I13.0, I13.10, I13.11, I13.2, I15.0, I15.1, I15.2, I15.9, I16.0, I16.1, I16.9
COPD	J41.0, J41.1, J41.8, J42, J43.0, J43.1, J43.2, J43.8, J43.9, J44.0, J44.1, J44.9, J47.0, J47.1, J47.9, J684
CKD	I13.11, I13.2, N289, Q613, N181, N182, N183, N1830, N1831, N1832, N184, N185, N186, N189, R880, N19
STEMI	I21.01, I21.02, I21.09, I21.11, I21.19, I21.21, I21.29, I21.3, I21.9, I21.A1, I21.A9, I22.0, I22.1, I22.5, I22.9
NSTEMI	I21.4, I22.2
Old MI	I25.2

(COPD = chronic obstructive pulmonary disease, CKD = chronic kidney disease, MI = myocardial infarction, STEMI = ST elevation myocardial infarction, NSTEMI = non-st elevation myocardial infarction).

Table 2.

SHF patient demographics and details of their hospital visit.

2016–2020, age >20	Total SHF	Normal Weight	Cachexia	Overweight	Obesity	Morbid Obesity
Total Population	7,364,023	5,911,238	110,755	30,640	679,350	639,765
Age (Mean±SD)	69.25 ± 14.12	70.59 ± 13.92	72.51 ± 13.43	67.80 ± 13.72	65.03 ± 13.09	60.78 ± 13.25
LOS (Mean±SD)	6 ± 8	6 ± 8	10 ± 13	6 ± 6	6 ± 7	7 ± 7
Total Charges $ (Mean±SD)	84920 ± 151322	84134 ± 153412	114549 ± 206703	81254 ± 116750	85501 ± 131882	86917 ± 140842
Mortality	5.34%	5.60%	15.22%	2.56%	3.12%	3.70%
Gender
Male	63.76%	81.01%	1.48%	0.44%	9.21%	7.95%
Female	36.24%	78.97%	1.54%	0.37%	9.25%	9.98%
Race
White	66.49%	80.97%	1.46%	0.41%	9.14%	8.12%
Black	19.85%	76.78%	1.68%	0.42%	9.88%	11.37%
Hispanic	8.33%	80.44%	1.26%	0.50%	9.63%	8.29%
Asian/Pacific Islander	2.13%	87.20%	2.19%	0.48%	5.64%	4.53%
Native American	0.59%	80.30%	1.42%	0.24%	8.81%	9.27%
Others	2.61%	82.64%	1.65%	0.43%	8.21%	7.18%
Primary Payer
Medicare	69.09%	82.64%	1.66%	0.39%	8.23%	7.18%
Medicaid	11.97%	75.11%	1.36%	0.48%	10.41%	12.78%
Private including HMO	13.56%	73.95%	1.00%	0.48%	12.60%	12.10%
Self-Pay	2.89%	75.65%	0.90%	0.43%	11.41%	11.74%
No Charge	0.23%	75.17%	0.95%	0.53%	12.24%	11.29%
Other	2.26%	79.49%	1.39%	0.62%	10.02%	8.57%

(LOS = length of stay, SD = standard deviation, HMO = health maintenance organization).

Table 3.

DHF patient demographics and details of their hospital visit.

2016–2020, age >20	Total DHF	Normal Weight	Cachexia	Overweight	Obesity	Morbid Obesity
Total Population	10,064,223	7,152,273	106,815	37,230	1,091,110	1,694,205
Age (Mean±SD)	73.64 ± 12.78	75.91 ± 12.26	77.33 ± 11.75	74.21 ± 12.26	70.84 ± 11.84	65.55 ± 11.88
LOS (Mean±SD)	6 ± 7	6 ± 7	9 ± 11	6 ± 6	6 ± 6	7 ± 7
Total Charges $ (Mean±SD)	70735 ± 107650	69821 ± 108281	98607 ± 183102	68878 ± 92710	69443 ± 92654	73823 ± 107626
Mortality	4.16%	4.58%	14.25%	2.08%	2.43%	2.93%
Gender
Male	41.02%	72.08%	1.03%	0.41%	11.13%	15.51%
Female	58.98%	70.36%	1.08%	0.34%	10.64%	17.75%
Race
White	72.63%	71.97%	1.04%	0.36%	10.69%	16.10%
Black	15.80%	64.97%	1.09%	0.38%	11.94%	21.85%
Hispanic	6.98%	70.94%	0.93%	0.42%	11.46%	16.44%
Asian/Pacific Islander	2.03%	84.25%	1.93%	0.42%	6.63%	6.84%
Native American	0.48%	68.05%	1.07%	0.17%	10.73%	20.10%
Others	2.08%	74.11%	1.18%	0.39%	10.24%	14.24%

(DHF = diastolic heart failure).

2.2. Statistical analysis

Patient demographic, clinical, and hospital characteristics are means with standard deviation for continuous variables and proportions, and 95% confidence intervals for categorical variables. Multivariable logistic regression was performed for variable significance in univariate analysis to adjust the odds of clinical outcomes relative to patient and hospital characteristics as well as ascertain the odds of clinical outcomes over time. All analyses are conducted following the implementation of population discharge weights. All p-values are two-sided, and p < 0.05 is considered statistically significant. Data was analyzed using STATA 17 (Stata Corporation, College Station, TX).

3. Results

A total of 7,364,023 patients with a diagnosis of SHF were stratified by weight groups and demographics, with a total all-cause inpatient mortality rate of 5.34% (Table 2). A total of 10,064,223 patients with a diagnosis of DHF were stratified by weight groups and demographics, with a total all-cause inpatient mortality rate of 4.16% (Table 3). Demographic trends were observed, particularly in the distribution of weight groups among race (Figures 1–3). For both SHF and DHF, all-cause inpatient mortality was lowest in overweight patients, followed by those with obesity and morbid obesity, whereas mortality was highest in those with cachexia when compared to normal weight (Table 4).

Figure 1.

SHF demographic trends in the distribution of weight groups with race. Normal weight category omitted to better visualize trends of other groups (SHF = systolic heart failure).

Figure 2.

DHF demographic trends in the distribution of weight groups with race. Normal weight category omitted to better visualize trends of other groups (DHF = diastolic heart failure).

Figure 3.

SHF all-cause inpatient mortality rates among weight groups (SHF = systolic heart failure).

Table 4.

Table 4a shows all-cause inpatient mortality based on weight categories in patient with systolic heart failure (SHF). Table 4b shows all-cause inpatient mortality based on weight categories with diastolic heart failure (DHF).

	Mortality	Normal weight Mortality	Odds Ratio (95% C.I.)	P-Value
SHF
Cachexia	15.22%	5.60%	3.02 (2.91–3.14)	<0.001
Overweight	2.56%	5.60%	0.44 (0.38–0.52)	<0.001
Obesity	3.12%	5.60%	0.54 (0.52–0.56)	<0.001
Morbid Obesity	3.70%	5.60%	0.65 (0.63–0.67)	<0.001
DHF
Cachexia	14.25%	4.58%	3.47 (3.32–3.61)	<0.001
Overweight	2.08%	4.58%	0.44 (0.38–0.52)	<0.001
Obesity	2.43%	4.58%	0.52 (0.50–0.53)	<0.001
Morbid Obesity	2.93%	4.58%	0.63 (0.62–0.64)	<0.001

For both SHF and DHF, this relationship persists after multivariate analysis for comorbid conditions and baseline characteristics, as evident by the adjusted mortality odds ratios of each group in comparison to normal weight. Mortality odds ratios of comorbidities in each group additionally show little variation in comparison to each other, indicating a lack of a confounding effect secondary to a comorbid condition (Figure 4 and 5).

Figure 4.

DHF all-cause inpatient mortality rates among weight groups (DHF = diastolic heart failure).

Figure 5.

Panel a shows SHF mortality odds ratios of comorbidities (SHF = systolic heart failure). Panel b shows DHF mortality odds ratios of comorbidities (DHF = diastolic heart failure).

4. Discussion

Cachexia can manifest in a variety of diseases, including HF (termed cardiac cachexia in this context), and is related to increased morbidity and mortality [15,16] Our findings support this and it reinforces the external validity of our study, but more importantly, cardiac cachexia may provide an explanation for the underlying physiology of the obesity paradox in CVD. It has been shown that in the context of altered levels of endocrine mediators such as insulin, insulin-like growth factor 1, leptin, ghrelin, melanocortin, growth hormone, and neuropeptide Y that lead to cardiac cachexia, cardiac obesity actually plays a protective role [15,16]. That is, visceral obesity is thought to mitigate the increased morbidity and mortality that would normally be witnessed in these patients secondary to cardiac cachexia. Furthermore, adipose tissue has been shown to be beneficial to cardiac cachexia in more than one way [6,17]. It follows that those who have a higher fat/lean body mass ratio may develop cardiac cachexia less, and have lower mortality rates in HF. This further highlights the paradoxical and multifaceted cardio protection that excess fat provides in the context of HF, as it mitigates cardiac cachexia, which is linked to significantly increased mortality in these patients.

Our data is consistent with these hypotheses and displays the weight at which excess adiposity is the most cardioprotective while still minimizing the detrimental effects of obesity. However, those with more adiposity than the overweight group face the detrimental effects that excess adipose tissue exerts on their bodies. Still, HF patients who are obese and morbidly obese have lower all-cause inpatient mortality rates than those of normal weight, suggesting that the benefit of cardio protection outweighs the cost of the negative metabolic effects of excess adiposity on the body in these populations.

Conversely, some have hypothesized that the patterns observed in the obesity paradox may not be secondary to excess fat mass playing a protective role, but instead are related to levels of lean mass [5,6,13,17–19]. In this context, lean refers to fat-free mass, such as skeletal muscle. A deficiency in lean mass, also known as sarcopenia, is independently associated with a poor prognosis in numerous chronic diseases, including HF [20,21]. If this was the case, findings would likely manifest with patients of lower BMI having worse outcomes in comparison to those with higher BMI. This may be due to the relationship between skeletal muscle mass and cardiorespiratory fitness (CRF) [14,22], defined by the minute ventilation divided by carbon dioxide production (/) slope. This is a measure that quantifies ventilatory efficiency and a lower V_E/ V_CO2 slope has been shown to correlate with better outcomes in HF patients, irrespective of body habitus [13]. It follows that higher amounts of lean mass exert protective effects on patients with HF, and/or the detrimental effects of sarcopenia manifest as increased mortality in lower BMI categories. This also suggests that among two HF patients of similar weights in the obese category of BMI, with one being nonsarcopenic obese and the other being sarcopenic obese, the former would have higher CRF and better outcomes [14].

Our data is also consistent with this idea of skeletal muscle’s role in CRF, and further ties in with the idea that cardiac cachexia plays a significant role in HF mortality. Still, these competing hypotheses function on different ideologies regarding the underlying physiology of the obesity paradox. Is it increased fat mass, increased lean mass, or a combination of the two that are protective from mortality in HF? This question sheds light on the primary limitation of those studying obesity in all fields of research, including our own HF study. We used ICD-10 diagnosis codes, which function on a basis of BMI, to identify and categorize our patients. BMI is a measure that has widespread standardized use in healthcare and therefore utilizing it has allowed us to identify and categorize an extremely large number of HF patients. However, this comes at a cost, since BMI does not distinguish between variations in body composition (fat mass versus lean mass) and fat distribution (subcutaneous fat versus visceral fat) [6,18,19,23]. As a result, there may be patients with different body compositions and, more importantly, different metabolic profiles, who get placed into the same BMI group. However, due to the variation in body composition within BMI groups, it has been argued that there are more effective measures of adiposity such as waist circumference, waist-to-hip ratio, body fat percentage, and fat mass index [3,6,24]. These alternative measures are not nearly as commonly used in medicine as BMI is. Furthermore, they require more time with patients, training of healthcare workers, and increased cost of visits [5]. As such, this may introduce challenges to those studying obesity in any field, as most patients obtain a BMI measurement when they are receiving healthcare, and very few obtain a measurement with one of the alternative techniques mentioned above. Even more so, it may be warranted to obtain measures of lean mass alongside these alternative measures of fat mass to obtain a more complete picture of the underlying physiology of the obesity paradox.

Although some have criticized the use of BMI as a predictor of mortality in CVD, it has been demonstrated that at the population level, BMI still predicts clinical outcomes and that BMI can be as clinically important as the other total adiposity measures [25]. In fact, it has also been shown that these other measures of fat mass still demonstrate the obesity paradox.⁵ Given these findings, BMI will likely continue to be the gold standard of body composition assessment due to its utility, simplicity, widespread adoption in literature and healthcare, and endorsement by groups such as the World Health Organization [2,25]. Overall, in our own study, we have demonstrated a strong pattern of inpatient mortality rates in HF patients of varying BMI, but we are still unable to say with confidence what the underlying physiology of the obesity paradox is to explain our findings. Now that the pattern of the obesity paradox has been mapped out by the largest study to date, further prospective studies with other measures of adiposity and lean mass are warranted to analyze the physiological relationship of HF and obesity and distinguish which metabolic processes are at play to generate these patterns. Obesity is an incredibly complex disease, and the pattern witnessed in the obesity paradox is likely multifaceted with physiology that cannot be explained by a single mechanism.

It should be noted that although we have confirmed and further characterized the obesity paradox in SHF and DHF, this should not be interpreted as a reason to promote habits leading to overweight or obese individuals in any population. Since obesity is a significant risk factor in the pathogenesis of many chronic health conditions, including CVD and specifically HF, individuals who endorse healthy habits are practicing primary prevention of the development of HF in the first place.

Lastly, by analyzing the demographics of our study, particularly race and median household income, we observe a clear example of the social determinants of health and take this opportunity to advocate for more equitable healthcare. It is apparent that White and Asian/Pacific Islander populations make up a larger proportion of the normal weight category, whereas those who are Black or Native American make up a much larger proportion of the obese and morbidly obese categories. This is the pattern we expect to see [26–29], serving as a source of external validity of our study, but more importantly highlights the importance of equitable healthcare. It is critical to be mindful of how these communities are disproportionately affected by systemic issues, such as limited access to healthcare, less educational opportunities, food deserts, and decreased economic stability [1,26,28], which all lead to a worsened overall health of an individual. An expected pattern is also witnessed with the median household income. We see a strong inverse relationship between median household income and rates of obesity and morbid obesity. This further emphasizes just how much of a role that economic stability and healthcare access play in an individual’s overall health. It is essential that healthcare workers and researchers alike continue to advocate for more equitable healthcare to address these disparities. It is important to note that we did not find a linear improvement with increasing body weights. The lowest mortality was seen in patients with overweight and a further increase in weight increased mortality compared to overweight but not to normal weight. Therefore, our data only revealed a partial obesity paradox suggesting that beyond overweight, further weight gain has a negative effect on mortality. Based on our findings, the overweight category had the best outcome. Therefore, our results can be used to motivate obese and morbid obese patients to lose enough weight to at least reach an overweight status that is much easier to achieve than obtaining a normal weight. This easier-to-achieve goal may improve our fight against obesity in an increasingly obese society. A preprint of this manuscript is available at medRxiv [30].

5. Limitations

We used administrative ICD-10 coding with inherent limitations. For example, coding for systolic vs diastolic heart failure was done based on the discretion of the physician and ejection fractions that could cause some inaccuracy in the definition of systolic vs diastolic heart failure. Furthermore, we do not have ICD-10 coding based on ejection fraction, which could have been a more accurate way to distinguish between diastolic vs systolic heart failure. Our study was a retrospective study needing confirmation in a prospective study. We adjusted for many comorbidities, but we cannot rule out other important comorbidities that could lead to obesity paradox not adjusted in our study. We do not have markers in the NIS database limiting our study.

6. Conclusion

Our data concludes that after multivariate adjustment for comorbidities, all-cause inpatient mortality in SHF and DHF is lowest in overweight patients. Obese and morbidly obese SHF and DHF patients also have lower all-cause inpatient mortality compared to normal-weight patients but to a lesser extent than those that are overweight. Although we confirmed the obesity paradox in HF patients, we found that increasing weight above the overweight threshold reduces the beneficial effect of additional weight in this population. Patients with cachexia by far have the highest all-cause inpatient mortality, nearly tripling the mortality rates of HF patients of normal weight. Further prospective studies with other measures of adiposity and lean mass are warranted to analyze the physiological relationship of HF and obesity and distinguish which metabolic processes are at play to generate these patterns.

Funding Statement

This paper was not funded.

Article highlights

• A phenomenon known as the obesity paradox has been reported in patients with heart failure (HF).

• We used the National Inpatient Sample (NIS) database for 2016–2020.

• We evaluated mortality based on body weight in a total of 7,364,023 patients with systolic heart failure (SHF) and 10,064,223 with diastolic heart failure (DHF).

• All-cause inpatient mortality was lowest in overweight followed by obesity and morbid obesity, whereas mortality was highest in cachexia for SHF and

• This relationship remains similar after multivariate analysis

• For SHF patients: overweight OR: 0.49 (0.41–0.58), obesity OR: 0.64 (0.62–0.66), morbid obesity OR: 0.85 (0.83–0.88), and cachexia OR: 2.78 (2.67–2.90); p < 0.001

• For DHF patients: overweight OR: 0.47 (0.40–0.56), obesity OR: 0.61 (0.59–0.63), morbid obesity OR: 0.83 (0.81–0.85), and cachexia OR: 3.09 (2.96–3.23); p < 0.001).

• All-cause inpatient mortality in SHF and DHF is lowest in overweight populations followed by obese and morbidly obese populations, whereas cachexia has the highest mortality.

Author Contributions

All authors had full access to study data and an active role in writing manuscript.

Mohammad Reza Movahed: Protocol design and the study, writing, finding ICD codes

Austin Mineer: Writing the manuscript, conception of method, finding ICD codes

Mehrtash Hashemzadeh: Performing statistical analysis and writing the manuscript

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.