Sage Journals: Discover world-class research

Abstract

Objective

To investigate relationships between prognostic nutritional index (PNI) during pregnancy and risk of all-cause mortality (ACM) and cardiovascular disease (CVD) mortality in persons with gestational diabetes mellitus (GDM).

Methods

A cross-sectional study was conducted using NHANES data from 2007 to 2018, and weighted Cox regression models were established. Restricted cubic spline analysis was used to unveil associations of PNI with risk of ACM and CVD mortalities in individuals with GDM. Receiver operating characteristic curve was employed for determination of threshold value for association of PNI with mortality. Sensitivity analysis was performed to verify the stability of the results.

Results

734 GDM individuals and 7987 non-GDM individuals were included in this study. In GDM population, after adjusting for different categorical variables, PNI was significantly negatively correlated with ACM risk. Subgroup analysis showed that among GDM populations with no physical activity, moderate physical activity, parity of 1 or 2, negative correlation between PNI and risk of ACM was stronger than other subgroups. Sensitivity analysis results showed stable negative correlations between PNI and ACM and CVD mortality of total population, and between PNI and ACM of GDM.

Conclusion

In individuals with GDM, PNI was negatively correlated with ACM risk, especially in populations with no physical activity, moderate physical activity, and parity of 1 or 2. PNI = 50.75 may be an effective threshold affecting ACM risk in GDM, which may help in risk assessment and timely intervention for individuals with GDM.

Keywords

Gestational diabetes mellitus prognostic nutritional index all-cause mortality risk cardiovascular disease mortality risk national health and nutrition examination survey

Introduction

Gestational diabetes mellitus (GDM) is a glucose tolerance abnormality initially diagnosed during pregnancy,¹ occurring when maternal insulin secretion is insufficient to satisfy increased insulin demands during pregnancy.² As one of common complications during pregnancy, global prevalence of GDM is approximately 14%.³ GDM not only affects women’s health but may also become a risk factor for pregnant women developing hypertension,⁴ dyslipidemia,⁵ and atherosclerotic cardiovascular disease (CVD). Women with GDM have a two-fold higher risk of future cardiovascular events compared with women without GDM, and women with GDM are at high risk for CVD.⁶ CVD ranks third among women of reproductive age and is one of primary causes of mortality for women.⁷ Additionally, it is worth noting that GDM not only leads to enhanced insulin resistance and impaired glucose tolerance, but is accompanied by significant inflammation dysregulation,^8,9 which increases complexity of GDM prognosis management.^8,9 Therefore, early identification of key prognostic factors for individuals with GDM may help guide clinical practice and reduce their mortality risks.

Prognostic nutritional index (PNI) is a measure of each person’s immune-nutritional state, calculated based on serum albumin and total lymphocyte count, which can reflect individual’s immune inflammation level and nutritional status.^10,11 PNI can be used to predict prognoses of persons with colorectal cancer, hepatocellular carcinoma,^11,12 and CVD.^13,14 PNI is also an independent predictor of mortality risk in people with type 2 diabetes, with low serum PNI levels implicated in increased all-cause mortality (ACM) and CVD mortality rates.^10,15 However, no relevant studies have explored the association between PNI and GDM. Given this, it is of great significance to unveil relationship between PNI and prognosis in individuals with GDM. Therefore, this study analyzed association between PNI and ACM and CVD mortality in individuals with GDM through National Health and Nutrition Examination Survey (NHANES) database system in the United States. It is hoped that through a deeper understanding of the correlation between PNI and the risk of death in individuals with GDM, the overall prognosis of such patients can be better assessed, and the research gap in this field can be filled by providing insights and clinical guidance for the treatment and management of GDM and its related complications.

Methods

Study population

NHANES is a thorough survey initiative carried out in the US by National Center for Health Statistics (NCHS), aimed at collecting data on nutrition and health status of American adults and children. The survey was administered by NCHS at Centers for Disease Control and Prevention and was authorized by Institutional Review Board. Each participant provided a written informed consent to participate in survey.¹⁶ For detailed information, please refer to relevant pages on CDC official website (https://www.cdc.gov/nchs/nhanes/index.htm).

We used data from 59,842 respondents from NHANES 2007-2018, and after excluding 47,627 participants who did not meet criteria for women who were >20 years old at the time of interview and had at least one live birth experience, preliminary a total of 12,215 study subjects were obtained. Then, 825 respondents who lacked questionnaire information on lymphocyte count, serum albumin, and GDM were excluded, resulting in 11,390 participants. Further exclusion of 2669 participants with missing covariate data (age at first live birth, race, poverty-income ratio (PIR), body mass index (BMI), smoking, alcohol consumption, physical activity, and parity) was conducted. In the end, a total of 8721 respondents were included in this study, assigned into two groups according to if they had GDM, including 7987 non-GDM individuals and 734 GDM individuals. The detailed inclusion and exclusion process is shown in Figure 1.

Figure 1.

Flow chart of inclusion and exclusion.

Assessment of PNI and GDM

PNI assessed an individual’s nutritional status based on clinical biomarkers, computed by the formula: PNI = 5 × lymphocyte count (10⁹/L) + serum albumin (g/L)¹⁰. Lymphocyte count was mainly provided by complete blood cell count test, which used Beckman Coulter counting and sizing methods for measurement. Serum albumin levels were typically utilized for assessment of nutritional status, measured using bromocresol purple dye method. GDM is defined as having been informed by a doctor or health professional of having diabetes, hyperglycemia, or GDM during pregnancy.¹⁷

Covariates

The covariates included in this study were demographic factors (gender, age, race, PIR), parity, BMI, alcohol drinking, smoking, and physical activity. Among them, age represented age at the time of interview, and race was divided into Mexican Americans, other Hispanics, non-Hispanic whites, non-Hispanic blacks, and other races. PIR was grouped into low income (PIR < 1.30), middle income (1.3 ≤ PIR ≤ 3.5), and high income (PIR > 3.50) according to household income.¹⁸ Parity was divided into 1 or 2, 3, and ≥4.¹⁹ BMI was divided into <25 kg/m², 25–30 kg/m² and ≥30 kg/m², corresponding to normal weight, overweight, and obesity respectively.¹⁰ Alcohol drinking was defined as consuming at least 12 ounces of beer, 5 ounces of wine, or 1.5 ounces of spirits per year.²⁰

Smoking was classified as never, former, and now smoking. Now smoking refers to those who have smoked more than 100 cigarettes and currently smoke every day or some days. Former smoking refers to those who have smoked at least 100 cigarettes in the past but currently do not smoke. Never smoking refers to a lifetime smoking amount of less than 100 cigarettes.²¹ Physical activities were divided into three groups: none, moderate, and vigorous. Moderate physical activity was defined as tasks that resulted in mild sweating or a little rise in breathing or heart rate. Vigorous physical activity was defined as tasks that caused a large amount of sweating or a considerable increase in breathing or heart rate.²²

Determination of mortality rate

ACM rate and CVD mortality rate were determined based on records from National Death Index, with data collection up to December 31, 2019. Disease-specific mortality rates were identified using International Classification of Diseases, Tenth Revision (ICD-10). Specifically, CVD mortality rate was defined based on specified ICD-10 codes: I00–I09, I11, I13, I20–I51, or I60–I69.¹⁵

Statistical analysis

R (V4.2.2) software was utilized for all statistical analyses. The ‘tableone’ package was implemented to plot baseline tables, grouping respondents based on whether they had GDM according to characteristics of overall population, with categorical variables represented by sample size and percentage (n(%)), and continuous variables represented by mean and standard deviation (mean(sd)). The n represented unweighted sample size; n (%) represented weighted proportion; mean represented weighted mean; SD represented weighted standard deviation. PNI was stratified using weighted tertiles, adjusting for different categorical variables to build two models: Crude model without categorical variables adjustment; Model I adjusting for age at first live birth, race, PIR, BMI, smoking, alcohol drinking, physical activity, and parity. Cox proportional hazards regression models were constructed using ‘survey’ package to estimate hazard ratios (HR) and 95% confidence intervals (CI) of PNI for ACM and CVD mortality, and restricted cubic splines were used to explore association between PNI and ACM, CVD mortality in Model I. Stratified analysis of categorical variables and forest plots were conducted in Crude model; in Model I, chi-square tests were used for p values of interaction terms, with p < .05 indicating significant differences, and subgroup analysis was performed for significant categorical variables and physical activity. Receiver operating characteristic curves were employed to help make decisions on the optimal cutoff value of PNI levels in association analysis, and Kaplan-Meier survival curves were plotted for PNI with ACM and CVD mortality. Finally, sensitivity analyses were performed to verify the stability of the results.

Results

Characteristics of study population stratified by presence or absence of GDM

A total of 8721 female participants with or without GDM were included, among which 734 had GDM. Baseline characteristics of non-GDM participants are presented in Table 1. The average age of participants was 52.66 years, with an average age of first live birth at 22.91 years. Distribution of number of children born (1 or 2, 3, and 4 or more) was 53.0%, 26.7%, and 20.3% respectively. ACM rate was 6.7%, and the CVD mortality rate was 1.5%. Compared to non-GDM group, GDM group had a lower age at interview [46.10 (11.97) vs. 53.28 (15.88), p < .001], higher age at first live birth [23.96 (5.54) vs. 22.81 (5.01), p = .003], higher prevalence of obesity (BMI ≥30 kg/m²) [56.5% vs. 40.8%, p < .001], higher PNI levels [53.00 (4.81) vs. 52.37 (5.23), p = .008], and higher lymphocyte count levels [2.33 (0.76) vs. 2.16 (0.83), p < .001].

Table 1.

Characteristics of NHANES participants between 2007 and 2018.

Characters	Total	Non-GDM	GDM	p value
Overall	8721	7987 (91.4)	734 (8.6)
Age at interview(years)	52.66 (15.71)	53.28 (15.88)	46.10 (11.97)	<0.001
Age at first live birth(years)	22.91 (5.07)	22.81 (5.01)	23.96 (5.54)	0.003
Race				<0.001
Mexican American	1454 (8.4)	1286 (8.0)	168 (12.3)
Other Hispanic	975 (5.4)	895 (5.4)	80 (5.4)
Non-Hispanic White	3706 (68.9)	3433 (69.4)	273 (63.7)
Non-Hispanic Black	1760 (10.9)	1635 (11.0)	125 (9.3)
Other race	826 (6.4)	738 (6.2)	88 (9.3)
PIR				0.563
≤1.3	3140 (24.6)	2854 (24.4)	286 (26.6)
1.3-3.5	3359 (37.1)	3098 (37.2)	261 (35.7)
>3.5	2222 (38.4)	2035 (38.4)	187 (37.7)
BMI (kg/m²)				<0.001
<25	2242 (28.9)	2115 (29.8)	127 (18.7)
25-30	2576 (29.0)	2384 (29.4)	192 (24.9)
≥30	3903 (42.1)	3488 (40.8)	415 (56.5)
Smoking				0.787
Never smoking	5465 (60.7)	5006 (60.5)	459 (62.3)
Former smoking	1743 (21.7)	1607 (21.8)	136 (20.8)
Now Smoking	1513 (17.6)	1374 (17.7)	139 (16.9)
Alcohol drinking				0.548
No	3206 (28.7)	2951 (28.5)	255 (30.0)
Yes	5515 (71.3)	5036 (71.5)	479 (70.0)
Physical activity				0.135
None	2081 (27.9)	1902 (27.9)	179 (28.0)
Moderate	3134 (38.9)	2842 (38.5)	292 (42.9)
Vigorous	3506 (33.3)	3243 (33.7)	263 (29.1)
Parity				0.608
1 or 2	3970 (53.0)	3646 (53.0)	324 (53.3)
3	2410 (26.7)	2198 (26.9)	212 (24.9)
≥4	2341 (20.3)	2143 (20.1)	198 (21.7)
All-cause mortality				<0.001
No	7999 (93.3)	7287 (92.8)	712 (98.2)
Yes	722 (6.7)	700 (7.2)	22 (1.8)
Cardiovascular mortality				0.014
No	8554 (98.5)	7829 (98.5)	725 (99.4)
Yes	167 (1.5)	158 (1.5)	9 (0.6)
Albumin (g/L)	41.56 (3.30)	41.58 (3.29)	41.36 (3.43)	0.227
Lymphocyte (10⁹/L)	2.17 (0.82)	2.16 (0.83)	2.33 (0.76)	<0.001
PNI	52.43 (5.20)	52.37 (5.23)	53.00 (4.81)	0.008

Note: Categorical variables are presented as sample size and proportion (n (%)), while continuous variables are presented as mean and standard deviation (mean (sd)).

Poverty-income ratio, PIR; Body mass index, BMI; Prognostic Nutrition Index, PNI.

Relationship between PNI and ACM and CVD mortality in total study population

As outlined in Table 2, in Crude model and Model I model, PNI was associated with ACM in general population (Crude: HR: 0.91, 95%CI: 0.89-0.93; Model I: HR: 0.92, 95%CI: 0.90- 0.94) and CVD mortality (Crude: HR: 0.89, 95%CI: 0.85-0.92; Model I: HR: 0.91, 95%CI: 0.88-0.94) were negatively correlated (p < .01). After further stratifying PNI into three tertiles, compared to the first tertile (PNI ≤ 50.5), the second (PNI = 50.5-54.0, Crude: HR: 0.89, 95%CI: 0.86-0.92; Model I: HR: 0.90, 95%CI: 0.87-0.93) and the third tertile (PNI >54.0, Crude: HR: 0.91, 95%CI: 0.89-0.93; Model I: HR: 0.92, 95%CI: 0.90-0.94) were significantly negatively correlated with ACM (p < .001); similarly, compared to the first tertile (PNI ≤ 50.5), the second (PNI = 50.5-54.0, Crude: HR: 0.88, 95%CI: 0.83-0.94, p < .001; Model I: HR: 0.90, 95%CI: 0.84-0.96, p = .002) and the third tertile (PNI >54.0, Crude: HR: 0.88, 95%CI: 0.85-0.92, p < .001; Model I: HR: 0.90, 95%CI: 0.86-0.93, p < .001) were significantly negatively correlated with CVD mortality. By exploring weighted Cox regression model in Model I with restricted cubic splines, a significant overall trend between PNI and ACM risk (p < .0001) and a significant non-linear association (P-<.0001) were observed (Figure 2(A)). A significant overall trend between PNI and CVD mortality rate (p < .0001) was seen but without significant non-linear association (p = .1600 > 0.05) (Figure 2(B)).

Table 2.

Multivariate Cox regression analysis of PNI with cause-specific mortality and cardiovascular disease mortality in all populations.

Participants	Crude	Model I
Participants	HR (95% CI)	p-value	HR (95% CI)	p-value
All-cause mortality
All participants	0.91 (0.89-0.93)	<0.001	0.92 (0.90-0.94)	<0.001
PNI
Tertile1 (≤50.5)	Ref.		Ref.
Tertile2 (50.5-54.0)	0.89 (0.86-0.92)	<0.001	0.90 (0.87-0.93)	<0.001
Tertile3 (>54.0)	0.91 (0.89-0.93)	<0.001	0.92 (0.90-0.94)	<0.001
Cardiovascular disease mortality
All participants	0.89 (0.85-0.92)	<0.001	0.91 (0.88-0.94)	<0.001
PNI
Tertile1 (≤50.5)	Ref.		Ref.
Tertile2 (50.5-54.0)	0.88 (0.83-0.94)	<0.001	0.90 (0.84-0.96)	0.002
Tertile3 (>54.0)	0.88 (0.85-0.92)	<0.001	0.90 (0.86-0.93)	<0.001

Note: Crude refers to unadjusted; model I adjusts for age at first birth, race, PIR, BMI, smoking, alcohol consumption, physical activity, and Parity. Prognostic Nutrition Index, PNI.

Prognostic Nutrition Index, PNI; Gestational diabetes mellitus, GDM; Confidence Interval, CI.

Figure 2.

The restricted cubic spline plots depicting the association between PNI and the risks of ACM (A) and CVD mortality (B).

Association between PNI and mortality rate in individuals with GDM

In GDM population, correlation between PNI and ACM risk was studied using a weighted Cox model. Stratified analysis results are presented in Supplementary Table S1. In model with no categorical variables being adjusted, PNI was negatively correlated with ACM risk (HR <1, p < .05) among populations with PIR >3.5, BMI <30 kg/m2, former smoking, never smoking, non-drinking, moderate physical activity, no physical activity, hypertension, and parity of 3. Additionally, a significant interaction between parity and PNI (p < .05) was observed, as shown in corresponding forest plot in Figure 3. After adjusting for different categorical variables, in populations with no physical activity (Crude: HR: 0.80, 95%CI: 0.74-0.87, p < .001; Model I: HR: 0.63, 95%CI: 0.52-0.77, p < .001), moderate physical activity (Crude: HR: 0.85, 95%CI: 0.78-0.92, p < .001; Model I: HR: 0.86, 95%CI: 0.76-0.98, p = .019), and parity of 1 or 2 (Model I: HR: 0.21, 95%CI: 0.19-0.24, p < .001), PNI was negatively associated with ACM risk (Table 3).

Figure 3.

Relationship between PNI and ACM in categorical variables.

Table 3.

Relationship between PNI and all-cause mortality by physical activity, parity.

Participants	HR (95% CI)
Participants	Crude	p-value	Model I	p-value
Physical activity
None	0.80 (0.74-0.87)	<0.001	0.63 (0.52-0.77)	<0.001
Moderate	0.85 (0.78-0.92)	<0.001	0.86 (0.76-0.98)	0.019
Vigorous	0.96 (0.87-1.06)	0.400	0.94 (0.86-1.03)	0.200
Parity
1 or 2	1.01 (0.95-1.08)	0.700	0.21 (0.19-0.24)	<0.001
3	0.81 (0.72-0.92)	<0.001	0.83 (0.67-1.01)	0.067
≥4	0.90 (0.81-1.00)	0.056	0.90 (0.72-1.13)	0.400

Note: Crude refers to unadjusted; model I adjusts for age at first birth, race, PIR, BMI, smoking, alcohol consumption, physical activity, and Parity. Confidence Interval, CI.

In Figure 4(A), the optimal cutoff value for PNI and ACM was 50.75, with corresponding sensitivity and specificity of 0.682 and 0.692, respectively. AUC = 0.702 > 0.7 indicated a high accuracy of model. In Figure 4(B), the optimal cutoff value for PNI and CVD mortality rate was 49.750, with corresponding sensitivity and specificity of 0.556 and 0.770, respectively. AUC = 0.638 > 0.5 reflected a high accuracy of model. According to the above optimal cutoff value, PNI was stratified using median, and weighted Cox regression model for associations of PNI with ACM and CVD mortality risk was constructed (Table 4). The model results revealed that PNI was negatively correlated with ACM risk (Crude: HR: 0.91, 95% CI: 0.85-0.97, p = .003; Model I: HR: 0.90, 95% CI: 0.84-0.96, p = .001). Compared to PNI ≤50.75, PNI >50.75 significantly reduced risk of ACM (Crude: HR: 0.27, 94% CI: 0.09-0.82, p = .022; Model I: 0.25, 95% CI: 0.09-0.74, p = .012), while PNI was not significantly associated with CVD mortality risk (Crude: HR: 0.92, 95% CI: 0.82-1.04, p = .200; Model I: HR: 0.96, 95% CI: 0.80-1.16, p = .700). After adjusting for all categorical variables, with increasing survival time, individuals with GDM in PNI >50.75 group had a significantly lower risk of ACM than those in PNI ≤50.75 group (Log-rank p = .023 < 0.05) (Figure 5).

Figure 4.

Receiver operating characteristic curve analysis for evaluation of the ability of PNI to predict ACM (A) and CVD mortality (B).

Table 4.

Multivariate Cox regression analysis of PNI with cause-specific mortality.

Participants	Crude		Model I
Participants	HR (95% CI)	p-value	HR (95% CI)	p-value
All-cause mortality
Patients with GDM	0.91 (0.85-0.97)	0.003	0.90 (0.84-0.96)	0.001
PNI
≤50.75	Ref.		Ref.
>50.75	0.27 (0.09-0.82)	0.022	0.25 (0.09-0.74)	0.012
Cardiovascular disease mortality
Patients with GDM	0.92 (0.82-1.04)	0.200	0.96 (0.80-1.16)	0.700
PNI
≤49.75	Ref.		Ref.
>49.75	0.29 (0.07-1.17)	0.081	0.56 (0.11-2.96)	0.500

Note: Crude refers to unadjusted; model I adjusts for age at first birth, race, PIR, BMI, smoking, alcohol consumption, physical activity, and Parity.

Prognostic Nutrition Index, PNI; Gestational diabetes mellitus, GDM; Confidence Interval, CI.

Figure 5.

Kaplan-Meier survival curves for ACM by PNI of two categories.

Sensitivity analysis

We further validated our results through sensitivity analysis to enhance the reliability of the findings. Similar to previous research, sensitivity analysis results in the total population after excluding individuals with follow-up times less than 2 years (N = 7821) and outliers (N = 8661) showed a negative correlation between PNI and ACM as well as CVD mortality (p < .01, Table S2). Similarly, in the sensitivity analysis of the population of GDM after excluding individuals with follow-up times less than 2 years (N = 655) and outliers (N = 726), PNI was negatively associated with ACM risk (p < .05, Table S3), but no significant association was observed with CVD mortality risk (p > .05, Table S3). These analyses indicated that the negative correlations between PNI and ACM and CVD mortality in the total population, as well as with ACM in the GDM population, were stable.

Discussion

To our knowledge, this is the first cohort study to look into the possible relationship between PNI and risk of ACM and CVD mortality in individuals with GDM. After adjusting for categorical variables, we found that in women over 20 years old with at least one live birth experience, PNI levels were negatively correlated with risk of ACM and CVD mortality. Among individuals with GDM, PNI was not significantly associated with risk of CVD mortality but was significantly negatively correlated with ACM, especially in those with no physical activity, moderate physical activity, and parity of 1 or 2. Specifically, when PNI > 50.75, ACM risk in individuals with GDM was significantly reduced. Our results revealed importance of PNI as an indicator of immune-nutritional status in assessing prognosis of pregnant women, especially in individuals with GDM. The level of PNI can serve as a useful indicator for assessing ACM, providing a theoretical basis for personalized medical management strategies for this population. This finding provides important guidance for clinical practice and helps improve quality of life for GDM individuals.

The ACM and CVD mortality rates may rise in correlation with the overall population’s worse nutritional state. In comparison with other nutritional scores (Geriatric Nutritional Risk Index, Controlling Nutritional Status, and Triglycerides × Total Cholesterol × Body Weight index), PNI has the highest predictive value.²³ Meanwhile, compared with other nutritional scores, the high predictive value of PNI may involve the following mechanisms: (1) PNI can assess not only the nutritional status of the human body but also effectively reflect the body’s inflammation and immune status.^24–26 This makes PNI not just a simple nutritional assessment tool but also provides more comprehensive health status information. (2) Compared to using categorical variables in the COUNT score, using albumin and lymphocyte counts as continuous variables to calculate PNI minimizes information loss and better reflects the nutritional status of the general population, thereby improving prediction accuracy. (3) During long-term follow-up, lymphocyte count is a more stable indicator of body composition, while indices used to calculate GNRI and TCBI (weight, TC, and TG) are more susceptible to factors such as age, diet, medications, smoking, alcohol consumption, and lifestyle habits. Therefore, PNI may be the most effective indicator for predicting adverse events in the general population. Additionally, the study results also supported the conclusion in this study that PNI in women aged >20 years with at least one live birth was negatively correlated with ACM and CVD mortality risk, indicating a significant predictive role of PNI in ACM and CVD mortality in the general population.

Our study also demonstrated that ACM in GDM individuals with PNI >50.75 was significantly reduced. Similar studies supported the view of a significant non-linear association between PNI and ACM and CVD mortality in persons with type 2 diabetes.¹⁵ When PNI levels are below 53 or above 80, risk of ACM and CVD mortality is significantly increased compared to appropriate range of PNI (53-80).¹⁵ This further confirms that PNI, as a comprehensive marker integrating immunity, nutrition, and chronic inflammation, can provide a comprehensive assessment of diabetes progression and prognosis.¹⁴ Chronic inflammation can accelerate immune dysfunction and malnutrition, thereby promoting inflammation and forming a vicious cycle, leading to progression of diabetes and other related diseases.^27–29 By comparing the above studies, significant differences in predicting CVD mortality risk were found in GDM and type 2 diabetes people based on PNI. This difference may be attributed to complex metabolic disorders of type 2 diabetes, characterized by persistent hyperglycemia, which can cause endothelial cell damage and promote atherosclerosis and plaque formation.³⁰ In contrast, GDM is a temporary high blood sugar state that occurs during pregnancy and usually returns to normal levels after delivery. Therefore, compared to type 2 diabetes, GDM has a smaller impact on arterial disease, with a lower risk of CVD occurrence and progression.³¹

During pregnancy, GDM is a frequent metabolic condition, and its pathogenesis involves insulin resistance and insufficient insulin secretion caused by high blood sugar.³² PNI is calculated based on serum albumin and lymphocyte levels, with low levels of serum albumin serving as a marker of malnutrition, which is implicated in dismal outcomes in diabetes.³³ In diabetic persons, albumin synthesis is influenced by insulin reserves³⁴; animal model studies have shown that insulin therapy can restore serum albumin levels to normal within a few days.³⁵ Hemoglobin A1C as an indicator of blood sugar control is negatively correlated with serum albumin concentration in outpatient cases, implying that hypoalbuminemia may indicate insulin deficiency, which leads to hyperglycemia.^36,37 Among people with diabetes, mortality risk and the frequency of renal illness are linked to lymphocyte count, another PNI component.¹⁰ Decreased lymphocyte count may lead to decreased immune function, increasing risk of severe illness.^38,39 The occurrence of insulin resistance is also closely related to chronic low-grade inflammation and immune system dysregulation,⁴⁰ where CD8⁺ and CD4⁺ T lymphocytes can infiltrate visceral adipose tissue and stimulate M1 macrophages, producing various cytokines such as TNF-α and IL-6, leading to local and systemic insulin resistance.^41–43 Furthermore, B lymphocytes are also involved in adaptive immune responses, regulating occurrence of obesity and insulin resistance,⁴⁴ possibly serving as one of reasonable predictive indicators of insulin resistance in women with GDM.⁴⁰ Therefore, PNI index, by integrating serum albumin and lymphocyte levels, has important clinical significance in assessing adverse outcomes such as mortality risk in GDM people.

This study also found that after adjusting for different categorical variables, risk of ACM was negatively correlated with PNI in populations with no physical activity, moderate physical activity, and parity of 1 or 2. This suggests that persons with no physical activity, moderate physical activity, and low parity may benefit from high PNI, as good nutritional status and strong immune function in these populations help reduce risk of ACM. In GDM individuals with vigorous physical activity and high parity, PNI was not significantly associated with ACM. Physical activity may affect insulin sensitivity and blood pressure, thereby improving cardiovascular health and metabolic status by regulating concentrations of factors such as Adiponectin, TNF-α, IL-6, resistin, and C-reactive protein.⁴⁵ Therefore, for individuals with GDM with poor immune-nutritional status, vigorous physical activity may help reverse adverse effects on mortality risk. However, individuals with GDM with moderate to low physical activity are more likely to be impacted by adverse effects of poor immune-nutritional status on mortality risk due to insufficient exercise. In addition, parity status is also an important factor for pregnancy outcomes,⁴⁶ which can not only alter impact of demographic factors such as advanced maternal age on obstetric outcomes^47,48 but also interact with other factors, such as fetal growth restriction, becoming one of main determinants of perinatal mortality rates.^49,50 According to our research results, parity status may also influence mortality risk through its interaction with immune-nutritional status of individuals with GDM. In conclusion, it is necessary to consider factors such as physical activity and parity status when assessing ACM in GDM individuals by using PNI.

Individuals’ lives and health are seriously at risk due to GDM. To this end, this study found a significant association of PNI with ACM in individuals with GDM, providing important evidence for assessment and intervention of mortality risk. Although our investigation is a comprehensive long-term study with representative samples, and multiple potential categorical variables have been carefully considered, we must acknowledge some limitations. First, due to the cross-sectional design of this study, causal relationships cannot be determined. Although we found an association between PNI and ACM and CVD mortality, the causality of this association needs to be further validated through longitudinal studies. Additionally, the data in this study primarily relied on participants’ self-reports, which may introduce recall bias and potentially lead to result bias. Furthermore, this study relied on a single baseline measurement of serum PNI and could not assess changes in PNI values at the time of endpoint events, making it impossible to determine the impact of PNI value changes over time on health outcomes. Additionally, the lack of detailed information on the severity of GDM prevented us from fully assessing the potential impact of GDM on PNI and health outcomes. Finally, since the study results are derived from a cohort of U.S. adults with GDM, their generalizability may be limited. Future research should be conducted in diverse populations and regions to provide broader evidence.

Supplemental Material

Supplemental Material - Associations of prognostic nutritional index with risk of all-cause and cardiovascular disease mortalities in persons with gestational diabetes mellitus: A NHANES-based analysis

Supplemental Material for Associations of prognostic nutritional index with risk of all-cause and cardiovascular disease mortalities in persons with gestational diabetes mellitus: A NHANES-based analysis by Jianfang Cao, Xiao Bu, Juping Chen, Xia Zhang in Diabetes & Vascular Disease Research.

Footnotes

Author contribution

JF C conceived of the study, and participated in its design and interpretation and helped to draft the manuscript. X B and JP C participated in the design and interpretation of the data and drafting/revising the manuscript. X B and X Z performed the statistical analysis and revised the manuscript critically. All the authors read and approved the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is supported by The Second Batch of Jinhua Major (Key) Science and Technology Research Projects in 2021(2021-3-031).

Ethical statement

ORCID iD

Jianfang Cao

Data availability statement

The data and materials in the current study are available from the corresponding author on reasonable request.

Supplemental Material

Supplemental material for this article is available online.

References

Rosta

Al-Aissa

Hadarits

, et al. Association study with 77 SNPs confirms the robust role for the rs10830963/G of MTNR1B variant and identifies two novel associations in gestational diabetes mellitus development. PLoS One 2017; 12(1): e0169781. DOI: 10.1371/journal.pone.0169781.

Christensen

Rubin

Petersen

, et al. Cardiovascular and metabolic morbidity in women with previous gestational diabetes mellitus: a nationwide register-based cohort study. Cardiovasc Diabetol 2022; 21(1): 179. DOI: 10.1186/s12933-022-01609-2.

Wang

Chivese

, et al. IDF diabetes atlas: estimation of global and regional gestational diabetes mellitus prevalence for 2021 by international association of diabetes in pregnancy study group’s criteria. Diabetes Res Clin Pract 2022; 183: 109050. DOI: 10.1016/j.diabres.2021.109050.

Tobias

Forman

, et al. Increased risk of hypertension after gestational diabetes mellitus: findings from a large prospective cohort study. Diabetes Care 2011; 34(7): 1582–1584. DOI: 10.2337/dc11-0268.

Retnakaran

Connelly

, et al. The graded relationship between glucose tolerance status in pregnancy and postpartum levels of low-density-lipoprotein cholesterol and apolipoprotein B in young women: implications for future cardiovascular risk. J Clin Endocrinol Metab 2010; 95(9): 4345–4353. DOI: 10.1210/jc.2010-0361.

Kramer

Campbell

Retnakaran

. Gestational diabetes and the risk of cardiovascular disease in women: a systematic review and meta-analysis. Diabetologia 2019; 62(6): 905–914. DOI: 10.1007/s00125-019-4840-2.

Sufa

Abebe

Cheneke

. Dyslipidemia and associated factors among women using hormonal contraceptives in Harar town, Eastern Ethiopia. BMC Res Notes 2019; 12(1): 120. DOI: 10.1186/s13104-019-4148-9.

Sharma

Banerjee

Krueger

, et al. Immunobiology of gestational diabetes mellitus in post-medawar era. Front Immunol 2021; 12: 758267. DOI: 10.3389/fimmu.2021.758267.

Visiedo

Vazquez-Fonseca

Abalos-Martinez

, et al. Maternal elevated inflammation impairs placental fatty acids β-oxidation in women with gestational diabetes mellitus. Front Endocrinol 2023; 14: 1146574. DOI: 10.3389/fendo.2023.1146574.

10.

Zhang

Chen

Zou

, et al. Prognostic nutritional index as a risk factor for diabetic kidney disease and mortality in patients with type 2 diabetes mellitus. Acta Diabetol 2023; 60(2): 235–245. DOI: 10.1007/s00592-022-01985-x.

11.

Wang

Xiao

, et al. Prognostic nutritional index and systemic immune-inflammation index predict the prognosis of patients with HCC. J Gastrointest Surg 2021; 25(2): 421–427. DOI: 10.1007/s11605-019-04492-7.

12.

Tokunaga

Sakamoto

Nakagawa

, et al. Comparison of systemic inflammatory and nutritional scores in colorectal cancer patients who underwent potentially curative resection. Int J Clin Oncol 2017; 22(4): 740–748. DOI: 10.1007/s10147-017-1102-5.

13.

Zencirkiran Agus

Kahraman

. Prognostic nutritional index predicts one-year outcome in heart failure with preserved ejection fraction. Acta Cardiol 2020; 75(5): 450–455. DOI: 10.1080/00015385.2019.1661139.

14.

Cao

Wang

, et al. Prognostic nutritional index predicts acute kidney injury and mortality of patients in the coronary care unit. Exp Ther Med 2021; 21(2): 123. DOI: 10.3892/etm.2020.9555.

15.

Ning

Pan

Guo

, et al. Association of prognostic nutritional index with the risk of all-cause mortality and cardiovascular mortality in patients with type 2 diabetes: NHANES 1999-2018. BMJ Open Diabetes Res Care 2023; 11(5): e003564. DOI: 10.1136/bmjdrc-2023-003564.

16.

Dorans

Bazzano

, et al. Lifestyle behaviors and cardiovascular risk profiles among parous women by gestational diabetes status, 2007-2018. Nutr Metabol Cardiovasc Dis 2022; 32(5): 1121–1130. DOI: 10.1016/j.numecd.2022.01.012.

17.

Ciardullo

Bianconi

Zerbini

, et al. Current type 2 diabetes, rather than previous gestational diabetes, is associated with liver disease in U.S. Women. Diabetes Res Clin Pract 2021; 177: 108879. DOI: 10.1016/j.diabres.2021.108879.

18.

Stebbins

Noppert

Aiello

, et al. Persistent socioeconomic and racial and ethnic disparities in pathogen burden in the United States, 1999-2014. Epidemiol Infect 2019; 147: e301. DOI: 10.1017/S0950268819001894.

19.

, et al. Gestational diabetes, subsequent type 2 diabetes, and food security status: national health and nutrition examination survey, 2007-2018. Prev Chronic Dis 2022; 19: E42. DOI: 10.5888/pcd19.220052.

20.

Dong

Sun

, et al. Association of coffee, decaffeinated coffee and caffeine intake from coffee with cognitive performance in older adults: national health and nutrition examination survey (NHANES) 2011-2014. Nutrients 2020; 12(3): 840. DOI: 10.3390/nu12030840.

21.

Zhang

Liu

Zhang

, et al. Association between blood lead levels and hyperlipidemiais: results from the NHANES (1999-2018). Front Public Health 2022; 10: 981749. DOI: 10.3389/fpubh.2022.981749.

22.

Wang

Mukherjee

Park

. Associations of cumulative exposure to heavy metal mixtures with obesity and its comorbidities among U.S. adults in NHANES 2003-2014. Environ Int 2018; 121(Pt 1): 683–694. DOI: 10.1016/j.envint.2018.09.035.

23.

Fan

Huang

Zhang

, et al. Association of four nutritional scores with all-cause and cardiovascular mortality in the general population. Front Nutr 2022; 9: 846659. DOI: 10.3389/fnut.2022.846659.

24.

Arroyo

Garcia-Martinez

Salvatella

. Human serum albumin, systemic inflammation, and cirrhosis. J Hepatol 2014; 61(2): 396–407. DOI: 10.1016/j.jhep.2014.04.012.

25.

Sheinenzon

Shehadeh

Michelis

, et al. Serum albumin levels and inflammation. Int J Biol Macromol 2021, 184, 857-862. DOI: 10.1016/j.ijbiomac.2021.06.140.

26.

Nunez

Minana

Bodi

, et al. Low lymphocyte count and cardiovascular diseases. Curr Med Chem 2011, 18 (21), 3226-3233. DOI: 10.2174/092986711796391633.

27.

Zha

Qian

. Protein nutrition and malnutrition in CKD and ESRD. Nutrients 2017; 9(3). DOI: 10.3390/nu9030208.

28.

Hickey

Martin

. Diabetic kidney disease and immune modulation. Curr Opin Pharmacol 2013; 13(4): 602–612. DOI: 10.1016/j.coph.2013.05.002.

29.

Shang

Huang

Zhan

, et al. The relationship between the prognostic nutritional index and new-onset pneumonia in peritoneal dialysis patients. Int Urol Nephrol 2022; 54(11): 3017–3024. DOI: 10.1007/s11255-022-03233-1.

30.

Ying

Zhang

, et al. Gestational diabetes mellitus and risk of long-term all-cause and cardiac mortality: a prospective cohort study. Cardiovasc Diabetol 2024; 23(1): 47. DOI: 10.1186/s12933-024-02131-3.

31.

Zhao

Zhou

Zhao

, et al. Metformin administration during pregnancy attenuated the long-term maternal metabolic and cognitive impairments in a mouse model of gestational diabetes. Aging (Albany NY) 2020; 12(14): 14019–14036.

32.

Gorar

Alioglu

Ademoglu

, et al.

Is there a tendency for thrombosis in gestational diabetes mellitus?

J Lab Physicians 2016; 8(2): 101–105. DOI: 10.4103/0974-2727.180790.

33.

Zhang

Wang

, et al. The level of serum albumin is associated with renal prognosis in patients with diabetic nephropathy. J Diabetes Res 2019; 2019: 7825804. DOI: 10.1155/2019/7825804.

34.

Tessari

Kiwanuka

Millioni

, et al. Albumin and fibrinogen synthesis and insulin effect in type 2 diabetic patients with normoalbuminuria. Diabetes Care 2006; 29(2): 323–328. DOI: 10.2337/diacare.29.02.06.dc05-0226.

35.

Peavy

Taylor

Jefferson

. Time course of changes in albumin synthesis and mRNA in diabetic and insulin-treated diabetic rats. Am J Physiol 1985; 248(6 Pt 1): E656–E663. DOI: 10.1152/ajpendo.1985.248.6.E656.

36.

Rodriguez-Segade

Rodriguez

Mayan

, et al. Plasma albumin concentration is a predictor of HbA1c among type 2 diabetic patients, independently of fasting plasma glucose and fructosamine. Diabetes Care 2005; 28(2): 437–439. DOI: 10.2337/diacare.28.2.437.

37.

Cheng

Hsu

Cheng

. Association between serum albumin concentration and ketosis risk in hospitalized individuals with type 2 diabetes mellitus. J Diabetes Res 2016; 2016: 1269706. DOI: 10.1155/2016/1269706.

38.

Warny

Helby

Nordestgaard

, et al. Incidental lymphopenia and mortality: a prospective cohort study. CMAJ (Can Med Assoc J) 2020; 192(2): E25–E33. DOI: 10.1503/cmaj.191024.

39.

Cardoso

CRL

Leite

Salles

. Importance of hematological parameters for micro- and macrovascular outcomes in patients with type 2 diabetes: the Rio de Janeiro type 2 diabetes cohort study. Cardiovasc Diabetol 2021; 20(1): 133. DOI: 10.1186/s12933-021-01324-4.

40.

Zhuang

Zhang

, et al. B lymphocytes are predictors of insulin resistance in women with gestational diabetes mellitus. Endocr, Metab Immune Disord: Drug Targets 2019; 19(3): 358–366. DOI: 10.2174/1871530319666190101130300.

41.

Olefsky

Glass

. Macrophages, inflammation, and insulin resistance. Annu Rev Physiol 2010; 72: 219–246. DOI: 10.1146/annurev-physiol-021909-135846.

42.

Odegaard

Chawla

. Pleiotropic actions of insulin resistance and inflammation in metabolic homeostasis. Science 2013; 339(6116): 172–177. DOI: 10.1126/science.1230721.

43.

Schober

Radnai

Spratte

, et al. The role of regulatory T cell (Treg) subsets in gestational diabetes mellitus. Clin Exp Immunol 2014; 177(1): 76–85. DOI: 10.1111/cei.12300.

44.

Winer

Chng

, et al. B Lymphocytes in obesity-related adipose tissue inflammation and insulin resistance. Cell Mol Life Sci 2014; 71(6): 1033–1043. DOI: 10.1007/s00018-013-1486-y.

45.

Rubin

Hackney

. Inflammatory cytokines and metabolic risk factors during growth and maturation: influence of physical activity. Med Sport Sci 2010; 55: 43–55. DOI: 10.1159/000321971.

46.

Bai

Wong

Bauman

, et al. Parity and pregnancy outcomes. Am J Obstet Gynecol 2002; 186(2): 274–278. DOI: 10.1067/mob.2002.119639.

47.

Chan

Lao

. Effect of parity and advanced maternal age on obstetric outcome. Int J Gynaecol Obstet 2008; 102(3): 237–241. DOI: 10.1016/j.ijgo.2008.05.004.

48.

Wang

Tanbo

Abyholm

, et al. The impact of advanced maternal age and parity on obstetric and perinatal outcomes in singleton gestations. Arch Gynecol Obstet 2011; 284(1): 31–37. DOI: 10.1007/s00404-010-1587-x.

49.

Gardosi

Clausson

Francis

. The value of customised centiles in assessing perinatal mortality risk associated with parity and maternal size. BJOG 2009; 116(10): 1356–1363. DOI: 10.1111/j.1471-0528.2009.02245.x.

50.

Gardosi

Madurasinghe

Williams

, et al. Maternal and fetal risk factors for stillbirth: population based study. BMJ 2013; 346: f108. DOI: 10.1136/bmj.f108.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB