Relationship Between Bmi And Risk Of Death And Immunological Response In Hiv-infected Patients With Antiretroviral Therapy

ObjectiveThe body mass index(BMI)is an indicator or a measure of obesity.Some previous studies have shown that BMI is associated with mortality in human immunodeficiency virus(HIV)-infected patients and recovery of CD4 cells when these patients received antiretroviral therapy(ART).However,it is unclear whether BMI is associated with mortality and recovery of immune function in HIV-infected patients after ART.Meanwhile,in China,very few studies focused on the impact of BMI on the risk of death among HIV-infected people according to the Chinese BMI standard,which has a slight difference with European and American standard.In this study,a large-scale,long-span retrospective cohort study was conducted to assess the relationship among baseline BMI,mortality and immune recovery among HIV-infected patients at the start of ART as well as throughout the treatment.Methods(1)A retrospective cohort study was conducted among HIV/AIDS patients from National Free Antiretroviral Treatment Program(NFATP)in the fourth People’s hospital of Nanning and Guigang Center for Disease Control and Prevention(CDC),who initiated ART during May 2005 to June 2016,aged 18 years or older,and had baseline and follow-up BMI data..The retrospective cohort was established by linking baseline information and follow-up infornation of HIV/AIDS patients.(2)For the effect of BMI on mortality or related risk factors,Chi-square analysis or Fisher exact test,when appropriate,was used to compare the baseline characteristics according to BMI classification at ART initiation;Cox proportional hazard ratio models were used to explore the effect of BMI on mortality and other factors affecting the risk of death;Kaplan-Meier model was used to assess the risk of death during ART in HIV-infected patients,and Log-rank test was used to compare the risk of death among different groups;In addition,1:1 propensity score matching(PSM)was used to match the baseline characteristics which were different among four BMI groups and to further verify the effects of BMI on mortality of ART in HIV-infected patients.Test level was defined as cc=0.05.(3)For the effect of BMI on immunological response or related risk factors,Chi-square analysis or Fisher exact test,when appropriate,was used to compare 1the baseline characteristics according to BMI classification at ART initiation;Cox proportional hazard models were used to explore the effect of BMI on immunological response and other factors affecting the immunological response;1:1 propensity score matching(PSM)was used to match the baseline characteristics which were different among four BMI groups and to further verify the effects of BMI on immunological response of ART in HIV-infected patients.Test level was defined as α=0.05.Results1.The association between BMI and mortality1.1 Baseline CharacteristicsA total of 5,101 eligible subjects were included in the study,with 12459.8 person-years follow-up.Among them,28.2%were underweight(BMI<18.5 kg/m2),59.7%were normal weight(18.5 kg/m2<BMI<24kg/m2),10.7%were overweight(24 kg/m2<BMI<28 kg/m2),and 1.3%were obese(BMI>28kg/m2).There were significant differences in the age,gender,marital status,CD4 cell count,initial antiretroviral treatment regimen,and WHO clinical stages among four groups when antiretroviral therapy was initiated.1.2 Analysis of mortality rates in four groupsAmong the 5,101 eligible subjects,177 died during follow-up.The mortality rate was 3.5%(177/5101).The mortality rates in underweight,normal weight,overweight,and obesity group were 5.9%(85/1439),2.7%(83/3047),1.1%(6/548),and 4.5%(3/67),respectively.There was significant difference in mortality among the four groups(P<0.001).The total cumulative mortality of the subjects was 1.4/100 person-year(95%Cl:1.2-1.6)during the whole follow-up period.There was significant difference in cumulative mortality among the four-BMI groups(P<0.001).1.3 Analysis of the effect of baseline BMI on mortality using Cox proportional hazards modelCompared with the underweight group,the hazard ratio(HR)of normal weight,overweight and obese group was 0.45(95%Cl:0.33-0.60,P<0.001),0.18(95%Cl:0.08-0.42,P<0.001),and 0.82(95%Cl:0.26-2.58,P=0.727),respectively.The cumulative mortality of underweight,normal weight,overweight and obese group was 2.4/100 person-year(95%Cl:1.9-2.9),1.1/100 person-year(95%Cl:0.9-1.3),and 0.5/100 person-year(95%Cl:0.1-0.9),and 2.4/100 person-year(95%Cl:-0.2-4.9).After adjusted the factors that had significant differences among four groups(except viral load),the multivariate Cox proportional hazard model shows that the adjusted hazard ratio of normal weight and overweight group was 0.48(95%Cl:0.35-0.66,P<0.001)and 0.23(95%Cl:0.10-0.53,P=0.001),compared with the underweight group,respectively.1.4 Analysis of other possible factors that affected mortality using multivariate Cox proportional hazard modelWe explore other possible factors that affected mortality and found there were several factors independently related to the risk of mortality,including age,gender,continuous or intermittent fever,yersinia pneumocystis pneumonia,WHO clinical stage.1.5 The result’of the propensity score matching(PSM)The 1:1 propensity score matching(PSM)showed that there was a significant difference in mortality between the normal weight and underweight(P<0.001),or between overweight and underweight(P<0.001).And there was a significant difference in cumulative mortality between the normal weight and underweight(Log rank:P<0.001),or between overweight and underweight(Log rank:P<0.001).Cox proportional hazard analysis showed that,compared with underweight group,normal weight and overweight group had a lower probability to death,with an AHR 0.47(95%CI:0.32-0.68,P<0.001)or 0.26(95%CI:0.10-0.70,P=0.008),respectively.2.The association between BMI and immunological response:2.1 Baseline CharacteristicsThis part included 1,762 eligible subjects,with a total follow-up of 3741.4 person-years.Of the 1762 subjects,465 patients(26.4%)were underweight,1094 patients(62.1%)were normal weight,181 patients(10.3%)were overweight,and 22 patients(1.3%)were obese.There were significant differences in age,gender,WHO clinical stage,CD4 cell count and antiretroviral treatment regimen at the beginning of ART among four groups(P<0.05).Of the 1,762 eligible subjects,1843 were immunological responders.2.2 Analysis of immunological response in four groupsThere were 1483 immunological responders among 1762 subjects whose viral load was controlled at the first year of ART.There was a significant difference in the immunological response rates among the four groups(P<0.001),and the immunological response rate increased along with the baseline BMI increased.And the time required for an immunological response decreased as BMI increased.The mean time required for the immunological response in the underweight,normal,overweight,and obese group was 2.3 years(2.2-2.4),and 2.1 years(2.0-2.2)2.0 years(1.9-2.1)and 1.7 years(1.3-2.0),respectively.2.3 Analysis the effect of BMI on immunological response using multivariate Cox proportional hazardThe multivariate Cox proportional hazard analysis showed that,compared with underweight group,the normal weight,overweight and obese group had a higher probability to have an immunological response.The adjusted hazard ratio(AHR)of normal weight,overweight and obese group was 1.21(95%Cl:1.07-1.38,P=0.003);1.31(95%CI:1.07-1.59,P=0.008),1.95(95%CI:1.25-3.04,P=0.003),respectively.2.4 Analysis of other possible factors that affected immunological response using multivariate Cox proportional hazard modelMultivariate Cox proportional hazard model was used to explore other factors that may be associated with the immunological response.Several factors were found to be independently associated with the immunological response,including gender,marital status,CD4 cell count,tuberculosis infection,continuous or intermittent fever,disseminated non-tuberculosis infection,cytomegalovirus infection,brain lymphoma and WHO clinical stage.2.5 The result of propensity score matching analysisAfter matching,there was no significant difference in the age,gender,WHO clinical stage,CD4 cell count and antiretroviral treatment regimen at the start of ART between the normal weight and underweight(P>0.05),or between overweight and underweight(P>0.05).Cox proportional hazards regression analysis showed that,compared with underweight group,the normal weight,overweight and obese group had a higher probability to have an immunological response.And the AHR was 1.20(95%CI:1.03-1.40,P=0.018),1.44(95%CI:1.11-1.85,P=0.005),2.26(95%CI:1.07-4.77,P=0.032),respectively.Conclusions:1.The mortality rate among HIV-1-infected subjects who received ART is related to baseline BMI.The risk of death of HIV/AIDS subjects with normal weight or overweight at the beginning of ART was lower.And the overweight HIV-1 subjects had the lowest risk of death.2.There is an association between BMI and immunological response in HIV/AIDS subjects with viral load control during ART.A higher baseline BMI is associated with a higher probability to have immunological response and a shorter immunological response time.Lower baseline BMI had a lower probability to have immunological response and a longer time to have immunological response.