Causes of the difference in absolute values of the foF2 and hmF2 negative trends obtained by different groups are analyzed. Three causes are considered in detail: wrong choice of the initial data used to reveal trends and the absence of the allowance for seasonal and diurnal variations in the trends. The influence of the years of a deep solar activity minimum of 2008-2009 on the foF2 trends is studied. It is shown that the years of the minimum cause some distortion in the curves used to derive the trends. In general, the trends for the period till 2014 are slightly lower than for the period till 2006-2008 analyzed in the previous publications. However, after 2009 a well pronounced negative trend is seen. The time behavior of the slope k of the foF2 dependence on hmF2 is analyzed. Since foF2 depends strongly on the atomic oxygen concentration (roughly speaking it is proportional to [O]) and hmF2 depends on [O] weakly, the slope k of the foF2(hmF2) dependence should decrease with time if there is a decrease in the atomic oxygen concentration. The analysis of the foF2(hmF2) dependence on time for several inospheric stations and various seasons shows that the value of k systematically decreases from the “etalon” period (1958-1980) to later periods. That confirms the hypothesis proposed by the author earlier that the negative trends in foF2 are caused mainly by the decrease in the atomic oxygen concentration in the thermosphere.